AI Evolution: Balancing Promise and Pitfalls

Google is making changes to settle lawsuits related to data collected during incognito mode. Meanwhile, in the world of Generative AI, great promise coexists with unexpected model hallucinations, an issue that the AI community is actively addressing.

All this and more in today’s newsletter

News Bytes

• Google has reached a proposed legal settlement to address privacy concerns related to its Incognito mode. The deal aims to resolve a class action lawsuit filed in the US in 2020, which accused Google of collecting user data even when they were browsing in “private mode.” Under the terms of the settlement, Google will delete billions of records and submit to restrictions on its tracking practices.

• OpenAI has made its ChatGPT accessible to everyone without requiring an account. Now, even users who haven’t created an account can engage with ChatGPT. However, there’s a catch: the experience won’t be identical. While you can chat freely, certain features—such as saving or sharing chats and using custom instructions—are exclusive to logged-in users.

• AI hallucinations occur when generative AI models deliver false or misleading information. These “hallucinations” can result from extrapolation, combining patterns, or creating imaginative results beyond what’s directly observed or explicitly instructed. While AI models have evolved, they still occasionally produce errors, even if less glaringly obvious.

• The Office of Management and Budget (OMB), part of the Executive Office of the President, has issued a memorandum directing federal agencies to advance artificial intelligence (AI) governance, innovation, and risk management. Each agency must designate a Chief AI Officer (CAIO) within 60 days. The CAIO will coordinate AI efforts across technical and policy areas.

• University of Texas at Austin have developed a groundbreaking wearable cap that enables users to control complex computer tasks using their brain activity. Unlike traditional brain-computer interfaces (BCIs) that require personalized calibration or invasive neurosurgery, this cap employs a novel machine learning program. Users wear the cap, which resembles a swimmer’s cap with red electrodes, and the program interprets their neural activity into actionable commands.

Productivity Tools

• ChatPDF : Interact with your PDF files like a human. Easily extract information from even large documents

• Humata AI : Chat your way through long documents. Command PDF AI to summarize for you.

• GPTExcel : Generate Spreadsheet Formulas with AI. An ideal solution for individuals and businesses looking to streamline their spreadsheet processes.

• Ask Your PDF : Say hello to documents that respond to you! With AskYourPDF, your reading isn't just simple, it's fun!

• Merlin : All-in-One AI extension to Write, Summarize, Code & Play.

Social Media Bytes

Get the facts on Large Language Model (LLM) hallucination

Feedback Bytes

Thanks for reading and If you’d like to sign up for this newsletter or share it with a friend or colleague, you can find us right here. 📬