The new playbook for founders: fewer employees, faster execution, smarter growth.

The lean startup methodology has entered a new phase where less truly becomes more. AI-powered startups today demonstrate unprecedented revenue-per-employee efficiency, often far exceeding traditional teams. Some lean companies now reach millions in ARR within a year, signaling a shift in startup economics.
What sets this new wave apart is the strategic use of AI to maximize output while minimizing headcount. Instead of scaling teams or chasing inflated valuations, founders are building ultra-efficient businesses with small teams. Several AI-native startups have publicly reported multi-million ARR with under 10 employees, proving the viability of this model.
AI-led lean startups now occupy a large share of global venture activity and show no signs of slowing down. This article explores how these companies achieve outsized results with minimal resources, the tech stack powering them, and why this approach signals the future of entrepreneurship.

What Is a “Lean AI Startup”?

A lean AI startup evolves the traditional lean methodology by combining iterative development with AI automation. These companies use AI to handle tasks that once required entire teams. They still follow the build–measure–learn loop but at far higher speed.

Common features include:

• Founding teams of 2–5 people
• Heavy reliance on AI for core operations
• Rapid iteration cycles
• Lower operational costs
  

By replacing early headcount with automation, these startups reduce burn, retain more founder equity, and often reach profitability sooner. This model fundamentally alters the labor-to-output ratio and represents a departure from traditional scaling.

The Tech Stack Behind Lean AI Startups

Lean AI startups rely on a combination of cloud services, open-source tools, and modular architecture.

Cloud platforms like AWS, Google Cloud, and Azure allow low-cost experimentation and scaling, aided by startup credit programs. Open-source software such as PostgreSQL, Kubernetes, and Apache Kafka provides enterprise-grade infrastructure without licensing fees. AI frameworks like TensorFlow, PyTorch, and OpenCV enable advanced development at low cost.

Modular architecture helps teams update or replace components without major rewrites. CI/CD pipelines (GitHub Actions, Jenkins) enable rapid deployment. Model libraries like Hugging Face Transformers give teams access to state-of-the-art language models with minimal engineering lift.

Case Studies: Small Teams, Big Impact

Multiple modern startups highlight the power of small teams achieving large-scale output.

Some early-stage SaaS companies have reached multi-million ARR with 5–10 employees, supported by AI automation. Companies like Super.com reached impressive ARR with relatively lean operational teams. Solo founder Pieter Levels demonstrated how automation and distribution leverage can power multi-million-dollar revenue streams.

Outside pure AI, lean principles still thrive. Melissa Wood Health generates seven-figure revenue with a small team, while GE’s FastWorks initiative showed how lean methodologies reduce costs for large-scale industrial projects.

Despite differences across industries, the pattern holds: lean operations plus technological leverage deliver outsized results.

How Lean AI Startups Shift the Founder Mindset

The rise of lean AI startups is reshaping how founders think about building companies. The shift isn’t only about adopting tools—it changes foundational assumptions about team size, cost, and speed.

Traditional founders scale by hiring early and often. Lean AI founders automate first and hire last. As coding, marketing, analytics, and content creation become partially automated, strategic thinking—not execution—becomes the bottleneck.

The build–measure–learn loop accelerates dramatically when AI handles prototyping and data analysis. This reduces the risk of launching products nobody wants. Lean teams tend to adopt cultures of ownership and continuous learning, supported by evidence-driven decision-making. The result is faster validation, more efficient growth, and earlier profitability.

Why Investors Love Lean AI Startups

Lean AI startups are attractive to investors because they offer capital efficiency, fast iteration, and high revenue-per-employee metrics. Even in tighter funding markets, AI companies continue attracting significant capital.

Public data shows AI startups often receive higher early-stage valuations due to their scalable, low-burn operating models. Automation replaces expensive headcount, turning fixed costs into variable costs and enabling flexible responses to market shifts.

This model reduces risk while preserving high-growth potential, making it appealing during uncertain economic periods.

Challenges and Limitations

Lean AI startups also face meaningful challenges.

Technical constraints remain: pushing models larger yields diminishing returns, and AI cannot fix organizational or interpersonal issues. Many AI products show early spikes followed by high churn, making traction difficult to interpret.

Hiring AI-literate talent is difficult due to limited supply. And regardless of methodology, scaling remains complex—around 75% of startups still fail.

Founders must avoid over-reliance on AI-generated insights and ensure consistent validation with real users.

The Future: Lean AI Startups in 2030

By 2030, lean AI startups may evolve into “micro-unicorns”—small human teams supported by large AI systems. These companies will compete through intelligence orchestration rather than workforce size.

AI will shift from task replacement to strategic augmentation, enhancing human problem-solving. Data will become the core driver of efficiency, powering real-time automated decision-making with human oversight.

Founders will design intelligent systems rather than large organizations, enabling small teams to achieve outputs comparable to large enterprises.

Conclusion: Entrepreneurship for Everyone

The combination of lean methodology and AI represents a fundamental shift in how companies are built. Small teams can now achieve extraordinary results with less capital and greater control.

This new founder mindset—automate before hiring, validate fast, and use evidence to guide decisions—creates a path for more people to build companies.

Lean AI startups point toward a future where small teams create massive impact, where intelligence replaces infrastructure, and where entrepreneurship becomes more accessible than ever.

Header Image from Pexels

The post Startup Success with 90% Less: A Lean Startup Guide That Actually Works appeared first on Jumpstart Magazine.

WealthRyse turns spreadsheets into rocket-fuelled AI portfolios for everyone.

Global private wealth has climbed past US$140 trillion as of 2022, and it continues to expand at about 7% annually. Yet many advisory desks are stuck in the past—shuffling spreadsheets, stitching together siloed data feeds and relying on risk tools built for a slower era. The fallout is clear: sluggish proposal times and cookie‑cutter portfolios right when investors demand Netflix‑level personalization.

WealthRyse, co-founded by David Lee, Bernard Lee, Anthony Choi and Jasmine Lee, was built to close that gap. The Hong Kong‑ and Singapore‑based startup blends neural‑network research, cloud scale and human advisors into what it calls a “bionic advisory” model. Its core mission is simple but ambitious: democratize wealth creation by giving every advisor the same super‑powered toolkit once reserved for top‑tier quant desks. 

A 2025 trophy cabinet that turns heads

Credibility matters in finance, and WealthRyse’s award haul speaks for itself. In 2025 alone, the startup won four FinanceAsia Awards this year in international categories for nonbank financial institutions: Biggest Sustainable Impact, Most DEI Progressive, Best Strategic Initiative and Most Innovative Use of Technology. The recognition spans strategy, technology, sustainability and DEI (diversity, equity and inclusion), proving that WealthRyse excels on more fronts than raw code alone.

The 2025 haul sits alongside earlier honors such as Best AI Solution for Portfolio Management (AsiaRisk 2024) and Best Client Advisory Solutions Platform (Global Private Banker Wealthtech Awards 2024). Together, these accolades give third-party weight to WealthRyse’s business model and growth.

 “At WealthRyse, innovation isn’t just about technology—it’s about reshaping the financial industry for a more inclusive, sustainable future. These awards affirm our mission to empower investors with AI-driven solutions while fostering a culture of responsible wealth creation,” notes Jasmine Lee, co-founder of WealthRyse.

Inside Genisys AI—the startup’s secret sauce

At the heart of WealthRyse’s offering is Genisys AI, a cloud, subscription‑based service that feels comfortably familiar—think a Bloomberg-style interface—yet runs on a cutting-edge AI engine. The B2B2C portfolio management solution is powered by cutting-edge graph neural network technology originating from the intellectual property of Bernard Lee, a former Managing Director in the Portfolio Management Group at BlackRock in New York. It is further developed thanks to over a decade of R&D grant support provided by Singapore’s National Research Foundation and its sister agencies, and is now commercialized by WealthRyse.

Advisors and asset managers can log in and manage multiple asset classes across currencies and languages with ease. Powered by proprietary AI algorithms and comprehensive data lakes, the platform generates real-time rebalancing strategies and analytics-driven investment reports, achieving consistent out-performance that has been scientifically validated in publications of Springer-Nature, Association for Computing Machinery, American Statistical Association and Institute of Electrical and Electronics Engineers.

Genisys AI constructs its contents using proven and fully-documented investment analytics and then makes them human friendly with language models. This approach is fundamentally different from language models that sound “roughly right” but give a different recommendation 5 seconds later, a phenomenon known as “AI hallucination” that is frowned upon by regulators.  WealthRyse’s tractability has been a key factor in its award-winning success, setting it apart from competitors and ensuring its continued reliability to users subject to stringent regulations.

Financial forecast: Rocket-ship growth ahead

Projections by Wealthryse paint a striking picture: management expects revenue to leap from US$2 million in FY 25 to more than US$33 million by FY 28, a compound annual growth rate of more than 150%. Wealthryse also expects that most of the expansion will come from recurring subscription fees, with distribution-partnership revenue adding further lift. By 2028, WealthRyse anticipates more than US$20 million in annual recurring revenue (ARR), a trajectory that would place the company firmly in top-quartile SaaS territory for both growth and margins.

“These numbers aren’t just projections—they’re signals,” says co-founder Anthony Choi. “Signals that our platform is changing how wealth is built, accessed and scaled globally. ARR isn’t just a metric—it’s a measure of trust, loyalty and lasting value.”

Why WealthRyse matters in a crowded fintech field

The history of fintech is littered with flashy dashboards that never moved the needle on client outcomes. WealthRyse stands apart by combining deep neural-network science, enterprise-grade infrastructure and a mission that extends beyond quarterly earnings. The true proof of the model shows up in faster proposals, happier clients and a tangible step toward narrowing the global wealth gap. Awards are nice; impact is better. Keep an eye on WealthRyse—the journey is only beginning.

Also read:

• WealthRyse-ing: Asia’s Next AI FinTech Unicorn?
• [Press Release] WealthRyse Wins Most Innovative Use of Technology at FinanceAsia Awards 2024

Header image by WealthRyse.Ai

The post WealthRyse 2025 Triumphs Prove Genisys AI Leads the Future of Wealth Tech appeared first on Jumpstart Magazine.

