• AI KATANA
  • Posts
  • How Singapore jumped ahead on AI

How Singapore jumped ahead on AI

Also: Microsoft unveils first professional certificate for generative AI skills

Author

AI KATANA
June 30, 2023

Hello!

In our latest AI KATANA edition, Singapore has made significant advancements in AI, ranking third globally despite its small size. A study revealed that AI-generated content, such as tweets from tools like ChatGPT, are perceived as more believable than human-written text, escalating concerns about the spread of misinformation. OpenAI, the creator of ChatGPT, is facing a lawsuit alleging the illicit scraping of private data from the internet. Microsoft has introduced a professional certificate for generative AI skills, aiming to meet the growing demand for AI knowledge and skills in the workplace. Furthermore, Microsoft has also released AI-powered shopping tools for Bing and Edge, promising an improved shopping experience. In venture capital news, Microsoft-backed AI startup Inflection AI raised $1.3 billion in funding, while AI company Runway has been valued at $1.5 billion after its recent fundraising effort.

Already sliced for you:

  • 🇸🇬How Singapore jumped ahead on AI

  • 🤖AI-written tweets using tools like ChatGPT easier to believe than human-written text

  • 😮 ChatGPT creator OpenAI sued for stealing private data

  • 👩🏻‍🏫 Microsoft unveils first professional certificate for generative AI skills

Singapore has made impressive strides in AI, ranking third on the Global Artificial Intelligence Index. Despite its small size and economy, Singapore's approach to AI has been strategic and cautious, focusing on neutrality in the face of US-China competition. Singapore's R&D spending on AI relative to GDP is 18 times that of the US, hosting 270 active AI startups and ranking first globally in per capita contribution to AI research papers. Its approach has been to maintain a balance between the two AI superpowers, the US and China, who contribute significantly to the foreign investment in Singapore's AI companies. Singapore's relaxed attitude towards AI regulation and its political climate, which fosters limited pushback against government initiatives, have also been key. While other small nations have similarly improved their AI standings, the UK, though still a leader in AI research and commercial ecosystem, must accelerate its pace to stay competitive.

A recent study conducted by researchers at the University of Zurich indicates that content generated by AI tools like ChatGPT, Bing AI chatbot, and Google Bard may be more believable than human-written text, potentially leading to an increased spread of misinformation. In their experiment, participants could only slightly better than randomly guess whether a piece of text was written by a human or AI, with an accuracy rate of 52%. Additionally, when comparing the credibility of synthetic tweets created by GPT-3 and human-written tweets on topics often affected by misinformation, participants recognized the synthetic tweets as accurate more often. They also took less time to evaluate these AI-generated tweets. However, humans still outperformed GPT-3 in evaluating the accuracy of information. The study also showed that GPT-3 generally produced accurate information, but occasionally refused to generate disinformation or made mistakes when asked to provide accurate information. This research highlights the need for vigilance and effective tools to identify and combat AI-generated misinformation.

OpenAI, creator of the AI model ChatGPT, faces a lawsuit filed by a group of anonymous individuals, alleging the company of unlawfully scraping 300 billion words from the internet for training its AI models, thus violating privacy laws and risking civilizational collapse. The lawsuit, demanding class action status, seeks $3 billion in damages and accuses OpenAI and Microsoft, its planned $13 billion investor, of stealing personal information from millions of internet users, including children. Plaintiffs assert that the defendants' actions have breached established protocols for the use of personal information, violated terms of service agreements, and broken state and federal privacy laws. The suit further claims OpenAI is misappropriating data to gain an edge in the "AI arms race", while moving away from its founding principle of benefiting humanity. The plaintiffs are calling for a temporary halt to commercial access to and development of OpenAI's products, in addition to monetary damages.

Microsoft has introduced a new professional certificate program focused on Generative AI, as part of its Skills for Jobs initiative. The program, which is offered through LinkedIn Learning, provides free courses on generative AI and offers a certification to those who pass the required exam. The courses cover the basics of generative AI, its ethical implications, thoughtful online search strategies, leveraging Microsoft Bing Chat, and an introduction to AI tools. Initially offered in English, the certificate will be available in multiple languages in the future. In addition to the certificate program, Microsoft is developing a toolkit for educators and trainers on the practical uses of AI. The company is also launching initiatives to foster AI learning, such as the Learn AI Skills challenge and the Generative AI Skills Grant challenge, the latter being a grant program aimed at organizations implementing AI for historically marginalized populations. This initiative is in response to the rising interest in AI and the need for workers to understand how to use AI, following a survey that showed that many workers are willing to offload work to AI to ease their workloads.

⚒️ AI tools updates

Microsoft has announced a series of new AI-powered shopping tools for Bing and its Edge browser. The tools include AI-generated buying guides which aggregate product specifications and purchasing options, a feature that could impact the existing ecosystem of buying guide websites. Bing's GPT-powered AI can now automatically generate these guides when users input queries like "college supplies". Another notable feature is AI-generated review summaries, which provide users with a brief overview of online product reviews, activated by asking Bing Chat in Edge to summarize reviews. Microsoft is also launching a Price Match tool to assist users in requesting price matches from retailers even after prices drop, with partnerships formed with top U.S. retailers. The buying guides are currently available in the U.S., with a worldwide rollout for Edge starting today.

💵 Venture Capital updates

Inflection AI, a startup backed by Silicon Valley heavyweights including Microsoft and Nvidia, has raised $1.3 billion in a funding round, bringing its valuation to $4 billion. The startup, co-founded by Google DeepMind co-founder Mustafa Suleyman and LinkedIn co-founder Reid Hoffman, has created a chatbot called Pi which uses generative AI technology similar to ChatGPT to interact with users. Inflection AI aims to build a personal AI to help users plan, schedule, and gather information. The startup, based in Palo Alto, California, and with about 35 employees, previously raised $225 million in early 2022. The recent funding will be mainly used to develop a more powerful foundational model, with plans to build a cluster of around 22,000 H100s.

AI company Runway has been valued at $1.5 billion after its latest funding round, in which it raised $141 million from investors, including Google and Nvidia, as well as Salesforce Ventures and existing investors. Runway's software, which is based on generative AI, is making significant transformations in the content creation industry by allowing users to alter existing videos or create new ones using text prompts or images.

🫡 Meme of the day

⭐ Generative AI image of the day

Before you go, check out how Meta’s details on how their AI systems rank content across their surfaces.

Have a great weekend 😊