OpenAI Stops Misuse of its AI for Geopolitical Deception
Plus: OpenAI introduces ChatGPT Edu for universities, Tech giants form an industry group to develop next-gen AI chip.
Hello Engineering Leaders and AI Enthusiasts!
Welcome to the 287th edition of The AI Edge newsletter. This edition features OpenAI’s reports of misuse of its AI for 'deceptive activity'.
And a huge shoutout to our amazing readers. We appreciate you😊
In today’s edition:
⚠️ OpenAI reports misuse of its AI for 'deceptive activity'
🏫 OpenAI introduces ChatGPT Edu for universities
💼 Tech giants form industry group to develop next-gen AI chip
🧠 Knowledge Nugget: Why Google's AI wants you to put glue on pizza by
Let’s go!
OpenAI reports misuse of its AI for 'deceptive activity'
OpenAI identified and disrupted five covert influence operations that used its generative AI models for "deceptive activity" across the internet.
OpenAI said the threat actors used its AI models to generate short comments, longer articles in various languages, made-up names, and bios for social media accounts over the last three months. These campaigns included threat actors from Russia, China, Iran, and Israel. It focused on issues including Russia's invasion of Ukraine, the conflict in Gaza, the Indian elections, and politics in Europe and the United States, among others.
OpenAI has also published a trend analysis report that describes the behavior of these malicious actors in detail.
Why does it matter?
This provides concrete evidence of generative AI being used to spread misinformation and manipulate global affairs. While there were no novel attacks this time, defending against them in the future will require continued investment in monitoring, safeguards, and public awareness from AI developers and platforms.
OpenAI introduces ChatGPT Edu for universities
OpenAI announced ChatGPT Edu, a ChatGPT version custom built for universities to responsibly deploy AI to students, faculty, researchers, and campus operations. It is powered by GPT-4o and can reason across text and vision. It also includes:
Enterprise-level security and controls affordable for educational institutions
Improved language capabilities across quality and speed supporting over 50 languages supported
Advanced capabilities like data analytics, web browsing, and document summarization
Ability to build GPTs, custom versions of ChatGPT, and share them within university workspaces
Significantly higher message limits
Conversations and data are not used to train OpenAI models
Why does it matter?
While there has been skepticism about using AI in teaching, universities like Oxford, Wharton, and more were seeing success using ChatGPT Enterprise. By tailoring its tech for them, OpenAI is setting a new standard in higher education.
Tech giants form industry group to develop next-gen AI chip
Intel, Google, Microsoft, Meta, and other tech heavyweights are establishing a new industry group called the Ultra Accelerator Link (UALink) Promoter Group to guide the development of the components that link together AI accelerator chips in data centers. The group also counts AMD, Hewlett Packard Enterprise, Broadcom, and Cisco among its members.
It is proposing a new industry standard to connect the AI accelerator chips found within a growing number of servers. It’s version one, UALink 1.0, will connect up to 1,024 AI accelerators — GPUs only — across a single computing “pod.” It is set to arrive in Q4 2024.
Why does it matter?
Nvidia, currently the biggest player in the AI chip market with an 80% share, is not in the group. Tech giants have been keen to reduce their dependence on Nvidia by working on their own custom chips. This seems to be the latest effort to break Nvidia’s dominance.
Enjoying the daily updates?
Refer your pals to subscribe to our daily newsletter and get exclusive access to 400+ game-changing AI tools.
When you use the referral link above or the “Share” button on any post, you'll get the credit for any new subscribers. All you need to do is send the link via text or email or share it on social media with friends.
Knowledge Nugget: Why Google's AI wants you to put glue on pizza
Last week, Google started testing a new feature they call "AI Overviews." Apparently, it is working perfectly!😆
Google promised to take swift action to correct these issues, pointing out that many other AI overview queries are helpful. But what is the goal here, and is it even feasible?
In this post,
discusses LLM hallucinations, the current method of tackling them, Retrieval-Augmented Generation (RAG), and what this means for web search.Why does it matter?
It highlights the fundamental limitations of LLMs as they are today. Though they are helpful for many applications, the author doesn't see them replacing web search in the future. (We’ll wait and watch.)
What Else Is Happening❗
🔗Anthropic’s Claude can now autonomously interact with external data and tools
Anthropic has announced the general availability of Tool Use for its AI, Claude. It allows Claude to autonomously interact with external data sources, APIs, and tools, making it a business-ready trusted AI solution that can be confidently adopted by enterprises across various industries. (Link)
✨Perplexity goes beyond AI search, launches publishing platform ‘Pages’
It is a tool to transform disorganized AI knowledge and research into easily digestible articles and reports for sharing. Users just have to describe the topic, select the target audience, and it will produce an in-depth article with a clear title, subheadings, associated media, and relevant citations. (Link)
🔍Google poaches AWS and Microsoft executives to strengthen AI offerings for cloud customers
Google hired two executives from these rivals as it builds out its AI ranks with the goal of boosting its cloud business. Saurabh Tiwary, a former corporate VP at Mircosoft, will join a newly created role as general manager of cloud AI. Raj Pai, a former VP at AWS, will oversee product management of the cloud AI team. (Link)
🏆Scale AI introduces SEAL Leaderboards, a new evaluation metric for frontier AI models
Trusted third-party evaluations are a missing part of the AI ecosystem, which is why Scale AI built them. These evals will be impossible to overfit, private, domain expert evaluations, unexploitable, and continuously updated with new data and models. (Link)
🎁OpenAI offers nonprofit organizations discounts on corporate ChatGPT subscriptions
It seeks to grow sales of its AI product to enterprises. Under the program, large nonprofits can get 50% off the enterprise-grade version of ChatGPT. Smaller nonprofits using ChatGPT Team will pay $20 per month per user instead of $25 or $30. (Link)
New to the newsletter?
The AI Edge keeps engineering leaders & AI enthusiasts like you on the cutting edge of AI. From machine learning to ChatGPT to generative AI and large language models, we break down the latest AI developments and how you can apply them in your work.
Thanks for reading, and see you tomorrow. 😊