Play stupid games and win stupid prizes! Learn how AI has prevented cheaters in esports.

In 2023, the esports market was valued at a jaw-dropping US$1.72 Billion, and is expected to grow with a CAGR of 20.7% to reach US$9.29 Billion in 2032. With massive investments in professional teams, esports tournaments are now growing bigger and their prize pool is reaching millions. 

This has significantly increased the competitive player base of these games. While all the players dream of winning the massive prize, not all of them dream it with honesty. Cheating has always been a problem in esports, but with the advent of AI-powered anti-cheat software, cheating can be significantly reduced. 

Most esports titles, such as Counter-Strike, Dota 2, League of Legends, among others, utilize dedicated anti-cheat software to prevent cheaters. The software detects and removes cheaters mid-tournament, and gamers may be penalized to ensure a level playing field for all participants.

How is AI preventing cheaters in gaming?

With the implementation of AI gaming, advanced machine-learning tools can be used to find and stop cheaters. AI-powered anti-cheat software offers a more adaptive solution compared to traditional methods, which are often bypassed by rapidly evolving cheating software. 

By utilizing machine learning, AI can analyze player behavior to identify any unusual movements in-game and further compare them with legitimate player patterns to detect discrepancies. Detecting such irregularities requires plenty of data, which has to constantly be fed to the system to ensure the bans and system flags are accurate. Along with finding suspicious player behavior, AI can be used in real-time games to weed out cheaters. 

Anti-cheat systems that use AI

Cheating remains a significant issue in the gaming industry, prompting developers to implement a range of AI strategies to address it. These strategies reflect the unique technological preference and operational requirements of each company. Below, we examine two prominent anti-cheat systems that illustrate how different developers are customizing their solutions. 

1. VACNet

Initially released for Counter-Strike in 2002, the critically acclaimed game developer Valve Software’s VAC has been an integral part of the anti-cheat space, some even looking up to it as a blueprint. VAC constantly checks the game files to look out for any third-party or unusual files, which,  if found, will result in a lifelong ban. 

Building on VAC, Valve introduced VACNet, an AI-powered system that takes a more advanced approach to combating cheaters by analyzing gameplay data rather than just scanning game files. VACNet 2.0, the prior iteration, used deep learning to examine player replays, identifying suspicious behaviors such as unnatural aiming, shooting patterns or movements that deviate from normal player actions. By comparing these patterns against a database of legitimate behaviors, VACNet 2.0 could flag potential cheaters with high accuracy.

With the rise of increasingly “undetectable” cheats that bypass traditional anti-cheat methods, Valve has released the AI-powered VacNet 3.0, which incorporates even more robust AI and machine learning models. While the system is still in development, flagged players are currently reviewed by human moderators to ensure accuracy before any action is taken. This hybrid approach allows VACNet to adapt to evolving cheat software while minimizing false bans.

Although completely eradicating cheating remains difficult, VACNet has already made significant progress. Valve has banned over 4.5 million accounts for cheating in Counter-Strike 2, their flagship first-person shooter game, and continues to refine its tools to stay ahead of cheaters.

2. Vanguard

The American game developer Riot Studios released Vanguard in 2020, their very own anti-cheat software to detect and punish cheaters in games such as Valorant and League of Legends. Vanguard uses a “kernel-level driver,” which gives it deep access to your computer’s operating system. This allows the software to monitor for cheats at all times, even when the game isn’t running, ensuring cheats can’t interfere with the game before it starts.

Vanguard combines machine learning algorithms to analyze player behavior and detect suspicious activity. If cheating is detected, Vanguard can issue penalties ranging from temporary bans to permanent account bans. Additionally, it can block specific hardware, preventing banned players from accessing their games.

AI-powered Vanguard has banned over 3.6 million accounts for Valorant in the past 4 years. This is a significant number, which, although it may sound concerning to the future of the game, is encouraging a culture of fair play and integrity. Riot Studios has always stood out with its zero-tolerance policy for cheating and malice in its games. 

Is AI the future of anti-cheat software?

AI has proven to be instrumental in the fight against cheaters in the esports and gaming industry. From detecting cheaters to auto-banning them, it has become a key element of anti-cheat software. As there are rising innovations and developments in AI and machine learning, we can only expect these software to get better. 

As more developers leverage AI and machine learning to tackle cheating, a specialized niche is emerging within the gaming industry. This shift promises a future where fair and competitive play becomes the norm, supported by stronger and more reliable anti-cheat systems. With advancements in technology, players can look forward to a more transparent and enjoyable gaming experience.

Also read

• Essential Gaming Accessories for the Ultimate Gameplay Experience
• 5 Essential Programming Languages Every Game Developer Should Master in 2024
• Paris Olympics 2024: How AI Is Changing the Game
• Essential Gaming Slang Terms for True Gamers

Header Image by Unsplash

The post AI in Esports: How Machine Learning is Transforming Anti-Cheat Systems in Esports appeared first on Jumpstart Magazine.

The investment aims to develop advanced digital infrastructure in the Middle East, Central Asia and Africa, promoting secure and equitable technology access. 

Microsoft has recently made a move in the AI realm with a hefty US$1.5 billion investment in G42, a prominent AI technology holding company based in the UAE. This investment aims to bolster the collaboration between the two entities, facilitating the introduction of Microsoft AI technologies and educational initiatives in the UAE and other global markets. Additionally, Brad Smith, Vice Chair and President of Microsoft, will join the G42 Board of Directors as part of the expanded partnership.

Enhanced collaboration and technological empowerment

The partnership will focus on enabling organizations of various sizes in new markets to utilize AI and cloud technologies while adhering to top-tier standards for safety and security. Building on their longstanding cooperation in AI and digital transformation, the investment by Microsoft will further solidify their mutual commitment to this strategic alliance. G42 will leverage Microsoft Azure to run its AI applications and services and jointly deliver advanced AI solutions to global public sector clients and large enterprises.

Regional development and infrastructure expansion

Together, G42 and Microsoft aim to extend advanced AI and digital infrastructure to nations in the Middle East, Central Asia and Africa. This initiative is designed to provide equitable access to services that address critical governmental and business needs, maintaining the highest standards of security and privacy.

H.H. Sheikh Tahnoon bin Zayed Al Nahyan, Chairman of G42, emphasized that Microsoft’s investment represents a critical moment in G42’s growth and innovation journey, highlighting the strategic alignment and shared goals of both organizations. “This partnership is a testament to the shared values and aspirations for progress, fostering greater cooperation and synergy globally,” he remarked.

Skilling initiatives and workforce development

The collaboration also includes a commitment to developing a skilled and diverse AI workforce. This includes a US$1 billion investment in a development fund aimed at nurturing developers, which will enhance innovation and competitiveness in the UAE and the broader region.

Commitment to safe and responsible AI

Brad Smith of Microsoft highlighted the partnership’s broader scope, which extends beyond the UAE to underserved nations, combining leading technology with stringent standards for safe, trusted and responsible AI. This collaboration will align closely with government regulations in the UAE and the US.

Intergovernmental assurance and compliance

The partnership is supported by a unique Intergovernmental Assurance Agreement (IGAA) developed in consultation with UAE and US governments. This agreement sets world-class best practices for secure, trusted and responsible AI development and deployment. Both companies commit to adhering to US and international regulations concerning trade, security, responsible AI and business integrity.

Peng Xiao, Group CEO of G42, noted that Microsoft’s strategic investment will significantly enhance their international market presence by combining G42’s AI capabilities with Microsoft’s robust global infrastructure. “Together, we are not only expanding our operational horizons but also setting new industry standards for innovation,” he said.

“Our investment in G42 stands as a testament to the thriving and dynamic tech landscape in the UAE and the broader region,” said Samer Abu-Ltaif, Microsoft Corporate Vice President and President, Central and Eastern Europe, Middle East and Africa. He added that Microsoft views the investment as a testament to the dynamic tech landscape in the UAE and surrounding regions. This strategic alliance is expected to spur innovation, accelerate economic growth and empower countries to advance their digital strategies by leveraging cloud and AI technologies.

Recent collaborative milestones

Over the past year, the collaboration between G42 and Microsoft has seen several key developments. These include plans to develop AI solutions tailored for the public sector and industry, introduce sovereign cloud offerings and leverage advanced AI capabilities on the Azure public cloud platform. In November 2023, Microsoft also announced the availability of G42’s Jais Arabic Large Language Model on the new Azure AI Cloud Model-as-a-Service offering, marking a significant milestone in their partnership.

Also read: 

• How Can Cities Manage Traffic Effectively with AI?
• NVIDIA Contributes to US$110 Million Initiative for AI Education in U.S.-Japan Partnership
• Samsung Galaxy AI Expands Language Support to Reach More Global Users

Header Image from Flickr 

Press release link: https://news.microsoft.com/2024/04/15/microsoft-invests-1-5-billion-in-abu-dhabis-g42-to-accelerate-ai-development-and-global-expansion/ 

The post Microsoft Invests US$1.5B in G42 to Boost Global AI and Cloud Technologies appeared first on Jumpstart Magazine.

Neuchips is focusing on inferencing chips and introducing technologies that boost efficiency in the semiconductor industry.

In 2023, the global semiconductor market faced significant challenges in 2023, with a noticeable decline in sales. According to the Semiconductor Industry Association (SIA), the market saw a decrease in sales to US$526.8 billion, marking an 8.2% drop compared to the previous year. This downturn was heavily influenced by a 37% decline in the memory sector, identified by Gartner Inc. as the segment with the largest decrease across the industry.

AI’s emerging role in market resurgence

Despite the overall downturn, the AI sector emerged as a bright spot, particularly in the second half of the year. Counterpoint Technology Market Research highlighted AI as a crucial driver of content and revenue within the semiconductor industry. This growth was propelled by the increasing use of AI-based applications across various domains, including data centers, edge infrastructure and endpoint devices.

Gartner forecasts a robust recovery led by AI technologies, with AI chips generating US$53.4 billion in revenue in 2023—an increase of about 21% year-over-year. The sector is expected to continue its rapid growth, reaching US$67.1 billion in 2024 and potentially doubling in size to US$119.4 billion by 2027. 

