AI hackathon underway

Reading time: 5 minutes  

 Newsletter | News | Tools 

Today we will discuss-

• 🤩India's central bank to use AI for supervision

• 🕵️‍♂️AI enthusiasts compete to find flaws in AI chatbots

• 👩‍🏫Swiss teachers call for stricter AI regulations

• 🧑🏻‍🎤9 amazing AI tools you might not have heard of

All this and more - Let's dive in!

👩‍🍳What’s cooking in the newsroom?

India's RBI selects Mckinsey and Accenture to develop AI model for regulatory supervision

👨‍🏫News - The Reserve Bank of India (RBI) has selected global consultancy firms McKinsey and Accenture to develop systems that use AI and ML to analyze large amounts of data, identify patterns, and flag potential risks.

This will help the RBI to better supervise banks and NBFCs and to take early action to prevent financial crises.

As per RBI, the contract is worth about ₹91 crore, By using these advanced technologies, the RBI can stay ahead of the curve and protect the financial system from harm.

👩‍💻What’s the backstory though? The RBI had, in September 2022, announced that it was looking for companies to help them use advanced analytics, artificial intelligence, and machine learning to supervise banks. The companies that submitted expressions of interest (EOI) revealed their fascination in the project and outlined their qualifications.

After reviewing the EOIs, the RBI selected seven companies to participate in the request for proposal process. It's a process where companies' document outlines the scope of work, the budget, and the desired timeline for the project.

In conclusion, the RBI's move is in line with the trend of financial regulators around the world using AI and ML to improve their regulatory supervision.

AI enthusiasts compete to expose flaws in 8 leading large-language models

👨‍💻News - White House officials and Silicon Valley giants are concerned about the potential for AI chatbots to be used for harm. In response, 2,200 hackers are competing in a three-day red-teaming exercise at the DefCon hacker convention in Las Vegas to find flaws in eight leading large-language models.

The competition is the first-ever independent effort to assess the security of AI chatbots. It is a sign of the growing concern about the potential for AI chatbots to be used for malicious purposes.

The findings of the competition are not expected to be released until February, but experts say that it will be difficult to find and fix flaws in these complex systems.

🧑‍🎨What do experts think about finding flaws in LLMs?

• Gary McGraw, a cybersecurity veteran and co-founder of the Berryville Institute of Machine Learning, warned that it is not easy to secure AI chatbots. He said that it is tempting to think that we can simply "sprinkle some magic security dust" on these systems after they are built, but this is not the case.

• Bruce Schneier, a Harvard public-interest technologist, agreed with McGraw and said that DefCon competitors are "more likely to walk away finding new, hard problems." He compared the security of AI chatbots to the security of computer systems 30 years ago, when it was much easier to find and exploit flaws.

Swiss teachers demand more oversight and call for caution on AI use in schools

👩‍🎓News - Swiss teachers are calling for more regulations on the use of artificial intelligence in classrooms, citing concerns about the potential for the technology to be used to collect and misuse sensitive data.

The Swiss Teachers' Association, on Monday, said that there are currently no clear guidelines for teachers on how to use AI in a safe and ethical way. This is a concern, the association said, because teachers often have to deal with sensitive data about their students, such as information about their learning disabilities.

What’s more? Beat Schwendimann, a member of the Swiss Teachers' Association management board, suggested that teachers be trained on how to use AI safely and ethically. He also stated that the Swiss government and cantons should develop ethical guidelines for the use of AI in schools.

Educa, a specialist agency commissioned by the government and cantons, warned that the world of AI is still very much unknown, especially when it comes to how some tools and apps store data.

Its director Tony Ritz suggests that teachers should not enter any personal data in these AI applications. This is because we don't know how the data is stored and used.

👩🏼‍💻What else is happening?

• UMP implements AI technology for enhanced chest X-Ray analysis efficiency

• Google-backed Anthropic raises $100M from South Korea's SK Telecom

• OpenAI may go bankrupt by 2024, a report suggests

• China start-up Fourier Intelligence eyes two-legged robots with AI brains

• AI will have a bigger impact on UK than industrial revolution: deputy PM

🕵🏼‍♀️Interesting facts about AI

• The media landscape is evolving as robots are now composing news pieces, even for major outlets like The Washington Post and BBC. Narrative Science's Quill, an AI platform, predicts that bots might craft 90% of news by 2025.

• Robo-dogs are the latest AI-powered robotic pets. While real pets demand care, the AI counterparts offer emotional connection without upkeep. Robo-dogs mimic real animals and adapt to owners' habits, curbing loneliness, and fostering relationships.

• In 2020, only 9% of businesses used AI tools like machine learning and voice recognition. However, the market was still worth $10 billion. It has continued to grow and is projected to reach $27 billion by 2026.

👩🏼‍🚒Discover mind-blowing AI tools

1. Relay - An AI-powered portfolio analyzer that helps investors extract key details from investment documents

2. YaatriAI - A travel-planning platform that uses advanced AI technology to provide personalized recommendations

3. Architect Render - An AI-powered platform that allows users to transform photos, sketches, and designs into realistic 3D renderings

4. Zuckisreal - A tool that allows people to chat with an AI parody of Zuck for an interactive live AMA

5. Bedrock - A serverless API that offers access to pre-built, foundational AI models from leading providers like Amazon

6. imagetocaption - An AI-powered tool that automatically generates SEO-optimized captions for social media images

7. Saruna - An AI-powered language learning app that improves users' speaking skills in over 20 languages

8. Blackbox - A web-based tool that enables users to copy code from videos, images, PDFs, and other formats

9. Deepswap - An AI video generator app that allows users to create face swap videos

 ✨Mythbuster

Myth - AI poses a threat to national security

Fact - The myth that AI poses a threat to national security is largely unfounded. While there are some potential risks associated with AI, it is unlikely that these risks will materialize in the near future. In fact, AI is more likely to be used to enhance national security than to threaten it.