Ken Lau, CEO of Neuchips, highlighted the potential for AI, particularly generative AI, which could evolve into a trillion-dollar market by the 2030s, with significant investments transitioning from training to inferencing technologies. “After training data, inferencing can significantly enhance operational efficiencies,” Lau said. He further explained that AI could transform customer interactions and digital marketing strategies, leading to direct consumer engagement and sales online.

“I think there are ways that we can’t even imagine going forward. The opportunities are limitless for AI. That’s how I see it. And a big part of that is going to be inferencing, not just training,” he added.

Inferencing as a strategic focus at Neuchips

Founded in 2019, Neuchips has strategically emphasized inferencing, particularly through its recommendation engine, which is extensively utilized in data centers to enhance online shopping experiences. 

Lau stated that its initial prototype used field programmable gate arrays (FPGAs) (a type of semiconductor device) to validate the design, leading to the development of the N3000 inference chip. The chip outperforms competitors by 1.7 times per watt according to MLPerf 3.0 Benchmarking.

Lau explained that when Neuchips designed this chip, it was with the recommendation engine in mind. However, when GenAI gained momentum, the AI company tested it with its chip and achieved compatibility due to optimized memory subsystems, applicable to both recommendation engines and GenAI.

Recognition and future plans

The N3000 chip won the “Best AI Chip” award at the EE Awards Asia 2023, reflecting the company’s high execution standards. Lau emphasized the efficiency of their development process, noting that the chip was successfully completed with a single production run—a significant achievement for a startup operating with limited resources.

Industry challenges and future outlook

Despite the optimistic growth projections for AI chips, the semiconductor industry faces several challenges, particularly in software integration and power consumption. Lau remarked on the integration challenges, citing examples from healthcare where data privacy and internal training are crucial. Neuchips addresses these challenges by offering low-power chips suitable for compact spaces and providing comprehensive software support, including software development kits (SDKs) and training services.

Looking ahead, Neuchips announced advancements in their Evo series, including the Raptor Gen AI inference chip, which supports Large Language Models (LLMs) and is set to launch in Q2 2024. As the AI applications ecosystem expands, Neuchips plans to innovate continuously in chip design, focusing on various form factors to accommodate diverse uses, including PCs and low-end workstations.

Also read:

• Top 7 Companies in the Semiconductor Chip Manufacturing Industry
• Elon Musk’s Neuralink Debuts Brain Chip Implant: A Bold Future with Ethical Questions
• The World Is Facing A Semiconductor Chip Shortage, Here’s Why

 Header Image from EETimes Taiwan via PR Newswire

Press release link: https://www.prnewswire.com/news-releases/neuchips-driving-ai-innovations-in-inferencing-302118337.html

The post Neuchips Drives AI Innovation with Advanced Inferencing Technologies appeared first on Jumpstart Magazine.

Smart traffic management can redesign urban traffic flow and make rush hour less rushed.

In metropolitan areas around the world, traffic congestion is a nightmare for many. It wastes our time, ramps up our stress and exacerbates pollution. To combat this major headache, there might be a promising solution on the horizon: artificial intelligence (AI). Many urban areas are already turning to AI-powered tools like responsive streetlights and systems that predict traffic patterns to smooth out the bumps in our daily commutes. Given the adaptability of AI, these systems can learn and improve over time, refining their algorithms to meet the distinct needs of different cities.

In this article, let’s take a closer look at how AI is improving road safety and transforming traffic management to pave the way for smarter, more efficient cities.

 1. Real-time traffic management

One of the most standout uses of AI in managing city traffic involves adaptive traffic signal control systems. These systems use real-time data to adjust traffic lights based on current traffic conditions instead of rigid, pre-programmed schedules. 

In places like Delhi, India—where residents spend 58% more time in traffic than in many other cities—the adoption of AI has been a game changer. Delhi, for example, utilizes 7,500 CCTV cameras, AI-powered traffic signals and 1,000 LED lights equipped with cameras and sensors to analyze real-time traffic data. This setup helps city authorities manage traffic flow more effectively and devise long-term strategies to decrease congestion.

A screenshot of Alibaba Cloud’s YouTube video about City Brain

Beyond Delhi, many countries in Asia, including China and Malaysia, have embraced AI-powered traffic management systems. Alibaba’s City Brain, one of the most extensively deployed systems, has significantly cut down congestion. For example, in Hangzhou, China, City Brain dramatically improved traffic conditions, lowering the city’s congestion ranking from fifth to 57th. This improvement comes from using real-time information from cameras at intersections and GPS systems to optimize over 1,000 traffic signals throughout the city.

2. Forecasting traffic with AI predictive analytics

AI-backed predictive analytics tools take both historical and real-time data to predict and mitigate traffic congestion. This technology examines previous traffic trends, weather conditions and various events—from protests to accidents—allowing urban planners to adjust traffic signal timings and reroute traffic ahead of time effectively.

Using platforms like mobile apps, digital signage and traffic advisory systems, authorities can give commuters early warnings about potential congestion hotspots. This proactive communication helps travelers plan their journeys better, making informed choices about their routes and travel times. Such strategic use of predictive analytics leads to smoother traffic flow and more efficient city travel management.

3. Monitoring road conditions and tackling potholes

Keeping an eye on road conditions is crucial, as potholes not only cause significant economic damage—costing U.S. drivers an estimated US$3 billion annually—but also pose a serious hazard to road users. In fact, nearly one-third of all car accident fatalities each year are due to poor road conditions, with potholes being one of the most common culprits. However, addressing these persistent road issues can be slow, and new problems may appear suddenly.

Image from Oregon City Public Works

Thankfully, AI systems equipped with computer vision are simplifying the detection and assessment of road defects. A great example is CityRover, an AI device mounted on street sweepers deployed by Oregon City, which functions like a dash camera to detect potholes. As it scans the road, it records the date, time, location and aerial photos of each pothole. This data is uploaded to the cloud, allowing city workers to efficiently track and plan pothole repairs.

Image from UK Research and Innovation

Over in the UK, a startup called Robotiz3d has taken things a step further by developing a pioneering AI robot. This autonomous vehicle doesn’t just find cracks and potholes—it repairs them. Armed with advanced detection and repair technology, it can assess the severity of road defects and seal them up before they get worse. This breakthrough is transforming road maintenance, leading to safer, smoother drives.

4. Boosting road safety through AI innovations

AI can be trained to prioritize protecting all road users and thus minimize accidents. One way to do so is by enhancing communication among connected vehicles, which is crucial during emergencies. 

For instance, systems like Emergency Vehicle Preemption (EVP) adjust traffic signals to let emergency vehicles move quickly and safely through intersections. This enhancement involves using pedestrian and bicycle counters, real-time data-gathering traffic cameras, smart streetlights and other facilities designed for vulnerable road users (VRUs). Together, these technologies create a comprehensive safety infrastructure that not only prevents accidents but also promotes smoother and safer travel for everyone.

Challenges of using AI to manage traffic 

Costs and complexity

AI-driven traffic systems come with significant upfront costs and operational complexity. These costs cover everything from upgrading networks to installing advanced vehicle technologies and setting up secure Internet-of-Things (IoT) systems. Moreover, maintaining these systems requires a workforce that is not only tech-savvy but also skilled in troubleshooting and ongoing system optimization.

Ethical implications

The extensive use of cameras and sensors in AI traffic systems also sparks significant privacy and surveillance concerns. It’s crucial for cities to comply with strict privacy laws, protect personal data vigilantly and enhance security measures such as data encryption and access controls. These steps are essential to reduce the risk of cyber-attacks and maintain public trust.

Another ethical issue is whether AI can consistently make the “right” calls. Despite streamlining many operations, AI can still exhibit errors and biases influenced by the data or algorithms it uses. Thus, the human element remains indispensable for critical decision-making.

Gaining public support

For AI traffic management to be fully integrated and effective, it must have the support of all stakeholders, including workers, designers, drivers, law enforcement and the general public. Governments have a crucial role in building this trust by addressing privacy, equity and potential biases in AI systems. Educating the public about the benefits of AI and demonstrating its positive effects are key to gaining support and ensuring the successful adoption of these technologies.

The road ahead for AI in traffic management 

As AI continues to advance, its role in traffic management is becoming increasingly essential yet complex. While AI can streamline traffic flow and improve route efficiency, we also face significant challenges, including high costs, privacy concerns and the need for ongoing oversight. Addressing these issues is crucial to ensure a balanced approach. By focusing on both the advantages and the challenges, we are moving closer to creating truly intelligent urban spaces that are not only efficient but also protect and earn the trust of their residents.

Also read:

• The Era of Flying Cars Begins: Alef Aeronautics Secures Special Airworthiness Certification
• TomTom and Microsoft Introduce AI-Powered Automotive Assistant
• Malaysia-Based Online Car Seller Carsome Secures US$200 Million in Latest Fundraising Round

Header Image from Freepik

The post How Can Cities Manage Traffic Effectively with AI? appeared first on Jumpstart Magazine.

Rockwell Automation and Microsoft’s latest AI advancements promise a more efficient and sustainable industrial future.

Rockwell Automation, Inc. and Microsoft Corp. have announced an expanded partnership to drive advancements in industrial automation design through generative artificial intelligence (AI). This collaboration seeks to empower the workforce and accelerate the development process for customers developing industrial automation systems.

Addressing industry challenges

Industry experts recognize the current skilled labor shortage as a critical challenge for industrial entities, with automation projects growing in complexity and demand. Matthew Littlefield, President of LNS Research, notes the potential of generative AI to enhance workforce productivity and sees the Rockwell-Microsoft alliance as a pivotal development in tackling structural challenges within the industry.

AI-driven industrial automation

Rockwell and Microsoft are leveraging AI to streamline a range of industrial roles, from decision-making to engineering and operations, aiming to simplify customer workflows and increase worker efficiency. Rockwell Automation’s Chairman and CEO, Blake Moret, emphasized the companies’ joint efforts to not only meet current industry demands but also to pioneer future technological advancements in industrial automation.

A key initiative from this collaboration is the incorporation of Microsoft’s Azure OpenAI Service into Rockwell’s FactoryTalk Design Studio. This tool will enable engineers to generate code with natural language prompts, streamlining design tasks and boosting efficiency. The technology also aims to fast-track development for experienced engineers and provide a more effective learning curve for newcomers, as well as offering improved access to extensive informational databases for further education.

Extending AI to wider industrial challenges

Looking ahead, Rockwell and Microsoft anticipate expanding this integrated technology to tackle broader industrial issues such as Quality Management, Failure Mode Analysis and training frontline workers. The goal is to develop chat-based collaboration tools and AI-driven chatbots that can assist in manufacturing processes, further enhancing operational productivity.

Judson Althoff of Microsoft emphasized the growing importance of AI in the industrial sector and Microsoft’s role as a trusted partner in this technological evolution. This fusion of Rockwell’s industrial know-how with Microsoft’s AI expertise promises to accelerate the creation of advanced control systems and catalyze innovation across industries.

A commitment to the industrial metaverse

Additionally, both companies are committed to fostering innovation within the industrial metaverse, utilizing IoT capabilities, cloud datasets, simulations and AI to design and build products in a way that is more effective, efficient and sustainable.

Also read: 

• What is AI and ML Development: Harnessing the Power of Artificial Intelligence and Machine Learning
• Drawing without a Paintbrush: An Evolution of Artificial Intelligence
• Top 3 Free Business Intelligence Tools of 2023

Header Image Courtesy of Pexels 

Press release link: https://news.microsoft.com/2023/10/26/rockwell-automation-and-microsoft-expand-partnership-to-leverage-generative-ai-capabilities-for-enhanced-productivity-and-faster-time-to-market/

The post Rockwell Automation and Microsoft Strengthen Ties to Bolster Industrial Automation with Artificial Intelligence appeared first on Jumpstart Magazine.

AI can play a crucial role in transforming such a sensitive field. Here’s how!

The beginning of the COVID-19 pandemic catalyzed a major shift towards telehealth, where people can access healthcare through telecommunication methods, like video calls. As families stayed home to prevent catching a disease that was largely unknown at the time, many had no choice but to turn to electronic health-related services, among them mental health therapy. Simultaneously, increased isolation-induced anxiety and depression aroused additional demand for therapy, swirling up the debate of how modern technology could reinforce mental health therapy. 

As mental health therapy became available online, the world turned its attention to the role artificial intelligence (AI) could play in improving the field. Here are some ways it could do so:

1. Training mental health professionals

AI can be used to train mental health professionals, thus reducing the risk of inexperienced counselors providing incorrect treatment to at-risk clients. For instance, the AI-powered Trevor Project, a suicide prevention and crisis intervention organization for young LGBTQ members, partnered with Google to launch a crisis contact simulator and counselor training tool. The simulator models digital conversations with LGBTQ youth in crisis, allowing counselors in training to first undergo practice conversations before taking on real-life ones, where the risk of endangering a client is far greater. It also allows for more flexibility in training counselors or volunteers remotely, opening up more opportunities for aspiring counselors to work in a highly required field.

2. Interpreting data

Using AI in mental health care can allow the technology to obtain valuable insight from colossal amounts of data, which humans may not necessarily be able to do. While there have been many innovations in mental health care, the Director of Digital Psychiatry at Beth Israel Deaconess Medical Center, John Torous, explained, “Intuitively, we know that these tools contain important information… but it’s been very hard to unlock that data for clinical insights”. Compared to human workers, AI can comb through heaps of data with greater accuracy and efficiency.

3. Chatbots 

Chatbots and virtual assistants are an additional way for AI to fortify mental health care and provide support to marginalized groups. While not nearly as effective as in-person therapies, chatbots can interact with individuals 24/7 in real-time and at no cost, expanding access to mental health support. They also allow clients space to sustain sensitive conversations that may bring them discomfort in the presence of another human. Researchers and developers continue to innovate virtual assistants as both standalone treatments and as a supplement to traditional counseling. 

4. Mental illness detection

AI may also be able to detect mental illnesses, like anxiety, through behaviors that may otherwise slip through the cracks. Researchers involved in a small study in Karachi identified specific behaviors that detect anxiety, including nail-biting, knuckle-cracking and hand-tapping, which can be recognized with motion sensors and deep learning techniques. The study as a whole delivered 92% effectiveness, and even though not very all-encompassing, it gives us considerable insight into how AI can be further developed in the future. 

5. Supporting mental health professionals

Right now, clients usually undergo therapy alone in a room with a single therapist. After training, most therapists are not further supervised, and thus it could be difficult for them to make the “right” call. In this case, AI can listen into those sessions and give professionals feedback on the client’s performance, like how much a person talked or what intervention strategies were used, without breaking therapist-patient confidentiality. Of course, therapists who want to use AI as a tool in this department would require consent from their clients, which may pose a challenge.

The future of AI and mental health care

While AI may help streamline the process of training mental health professionals, identifying patterns in data and recognizing mental health issues in otherwise undiagnosed patients, it will never truly replace traditional therapy. As data and AI Product Manager at the Trevor Project Kendra Gaunt said, “There will always be a need for human-to-human connection.” 

While a machine can draw on the knowledge gained from data and learning, it lacks the human emotions and feedback most people appreciate during treatment. Gaunt continues, “AI’s role in this space shouldn’t be to replace humans, it should be to support them.” Still, the future for AI in mental health care is uncertain. With many exciting prospects on the horizon, it is important to remember that, ultimately, AI is a tool to streamline the ease with which humans can conduct their work.

Also read: 

• Mental Health Initiatives to Watch in 2022
• Introducing Mental Health First Aid in the Workplace
• How AI Can Help Solve the Growing Mental Health Crisis
• Interview with Kai.ai’s Founder Alex Frenkel: How AI Can Be a Game Changer in Mental Health Support for Young People
• Mental Health Tips for When You Get Back in the Office
• Tackling the Mental Health Challenges of Working from Home
• Why We Need To Talk About Founders’ Mental Health
• Week-Long Mental Health Break to Cope with Employee Burnout

Header Image by Unsplash

The post How AI Is Improving Mental Health Therapy appeared first on Jumpstart Magazine.

Is life too short to learn a new language? VR might be a gamechanger in second language acquisition.

Bilingual speakers make up approximately 43% of the world’s population. It is almost guaranteed that you have previously spoken with someone who is competent in speaking multiple languages. You might wonder, how do they do it so effortlessly? If you have had experiences learning a new language, you know it is the opposite of effortless. The process is painful, exhausting and frustrating—so frustrating that many of us might just give up because there are too many grammar rules to memorize. So kudos to you if you have mastered various languages! But what if there’s a simpler way to acquire a language… Well, virtual reality (VR) might be your answer. Here’s how:

Providing an immersive language-learning environment

We all know it is vital to immerse yourself in a language if you want to speak the language fluently. Through interactions with native speakers, learners can receive many language development opportunities, such as listening, speaking and feedback. However, due to geographical constraints, learners may not have the opportunity to communicate with native speakers of their target language. VR technology is thus an excellent solution to this problem, as it can provide users with an immersive language-learning environment to up their language learning game.

VR technology allows learners to enter the virtual world and interact with other people around the globe without traveling to another country. For language learners, this means they can surround themselves in an “authentic” language context, thereby boosting their reading, listening and speaking skills. VR platforms have also been proven to increase learners’ overall language abilities across all linguistic levels, making VR technology truly beneficial for second language acquisition (SLA).

Motivating language learners 

Apart from interaction and immersion, the learner’s motivation is another crucial factor in learning a new language. To effectively learn a second language, learners need to be motivated by their goals, such as to achieve fluency in English to live in an English-speaking country (an instrumental motivation) or to become a target community member (integrative motivation).

Counter to our assumption, studying a language with textbooks in the classroom is not the only way to achieve these goals. After all, we can’t be purely incentivized by our own set of goals; the means of learning a language needs to be motivating and intriguing as well, and gaming can do just that. 

As per a study, playing commercial digital video games can increase learning motivation in SLA. The reason is simple: it is more engaging and interesting than memorizing vocabulary from a textbook. With video games, players can pick up and practice everyday language rather than complicated, academic-oriented vocabularies through communicating with other players. Having the opportunity to use the language in such a setting will encourage learners to speak and use the language more confidently. 

However, this is not to say that we should replace textbooks with video games. As per the same study, classical educational setting is still necessary in SLA, and the increase in language learning motivation as a result of playing video games can transfer to classroom learning. In other words, video games can facilitate language learning at schools, even if language learning is not the games’ primary purpose.

With the rising popularity of VR technology, we can expect more and more VR games and headsets will be released in the market. Compared to commercial video games, these VR games can offer an even more engaging and aesthetic gaming experience, such as better graphics and audiovisual effects,  which can further enhance motivation. If VR games are to be integrated into language classes, maybe we will be seeing fewer students dozing off during class and more being excited to go to class!

Examples of VR language learning apps

In 2021, the global language learning market was worth US$14.2 billion, and it is projected to surpass US$28.5 billion by 2028. Surely, we will be seeing more VR language learning apps in the market soon. For now, if you are learning a new language or considering doing so, here are three apps you should check out :

ImmerseMe

This VR platform helps learners increase their fluency in their second language by simulating authentic and real-life scenarios, such as buying a baguette in Paris or a bento box in Tokyo, where they have to interact with native speakers. It offers more than 3,000 real-life scenarios across nine languages, i.e. English, Spanish, French, German, Japanese, Chinese, Italian, Greek and Indonesian. 

Mondly VR

The VR app which was acquired by UK education giant Pearson in April this year. Similar to ImmerseMe, the VR app provides an opportunity for users to learn more than 30 languages through interacting with a virtual language teacher. After each interaction, users will receive pronunciation feedback as well as vocabulary suggestions. It also has an augmented reality (AR) app in which their virtual language teachers can appear in your living room through your phone camera.

Panolingo 

Panolingo is a VR app that uses a gamification approach where users can collect points and bonuses from doing tasks or competing with friends. An example of said task is following instruction prompts, such as identifying what a refrigerator is in the user’s kitchen. With the gamification of the VR app, learners can be more motivated in picking up a second language via a relatively more interesting and incentivized approach. 

All in all, interaction and motivation are two vital factors of an effective SLA. VR technology’s ability to create a virtual environment that can motivate learners and provide them with lifelike and authentic interactions makes it especially valuable in the SLA market. With all the benefits of learning a language via VR platforms and the great potential of the language learning market, we will be seeing more tech solutions to help language learners around the world master as many languages as possible. 

Also read:

• Why Computers Can’t Master Human Language -for Now
• What is Natural Language Processing?
• LingoAce, Singapore-based online language learning platform, raises US$6 million in Series A+ led by Sequoia Capital India

Header image courtesy of Freepik

The post How Can VR Technology Facilitate Second Language Acquisition? appeared first on Jumpstart Magazine.

From identifying species to predicting poaching activities, here’s how AI is helping save wildlife.

What comes to your mind first when you think of wildlife conservation tools? A pair of binoculars for observing distanced large animals? Perhaps a magnifying glass for studying small creatures living on leaves? These are what zoologists and ecologists use when they carry out their studies; but as technology advances, researchers have begun deploying modern technology, namely artificial intelligence (AI), to optimize wildlife conservation.

AI has been regarded as one of the top three skyrocketing technologies used in wildlife conservation; there are also different conservation projects worldwide leveraging the power of AI to preserve wildlife species. Here are some practical and successful AI applications in wildlife conservation:

Animal identification

Occasionally, we’ll learn about the extinction of a particular animal from news reports. Seeing an unfamiliar species go extinct may not be disturbing to some, but what if we told you that, right now, there are one million species facing extinction? According to a study carried out by the United Nations Educational, Scientific and Cultural Organization (UNESCO) back in 2019, one in four (or one million) species in both animal and plant groups was vulnerable to extinction due to human activities.

Losing a species as small as a type of fungus is enough to disrupt biodiversity—the variety of life on Earth that sustains life. People who live in cities may not be concerned by the crisis; but the truth is, all species are interconnected and interdependent. For instance, without bees helping in pollination, we won’t have fruits and vegetables. Therefore, maintaining biodiversity is everyone’s duty.

Sadly, we didn’t know where to start for a long time. The world’s largest natural conservation organization, the International Union for Conservation of Nature (IUCN), keeps a red list containing endangered animals’ names. Within 149,676 endangered species, the population trends of nearly half of them (72,909) are unknown. The lack of data has left us no clue on how to act or allocate resources to help preserve species at the brink of extinction. Wildbook, an open-source platform containing the living status of millions of species, is here to be the leading light.

Initiated by the Translational Data Analytics Institute (TDAI) at Ohio State University, Wildbook collects a public stream of photos and videos of animals and uses AI to analyze them. The team makes use of machine learning and computer vision to identify different animals’ biological traits, like zebra stripes and whale flukes. Since these traits are as unique as our fingerprints, identifying them can help computers calculate different species’ populations without being bewildered by their similarity. When more data sets are being processed, ultimately, we can have a better idea of the population of different species and act accordingly to preserve them.

Poacher detection

Poaching has always been one of the largest stumbling blocks to animal conservation. Species that have economic value are vulnerable to this illicit activity. Their invaluable tusks and horns can fetch poachers a lot of money on the international market. Some of these animals include elephants and rhinoceros, which may go extinct in the next ten years due to extensive poaching. Thanks to AI, we can now protect these species from malicious hands round the clock with a human-free monitoring system.

Now, AI-backed drones and security cameras equipped with night vision are detecting poachers and any suspicious activities on the ground and alerting the rangers straight off. In addition to monitoring poaching in real-time, some also make good use of the technology to predict poaching activities effectively. Protection Assistant for Wildlife Security (PAWS), a collaborative effort that began at the University of Southern California, draws on AI to estimate the time when poachers lay snares or sneak into parks. Using its estimates, PAWS put together a ranger patrol schedule that suggests areas where poaching activities are likely to take place.

A similar project comes from Resolve, a non-profit environmental group, which partnered with mobile satellite communications provider Inmarsat to set up an anti-poaching AI system to play its part in animal conservation. Named TrailGuard AI, the system can detect humans with 97% accuracy. Given that poaching tends to happen in remote areas, the system can function in remote regions as well, thanks to the strong satellite connections supported by Immarsat. When poachers are spotted by the system, rangers will receive geographical data that help them track the activity and intervene before any animal’s life is put at risk.

Acoustic surveillance

Acoustic surveillance, or acoustic monitoring, is a common method adopted by ecologists and wildlife conservation groups to monitor and study particular species. By installing acoustic sensors, researchers can collect data that helps them identify the habitats different species live in and the size of their populations and their habits. What ecologists and researchers may not know is that the rich data they collect can support wildlife conservation. This again has something to do with poaching, the ultimate culprit threatening wildlife.

Apart from appreciating the sound of nature, a recording containing a bird’s tweets doesn’t mean much to most of us, but this is not how AI sees it. Unlike ordinary acoustic sensors, AI-enabled ones are trained with acoustic data gathered from forests, making them capable of identifying sounds that are not normally heard in a designated area, like gunshots and motor vehicles. When these sensors pick up such noises among animal sounds, they will report to related units for them to take instant actions to fend off poachers.

Apart from poachers, loggers are also to blame for wildlife destruction. Deforestation not only destroys the natural habitats of arboreal animals (those spending most of their lives on trees), such as sloths and parrots, but also leads to a short supply of food, like leaves and nuts, for other animals. The AI-enabled acoustic sensors mentioned above can also pick up chainsaw noises and report them to authorities. By doing so, we can protect both shelters and food sources for the wildlife.

The future of wildlife is still uncertain. Wildlife conservation is of utmost importance these days as experts point out that we’re now experiencing the sixth mass extinction. Over the past four decades, we have lost one million species per year. Is AI the ultimate weapon we have against mass extinction? Only time will reveal. But for the time being, AI has opened a new door for scientists to examine wildlife from a brand-new perspective.

Also read:

• How Are AI and Machine Learning Used in the Animation Industry?
• Boutiques in Your Phones: How Has Technology Changed the Fashion Industry?
• How Should Luxury Brands Fight against Counterfeiting?

Header image courtesy of Unsplash

The post Technology in Ecosystem: How AI Optimizes Wildlife Conservation? appeared first on Jumpstart Magazine.

Next time you have a perfect cup of coffee, you may be thanking AI.

When it comes to coffee, what pops up in your mind may be a coffee machine that crackles or a coffee cup with your name on it. Regardless, to many of us, artificial intelligence (AI) is not even remotely related to coffee. As irrelevant as they may seem to each other, AI has infiltrated the global coffee supply chain and impressed us with its coffee selection, roasting and pouring abilities. But what are some practical AI usages in the industry? Read on to find out.

Coffee bean assessment

In case you need a refresher, the coffee beans we usually come across can be classified into four varieties: Arabic, Robusta, Liberica and Excelsa. These varieties of coffee beans vary distinctly in aroma, flavor and price. Unless you’re a coffee enthusiast or a “Q Grader” (a professional specializing in coffee tasting and grading), it’s unlikely for you to distinguish between each type of coffee. However, with the help of AI, you can also examine coffee beans like a pro.

Demetria, an Israel-based AgriTech start-up, is launching a handheld device equipped with a near-infrared sensor that works with its AI platform “e-Palate” to determine coffee bean quality. To start with, users need to input quality parameters, such as fragrance, aftertaste, acidity and sweetness, into the device. Then, users can lay the device above the coffee beans, and the near-infrared sensor will analyze biochemical markers (a set of distinctive genetic traits that tell the crop properties) on the beans. The data obtained will then be sent to “e-Palate”, where AI matches the bean profiles with their unique flavors defined by the nonprofit, membership-based coffee trading organization Specialty Coffee Association (SCA) taster’s flavor wheel.

With Demetria’s AI-backed device, coffee assessment becomes more accessible to general coffee traders, roasters and farmers, empowering more business practitioners to examine and monitor coffee quality during trading.

Roasting coffee to perfection

Everyone can be a roaster as long as you have the right equipment with you. Meanwhile, roasting coffee well is way more demanding. Although roast profiles are personal (some may prefer a sourer or more acidic taste, while some prefer or more bitter flavor), coffee beans should neither be under-roasted (tastes grassy and sour) nor over-roasted (tastes burnt). Otherwise, they will both overpower the natural flavors of the beans.

So, how can we ensure coffee quality during roasting? The answer, again, is AI. Given that steam builds up in coffee beans during roasting, they will undergo two “cracks” (audible popping sounds) throughout the process, indicating the coffee roasting stage. For those who prefer light to medium roasts, they should stop roasting when the first crack is heard, whereas those who opt for a dark roast should wait until the second crack. As such, it’s crucial to control the roasting duration precisely to achieve our desired coffee flavors.

Norwegian sample coffee roaster manufacturer Roest teamed up with audio and machine learning (ML) firm Soundsensing to devise an AI-powered coffee roaster. Unlike other general coffee makers, Roest’s machine comes with an in-built microcontroller containing a microphone associated with a sound-sensing firmware (a software installed on a hardware) that monitors “cracks” during roasting. Once the microphone detects the first crack, the audio data will be processed by the pre-trained ML algorithm installed in the microcontroller that prompts the roaster to stop the roasting procedure, preventing the coffee beans from being over-roasted.

Although Roest’s machine may not be what dark roast fans need because it only looks for the first crack, it prevents us from drinking burnt coffee, at the very least.

Enjoy your exclusive coffee

The use of AI in the two settings mentioned above may have impressed you (hopefully). But, if you’re not a DIYer, and you’d rather remain an experienced coffee consumer, you may care more about how AI can optimize your purchasing experience. Learn more from the two case studies below!

Virtual barista

If you want to enjoy a cup of “personalized” coffee that matches your taste but don’t have much to spend on an AI coffee roaster like Roest’s, Javaya’s one-stop coffee online site may be what you need.

Javaya is a Chicago-based start-up that runs an AI-driven coffee bean marketplace. Inside the space, customers will meet a virtual AI barista asking them three questions surrounding their coffee preferences, including taste (bitterness, sweetness, sourness), body (also known as mouthfeel, a structural characteristic of coffee) and texture (grainy, creamy, smooth). Upon expressing the preferences, the AI algorithm will render instant coffee recommendations for customers and help locate the best batch of coffee beans for them.

To supply customers with coffee beans of premium quality, Javaya collaborates with only the top roasters in the U.S. who roast beans with exceptional skills. Javaya even promises that their coffee beans will be roasted, packed and shipped to customer’s doorsteps when they’re at “peak flavor”, ensuring that consumers can taste the best coffee at home.

A cup of “thoughtful” coffee

Can AI customize coffee based on more than inputted preferences? Yes, and this is what Preface Coffee has been doing. A coffee concept store established collaboratively by coding school Preface Coding and brand experience agency Secret Tour HK, Preface Coffee worked with a group of young programmers and devised a “coffee” big data AI system.

The system contains big data ranging from different times of day (sunrise, morning, afternoon, evening) and outdoor temperature to global news headlines (positive or negative news). These datasets will be analyzed by the AI system, which automatically adjusts the brewing process to create coffee that suits customers’ needs. Specifically, if customers come on a sunny day in summer (and no glaring global disaster has occurred), the AI system will increase the caffeine ratio—to energize consumers in the morning; lower the coffee temperature—or maybe add some ice to make the drink refreshing; and maintain moderate sweetness—without requiring extra sugar to boost positivity.

Since time, temperature and world news are constantly changing, the coffee recipe will also change accordingly. To help customers memorize the moment they enjoy the coffee, the store will print the coffee recipe on the coffee cup sleeve, which serves as an invaluable memento.

From what has been introduced above, AI has undoubtedly upgraded the coffee production game, taking roasting consistency in as well as the ability to customize and create optimal flavor to the next level. But will human baristas likely be put off work when more coffee shops deploy AI? Not really, because when everything surrounding us is ice-cold technology, we’d rather spend five minutes chatting sincerely with a barista while waiting for that cup of imperfect yet heartwarming coffee.

Also read:

• Boutiques in Your Phones: How Has Technology Changed the Fashion Industry?
• How Should Luxury Brands Fight against Counterfeiting?
• No More Head-Scratching: How Do Pet and Baby Translators Redefine Communication?
• When Old Meets New: How Do We Decode Ancient Texts with AI and Machine Learning?
• Will AI Replace Artists?

Header image courtesy of Freepik

The post Brewing with Artificial Intelligence: How Does Technology Breathe New Life into the Coffee Industry? appeared first on Jumpstart Magazine.

It is normal to feel sad about things in life. Why not try and talk to an AI about it? 

We often have to deal with stress in our everyday life, such as academic studies, work, family or even our romantic relationships. Amid the devastating COVID pandemic, there has been a 25% increase in depression worldwide, especially among young people. In light of this, platforms aiming to improve the mental well-being of adolescents have been emerging, and one of them is Kai.ai. 

What is Kai.ai?

Soft-launched in January this year, Kai.ai is an AI-powered companion that helps adolescents, who are often persistently battling anxiety, depression and hopelessness, nurture a better sense of well-being. As of July, more than 125,000 young people actively used Kai, with half-a-million messages being sent to Kai each week. More than 20,000 people use the AI on a daily basis, and 15,000 surface and express gratitude to start their morning. 

“We’re focused on ongoing well-being because there is such an alarming shortage of therapists compared to the demand of mental health needs, and developing routine for positive reflection can help people feel more grounded,” said the CEO and Co-founder of Kai, Alex Frenkel. “We’re making sure that more people feel in control of their feelings and less people feel like they’re in isolation so that overall, crises can decrease and fulfillment can thrive.”      

Using therapeutic and coaching approaches, Kai helps users mitigate depression, anxiety, sleeping disorders and many other psychological stressors. The platform’s name, “Kai”, is simple but concise and meaningful, echoing the company’s mission of using technology to support the emotional well-being of people. Frenkel said, “‘Kai’ means the ocean in Hawaiian and recovery in Japanese. It also means being strong, happy and lovable in other cultures.” 

Why does Kai.ai focus on young people?

Kai focuses on high school and college students, of whom many are struggling with mental health issues. A study published in 2019 comparing mental health concerns across different age groups found that people below 25 years old had the highest depression rate, with the group of college-age adults (20-21) showing the sharpest increase. 

Moreover, in the past two years, as the world was dealing with the pandemic, waves and waves of lockdowns made therapy and counseling not as accessible as before. Yet, amid the Covid crisis, more than 44% of high school students in the U.S. persistently felt sad or hopeless. Kai, which provides mental healthcare services for users to use from home, has therefore stepped in on time to help those in need to walk through such difficult times. 

The founding story

Frenkel started as a clinical therapist working in clinics and at schools. During his work, he noticed the limitations in the traditional mental health resolution model, in which students would only pay for a few meetings with the therapist when they experience mental health problems. However, talking to therapists only occasionally is far from enough to tackle mental health issues.

In addition, he was aware that there were not enough therapists at most of the schools. Hence, students in need are only receiving limited care and support, reducing the consultation’s effectiveness and even causing their mental health to further deteriorate. 

Compared to seeking help from therapists, young people would feel less ashamed or scared of being judged for seeking help on an AI platform. Frenkel shared, “The younger generation trusts technology, and they can talk to the AI more comfortably.” All these have prompted Frenkel to launch Kai. to support people’s mental well-being, especially the youth, with an AI solution.

The working model behind Kai.ai

Built by a team of psychologists, engineers and content creators, Kai integrates scientifically proven techniques according to the Acceptance Commitment Therapy (ACT) model to help people identify their emotions and perform self-reflection. 

Kai echoes mental healthcare 3.0, a system that provides fully digital care or a hybrid of human and digital care. Kai asks users different questions to pause and reflect on their lives, such as, “Hey, good morning, how are you feeling this morning?” or “Tell me something that you have in your life today.” Users can reply to the prompts, and it will remember the information that they have told the AI, building a long term relationship between the user and the AI. Besides allowing for a more personalized interaction experience, these daily questions can also help users build the habit of working on their mental well-being every day. 

What makes Kai.ai different?

There are many aspects of Kai that make it different from other mental health companies and other AI bots.

The first is in the way the product is made available. Instead of downloading an app, users, or Kai-ers, can access Kai on messaging platforms popular among young people—iMessage, Discord, WhatsApp and Telegram. Kai functions just like a contact in someone’s messaging list. It meets the person where they are at. 

On the mental healthcare side, unlike other telehealth companies, Kai is not offering a formal health service. Rather, it provides users with an ongoing experience to find a better sense of base-layer well-being. According to Frenkel, typical telehealth companies usually connect users with a therapist via video calls. However, with Kai, the user starts the journey with a simple question—“How happy are you now?” Then, the platform will help the users perform an initial assessment of the happiness score to evaluate their happiness level. 

In terms of quality assurance, Kai stands out by taking the way it measures impact very seriously. The team isn’t just focused on adoption and engagement but also being able to help people feel better from using the AI. “Kai is the only company that utilizes the World Health Organization Five Well-Being Index (WHO-5),” said Frenkel. 

WHO-5 is a short self-reported measure of current mental well-being. Users will answer five statements using a scale of one to five, and the final score represents the mental status of the user at the moment, with 0 and 100 being the worst and the best imaginable well-being. Using the WHO-5 well-being questionnaire, a study found that users’ well-being improved from an average of poor well-being to an acceptable well-being score after using Kai for four months. This shows Kai’s capability to improve users’ mental health.

Frenkel also added, “While interacting with Kai, the responses given by Kai also contain psychological theories.” Indirectly, teenagers can also learn about psychology by interacting with Kai.

Future plans of Kai.ai

A successful business, Kai has a brilliant vision for its future. Frenkel declared, “Kai would continue to grow and support millions of people’s mental health worldwide.” Currently, Kai mainly focuses on its U.S. audience, and the AI only speaks English. 

To broaden the user demographic, Kai is looking to support other age groups and expand its language offerings to reach people from non-English speaking countries as well. Furthermore, considering the growing popularity of the metaverse recently, Kai is planning to launch as a metaverse character to interact with users beyond texts.

As per Frenkel, maintaining happiness takes work and effort, and living happily is a life-long task with countless ups and downs. The first step towards that goal is to slow down and reflect on our lives from time to time. By being mindful and grateful for what we have and how powerful we have become, we will harness the abilities to tackle adversities in the future. 

Also read:

• Does Success Make Us Happy?
• Jumpstart Magazine Issue 31: The Happiness Issue
• How Is the Metaverse Changing the Medical field

Header image courtesy of Kai.ai

The post Interview with Kai.ai’s Founder Alex Frenkel: How AI Can Be a Game Changer in Mental Health Support for Young People appeared first on Jumpstart Magazine.

The sign language translation technology is on its way.

According to the Ethnologue guide, there are 7,151 spoken languages in the world. While the number might be surprising, did you know that there are also more than 300 sign languages worldwide, with American sign language (ASL) as the most widely used one?

As of 2021, 430 million people globally are suffering from some form of hearing loss, and that number is likely to double by 2050. The enormous number suggests that more people may need to learn sign language to communicate with deaf people in the near future. Some may even feel discouraged from learning the skill, since mastery of sign language takes time and extensive practice. This is where sign language translation comes into play. However, despite strenuous efforts made in developing sign language translation technology, it remains challenging to perfect the task. Read on to learn more about the obstacles facing sign language translation and how they are addressed by the newly-invented tech solutions.

Why are sign languages difficult to translate?

There is no universal sign language

As mentioned earlier, there are over 300 sign languages worldwide. From a linguistic perspective, a sign language is similar to a distinct spoken language in that they’re not mutually intelligible among themselves. To put it simply, an ASL user cannot understand British sign language (BSL) and vice versa because sign languages are developed based on regions’ dialect and culture.

Since sign language translation remains relatively experimental, there hasn’t been any system or device that translates ASL to BSL or allows users to translate sign language to any foreign language. Researchers from around the world, however, are developing systems that translate their regions’ sign language.

For instance, researchers from Complex Software Lab (a software engineering lab at University College Dublin, Ireland) have put together a new artificial intelligence (AI) -based technology that can translate Irish sign language (ISL) into spoken words. The team behind the AI understands that about 70% of communication in sign language comes from facial expressions. So, they also leverage computer vision and deep learning to capture facial expressions for more accurate translation. We will delve more into the other elements of sign language in the later part of this article.

Besides ISL translation, researchers from Australian video consultation services company Coviu have also developed a web application that translates Auslan (the national sign language of Australia) alphabets using machine learning. The team started off with building their own Auslan alphabet image dataset containing photos of different people signing the alphabets. Those photos served as data “teaching” the computer to match the signs with the corresponding alphabets. When signers sign in front of the computer webcam, machine learning helps the computer translate the Auslan alphabets in real-time.

Understanding individual signs is not enough

Except when we are replying or chit-chatting with someone (e.g., “yes,” “no,” “sure,” “how’s it,” etc.), we normally make complete sentences when communicating. The same also applies to sign language, and hence a sign language translator or software should be able to translate not only single, individual signs but also complete sentences.

Back in 2016, two sophomores from the University of Washington won the Lemelson-MIT Student Prize for their invention of a pair of sign language translation gloves named SignAloud. Inside the gloves, there are sensors measuring hand position and movement. The sensor data is then sent to a nearby computer via Bluetooth that matches the hand movements with corresponding gestures. If the data matches a gesture in the system, the translated signs will be read aloud through speakers. While the idea of translating sign language via gloves is undoubtedly groundbreaking, translating only words or phrases is far from enough.

Three years later, in 2019, researchers from Michigan State University rolled out a deep learning-backed sensory device called DeepASL, which can translate complete ASL sentences without requiring users to stop after each sign. To boost system performance, the team collected approximately 7,000 samples covering 56 common ASL words and 100 sentences to train the algorithm. The DeepASL device is powered by leap motion (a small movement tracking computer interface) which maps and tracks human hand movements through its cameras.

Sign languages involve more than just gestures

Since we see signers mainly using their hands while signing, we may think sign language is only about hand gestures. The truth is, sign language comprises three elements: hand gestures, body movements and facial expressions. All these elements help signers express meaning, such as raising the eyebrows to turn a phrase into a question. The importance of non-verbal cues in sign language communication suggests that researchers must consider more than hand gestures when developing their applications; otherwise, the translation can never be accurate.

SignAll, a Hungary-based startup that strives to enable spontaneous communication between deaf people and the hearing, has put together a device that translates sign language using computer vision and natural language processing. To unpack some technical terms, computer vision refers to enabling computers to process and analyze information (e.g. videos and pictures) through using high-performance camera and image processing software. As for natural language processing, it is the use of technology, like AI, to understand text and spoken words as humans do. 

For translation, signers need to wear gloves and sign in front of three cameras that capture their hand gestures, facial expressions and body movements. This data is then sent and processed by a central computer that transcribes complete ASL sentences for the hearing. On the other hand, the computer can also transcribe speech for deaf people using natural language processing.

Speech to signs—sign language translation goes both ways

Researchers from the Complex Software Lab at University College Dublin also noticed that sign language communication should be for both deaf people and those who can hear. Therefore, their technology focuses on more than just transcribing sign language into spoken language but also the other way around. When the software is translating speech to ISL, the signers will see an avatar signing on their screen. If signers want to translate signing to speech, all they need to do is sign in front of a Microsoft Kinect (a motion-sensing device) and let the system transcribe for the hearing.

While many people welcome and advocate sign language translation technology, some remain skeptical about the invention because sign language is too sophisticated to be translated accurately. We should, however, appreciate the efforts and time researchers have invested in building the technology and their attempts to enable effective communication between deaf people and the hearing.

Also read:

• How Formula 1 Incorporates Amazon’s AI and Machine Learning to Enhance Viewing Experience
• Boutiques in Your Phones: How Has Technology Changed the Fashion Industry?
• Explaining TikTok’s Machine Learning Algorithm
• How Do Machines Learn?

Header image courtesy of Freepik

The post Sign Language Translation: Why It Is Hard to Translate Hand Gestures to Spoken Words appeared first on Jumpstart Magazine.

NFT infringement is plaguing the art community.

The dark side of the non-fungible tokens (NFTs) market started to reveal itself after NFTs became mainstream last year. Previously, we have explored how the NFT marketplace is riddled with scams, such as artist impersonations and insider trading. Not only have such hoaxes become a big issue for buyers hoping to get their hands on genuine artist works but also for the artists themselves. 

On 18 December, 2021, British comic book artist Liam Sharp revealed in a viral tweet that people keep stealing his art to make NFTs. The artwork in question, Minotaur, was put on OpenSea, the largest NFT marketplace, by a user called 7D03E7 and priced at 0.0008 ETH (US$3.14 at the time). To stem the tide of artwork stealing, Sharp had to shut down his online gallery on the social networking site DeviantArt. Sharp’s situation points to a bigger problem with NFTs. Many NFTs currently circulating online are a result of copyright infringement and art theft. Let’s look at some steps being taken to combat art theft in this space.

How technology is addressing art theft

Blockchain technology

Since August 2021, DeviantArt and OpenSea have come together to develop a system that detects potential NFT infringement. Those who want to secure their artwork on the platform can submit their work to DevianArt. The system will monitor and scan public blockchains to find any near-identical matches of original content on DeviantArt and alert the registered artists on the platform. In the first two months of the launch, 86% of the infringement cases detected were verified as matches.

Artificial intelligence (AI)

Besides blockchain technology, AI also contributes to spotting art forgeries. MarqVision, a company dedicated to fighting against counterfeiting using AI, put forward an AI model that helps clients tackle counterfeit NFTs. Unlike the platform established collaboratively by DeviantArt and OpenSea, the removal solution MarqVision proposes does more than identify fake NFTs. It also removes them from the marketplace. The co-founder of MarqVision, Mark Lee, points out that their AI model can identify fake NFTs with 97% accuracy. Given that, MarqVision appears to be a comprehensive and long-term solution to art theft.

The issues associated with detecting art theft

Despite the detection system, many artists, including Sharp, still find securing their artwork challenging. Referring to his tweet, Sharp said he reported his stolen artworks every time, but his reports were “consistently ignored”. He also finds it “sad and frustrating” because DeviantArt uses defective testing to prevent similar crimes from happening. This flawed system not only fails to do its job but also disallows Sharp to prove himself as the rightful owner of his artwork. 

What’s worse, the detection tool serves no other practical usage than notifying artists about potential art-stealing behaviors. Simply put, the system will not recall the artwork on behalf of artists to prevent them from being stolen again.

Even if DeviantArt did recall the art, it wouldn’t accomplish much. The marketplace on which the artwork was listed would simply take it down, but it wouldn’t be removed from the blockchain itself. This is because the decentralized nature of the blockchain makes it hard to track NFTs. The art is being sold worldwide and continues to get re-sold, making it hard for the artist to figure out which jurisdiction they should file their complaint under.

Moreover, both the persons who minted these NFTs and those who are purchasing them are anonymous. OpenSea doesn’t require any of its sellers to provide proof of ownership or even their real name. However, artists who are filing copyright claims are required to prove that they are the real owners of a piece. In some cases, artists don’t even know how to file a copyright claim because the NFT in question has been minted by automated bots.

Where can artists go outside the NFT marketplace?

While many artists manage to make money minting and selling NFTs, many more have to deal with the art theft nuisance head-on. Artists who have their work copied and stolen as NFTs are recommending art enthusiasts and fans to support them in real life by commissioning them, signing up to their Patreon accounts or purchasing their actual work from their shops.

Ironically, NFTs are supposed to be all about authenticity, but NFT platforms haven’t been trying their best to protect this core value. If NFTs are to inspire more than frustration and distress from the creatives, these marketplaces must take instant actions to improve the current system so that artists feel safe to create and share their work online.

Also read:

• What Is NFT Renting?
• How Brands Are Leveraging NFTs for User Engagement
• What Is NFT Bridging?
• Can NFTs Be Used to Combat Climate Change?
• NFTs, Metaverse & Web 3.0: Is It Worth Investing in These Buzzwords?

Header image courtesy of Unsplash

The post Art Theft: The Disturbing New Issue on NFT Platforms appeared first on Jumpstart Magazine.

Developed by the LG AI Research team, AI Artist Tilda is spreading awareness on fashion sustainability by upcycling digital and physical waste from the industry.

By 2030, more than 134 million tons of textiles will be discarded. Yet, the environmental issues and impacts brought about by the textile industry tend to come across as abstract and far away, especially to those living in more developed regions away from manufacturing plants. The reason is simple—out of sight, out of mind. Since textile factories are mostly located in developing countries, the consequences of textile pollution are often borne by those regions. Thanks to advancing technology, perhaps the collaboration between artificial intelligence (AI) and humans can be a solution to encourage everyone, from the production to the consumer level, to make a change. 

Introducing the first ever AI artist, Tilda, which was developed by the LG AI Research team to bring environmental impacts associated with the fashion industry to light. Given the alarming level of both digital and physical waste generated by the fashion industry, Tilda has two primary missions. First, to reduce physical waste and the misuse of natural resources by reusing secondhand fabrics and materials. Second, to repurpose and upcycle discarded digital waste from unused designs. By collaborating with designers and craftsmen, Tilda demonstrates how AI can be a problem solver and a friend to the human race. To find out how Tilda designs pieces, its contribution to the sustainability movement and what LG AI Research has in store for its future, the Jumpstart Team spoke to Tilda’s creator.

To boost upcycling—AI artist Tilda’s “Digital Upcycling Project” debut

In June this year, Tilda debuted the “Digital Upcycling Project” to battle environmental issues in virtual space and reality. Tilda presented her initial collection of upcycled garments created from patterns from digital waste (discarded, unused designs that take up storage energy and thereby leave a carbon footprint). Upcycling is the process of recycling or reusing materials to create new designs and hence increase its original value of the material. 

First, Tilda selects digital waste images that can be upcycled. Then, she generates patterns from said images to design clothes, such as a jacket with the pattern of waves. The patterns are printed onto second-hand denim or fabric and the garment with the reinterpreted design is produced. 

“We utilized fashion so people could directly experience the impact of digital upcycling,” the LG AI Research team shares. Their goal was to present Tilda’s collection to help the audience understand what upcycling is and how it can help the environment by achieving sustainable fashion. “We hope this method of upcycling digital and physical waste will continue to grow as a widespread movement in the future,” they add. They see the opportunity to use fashion to promote sustainability and promote the concept of upcycling to the public.

The project was launched at Tilda’s metaverse store on World Environment Day, June 5, to showcase her collection of upcycled garments. Proceeds from the collection were donated to support marginalized artists who work for environmental causes.

To promote human-AI collaboration

Apart from boosting upcycling through fashion, Tilda collaborates with human artists to create novel ideas and set an example of human and AI collaboration. In her collaboration with Greedilous as a part of the New York Fashion Week in February this year, Tilda worked with the label’s designer YounHee Park. LG AI Research says, “Designer YounHee Park commented that her collaboration with Tilda inspired and incited her to try new and unfamiliar styles.” 

Tilda’s ultimate goal is to spread awareness about reducing the fashion industry’s carbon footprint. The LG AI Research team expresses, “Realistically, it’s difficult for Tilda’s influence alone [or that of any one individual] to shape an entire movement. An initiative of this scale requires a united mindset and the joint efforts of many people to succeed.” Hence, to effectively spread awareness of sustainable fashion, Tilda is looking to collaborate with and inspire like-minded artists. With Tilda’s ability to turn digital waste and unwanted fabric into new clothes, her actions can educate the public about the process of upcycling and its benefits for the environment and thus persuade the public to join the movement.

LG AI Research understands that people associate AIs with job loss, but with Tilda, LG AI Research is not developing an AI that will displace humans out of the workforce. Instead, LG AI Research aims to encourage the collaboration of AIs and humans to show that AIs can offer innovative drawings, patterns and music to facilitate the creativity of humans. “Via a steady stream of activities, we hope to make people understand that AIs are here to support and assist us as we continue to create and explore new subjects,” says the team. The collaboration of Tilda and YounHee Park is indeed a remarkable example to prove how AI can help humans through partnerships. 

As an AI artist, Tilda is designed to have the ability to generate personal emotions and react to environmental issues. Furthermore, the LG AI Research team mentions, “When it comes to fashion, people don’t simply wear clothes—they simultaneously consume the thoughts and emotions of the designers behind the clothing.” Whenever people wear clothes designed by Tilda and human designers, people are also perceiving the idea that AI and humans are contemplating a better future together.

Future possibilities for Tilda 

Moreover, the LG AI Research team states, “LG AI Research Team is planning to create a Metaverse arena for Tilda and her friends to interact and explore their ideas. Besides expanding virtually, the team plans to do street art free of negative environmental impacts.” Their plans for Tilda are only getting more extensive and prominent in scale. 

Human and AI collaboration is in itself a unique concept. Tilda’s AI collaboration with humans aims to promote creativity and give ideas about fashion, which is significantly different from other AI and human collaboration, such as SIRI, in which the AI is assisting the user to perform tasks. Tilda is set to accomplish a lot more in the future. “The collaboration between humans and AI opens up the possibility for Tilda to act as an omnipresent figure in the metaverse that can engage and create with users all the time,” the LG AI Research team divulges. 

One of the biggest perks of working with an AI in the metaverse is that the AI is unbounded by time and space, allowing it to be available for duty at any time. Furthermore, in the metaverse, Tilda can guide users to come up with innovative and creative ideas in the fields of art and music. Eventually, apart from creating patterns for fashion design, Tilda will also be able to help people create other kinds of media art. Tilda’s ultimate dream is a world where everyone can become a creator—an environmentally-conscious one.

Also read:

• Can AI Replace Sports Referees?
• How Are AI and Machine Learning Used in the Animation Industry?
• How the Metaverse can Protect the Environment
• AI Tools for Businesses

Header image from DUP Press Release

The post LG’s AI Artist Tilda: How Can Tilda Collaborate with Humans to Promote Fashion Sustainability? appeared first on Jumpstart Magazine.

Will you buy an AI-generated painting and hang it on the wall?

Before the advent of artificial intelligence (AI), we might never have imagined what the technology could accomplish. But now, as most of us have got the chance to experience the power of AI—be it through talking to voice assistants, like Siri, or enjoying personalized shopping, seeing AI moving forward in the art space is expected.

The emergence of AI art makes us wonder if AI will eventually replace human artists. Should we be concerned? Or is AI art not threatening at all? Read on to find out.

Introducing robotic art

Before answering the questions above, let’s meet two famed AI artists with talents that are out of this world.

Ai-Da the robot artist

Jumpstart has previously presented the top five humanoid robots that came into our world. While most of them are social robots powered by AI, it doesn’t mean that the technology stops there. Named after pioneering mathematician and scientist Ada Lovelace, Ai-Da is the world’s first AI-powered ultra-realistic humanoid robot specializing in art.

Ai-Da started achieving fame when it debuted back in 2019. After publicly performing poetry at a recital last year, Ai-Da marked another achievement recently by painting an “algorithm Queen” to celebrate the Queen’s platinum jubilee. According to BBC News, the “algorithm Queen” was layered and scaled to produce the final multidimensional portrait.

Ai-Da has a human-like appearance and is equipped with facial recognition technology backed by AI. With the cameras in her “eyes”, Ai-Da can analyze and focus on her painting’s subject. The data obtained from the analysis will be fed to the in-built AI algorithm that prompts Ai-Da’s robotic arm to sketch. Besides painting, Ai-Da can also write and perform poetry and create sculptures.

Dall-E mini

Creating artwork with robots is remarkable, but devising a robot artist as proficient as Ai-Da can be costly. Enter Dall-E mini. It is much more cost-effective because it’s entirely free to use. Put together by AI artist Boris Dayma, Dall-E mini is an AI artwork generation model that generates illustrations based on the prompts that users type in. The tool is similar to a search engine because they both generate results based on inquiries. But, while Google gives us search results in the formats of news, studies and images associated with the keywords, Dall-E mini is all about the fun because people typically use it to create memes! For example, if you type in “a pizza flying in the sky”, you’ll get something like this:

Given that Dall-E mini can generate more absurd pictures, such as “astronauts cooking in the sea” and “a penguin taking a selfie with a polar bear” (as seen in the images below), people like to use it to create memes to bring their imagination to life. Even though the illustrations Dall-E mini generates are mostly blurry and deformed, the AI program does empower us to let our creativity go wild.

AI art is impressive, but it isn’t “art”

To understand whether AI-generated artworks can be considered art, we must first understand what motivates us to create art. Generally speaking, we humans make art for four reasons— to evoke an emotional response, recall past events and emotions, communicate and educate. Humans are born to have feelings because we need them to respond to our ever-changing surroundings. Say, when a tiger is pacing beside us, we’ll be alert and our fear will drive us to run away or fight against the beast (not recommended). Fear is just one of our many feelings, but it helped our ancestors survive in the wild.

Our feelings are mostly inexpressible owing to their complexities. Other than making facial expressions, creating art is another ideal way for us to express the inexpressible because art is the projection of our emotions. With this conclusion, it’s safe to say that the greatest artists of all time are not only gifted for having excellent art skills but also having moving life stories to share. 

After meeting the two AI artists introduced above, there’s no doubt that AI can also make artwork, illustration and paintings. That said, machines can at most imitate human artists’ painting styles but not emotions because computers don’t have life experiences. Based on that, as long as we consider art as an expression of our emotions and feelings, it’s hard to say what AI can create is also art.

AI art is not copyright protected

This may be unfair to AI artists like Ai-Da but no, AI-generated art is ineligible for copyright protection. This February, a three-person board from the U.S. Copyright Office reviewed a 2019 ruling against Steven Thaler, a scientist and AI system creator who tried to seek copyright protection for an AI-created picture. The office rejected the application again after three years since copyright protection requires “human authorship” as per the committee. That is to say, human involvement during the creation process is a prerequisite in obtaining copyrights.

Paradoxically, AI is supposed to function with minimal human intervention. Insisting on human participation in AI art creation implies that there’s currently no chance for this type of art to be copyright protected. When more and more artists start using AI to create artworks, then perhaps AI-made art might be whitelisted. It might even get a separate mechanism to process copyright-related requests on a case-by-case basis.

AI art can be deceiving

In case you’re wondering how perfect AI art is, it’s good enough to fool our eyes and make us feel at sea while identifying them. After seeing the world’s first AI-generated portrait auctioned for US$432,500 at Christie’s, a researcher Harsha Gangadharbatla carried out an online survey to figure out how people perceive artwork. Like most other researchers, Gangadharbatla aims to find out if people can distinguish between human- and computer-made artworks.

Gangadharbatla recruited 211 subjects on Amazon and provided them with five landscapes created by human artists and AI algorithms. The subjects were asked to determine if those illustrations were human or machine-generated. At the end of the day, a majority of the respondents could only identify one out of five landscapes’ creators correctly. 

What does that imply? AI can produce quality artworks indistinguishable from those created by humans. With more people accepting machine-made artworks, the sales of conventional paintings may eventually drop. This may be why Gangadharbatla finds the result unsettling. “Already, the role of humans is shrinking every day and the role of big data is growing. So creativity is the last bastion, the line that humans hold in advertising,” he points out. When AI is equipped with creativity, machines may completely dominate the creative industry, leaving no space for human artists to survive. Thankfully, this hasn’t become the reality.

Returning to the question raised at the beginning: should we be concerned by AI art in replacing human artists? Not really. The claim that the gap between human and AI artists is narrowed to creativity may sound intimidating, but if we take a closer look, it’s in fact reassuring because machines haven’t been given creativity, and it remains our strength that makes us irreplaceable.

Also read:

• Will AI Replace Artists?
• How Are AI and Machine Learning Used in the Animation Industry?
• Boutiques in Your Phones: How Has Technology Changed the Fashion Industry?
• AI Tools for Businesses
• How AI Can Help Solve the Growing Mental Health Crisis

Header image courtesy of Unsplash

The post Drawing without a Paintbrush: An Evolution of Artificial Intelligence appeared first on Jumpstart Magazine.