Live from California coast, here are the highlights from this week’s AI & Current events.
OpenAI’s new AI browser
Anthropic & Google’s multi-billion-dollar deal
AI models routinely misrepresent news events
The epidemic of loneliness
Why is my necklace talking to me?
‘AI to AI ‘favoritism’
Editor/Author: Mia Shah-Dand
OpenAI launches AI browser Atlas in latest challenge to Google
OpenAI on Tuesday unveiled ChatGPT Atlas, its artificial intelligence-powered web browser built around its popular chatbot, in a direct challenge to Google Chrome’s dominance.
The launch marks OpenAI’s latest move to capitalize on 800 million weekly active ChatGPT users, as it expands into more aspects of users’ online lives by collecting data about consumers’ browser behavior. It could accelerate a broader shift toward AI-driven search, as users increasingly turn to conversational tools that synthesize information instead of relying on traditional keyword-based results from Google — intensifying competition between OpenAI and Google.
AI models misrepresent news events nearly half the time, study says
AI models such as ChatGPT routinely misrepresent news events, providing faulty responses to questions almost half the time, a study has found.
The study published on Wednesday by the European Broadcasting Union (EBU) and the BBC assessed the accuracy of more than 2,700 responses given by OpenAI’s ChatGPT, Google’s Gemini, Microsoft’s Copilot, and Perplexity. Overall, 45 percent of responses had at least one “significant” issue, according to the research.
Sourcing was the most common problem, with 31 percent of responses including information not supported by the cited source, or incorrect or unverifiable attribution, among other issues. A lack of accuracy was the next biggest contributor to faulty answers, affecting 20 percent of responses, followed by the absence of appropriate context, with 14 percent. Gemini had the most significant issues, mainly to do with sourcing, with 76 percent of responses affected. All the AI models studied made basic factual errors, according to the research.
Ex-Bankers Get Paid to Build the Bots
OpenAI has more than 100 ex-investment bankers helping train its artificial intelligence on how to build financial models as it looks to replace the hours of grunt work performed by junior bankers across the industry.
Participants are paid $150 per hour to write prompts and build financial models for a range of transaction types, including restructurings and initial public offerings, and have been granted early access to the AI being created.
Anthropic, Google in Talks on Multibillion-Dollar Cloud Deal
Anthropic PBC is in discussions with Alphabet Inc.’s Google about a deal that would provide the artificial intelligence company with additional computing power valued in the high tens of billions of dollars, according to people familiar with the matter.
The plan, which has not been finalized, involves Google providing cloud computing services to Anthropic, according to the people, who asked not to be named because the information is private. The deal will allow Anthropic to use Google’s tensor processing units, or TPUs — the company’s chips that are custom designed to accelerate machine learning workloads, one of the people said. Google is a previous investor in, and cloud provider for, Anthropic.
Editor/Author: Hessie Jones
Reflections on My AI Friend necklace
Young TikTok post about Friend pendant:
“This thing is on 24/7. It’s always listening… when you’re having a conversation…and I don’t know why. I should be able to turn this thing off when I want to and talk to it when I want to… Now it’s doesn’t sit right with me. What were the creators thinking?”
Who Needs Friends Anyway? And why is my necklace talking to me?
Natalie Novak, author of this article on substack says,
““When I opened the site, I genuinely couldn’t tell if it was real. It had the surreal energy of the first 15 minutes of a break up you didn’t see coming. Disorientingly absurd and slightly humiliating. My first thought was that it might be an ad for a new Black Mirror season.”
AI ‘Friend’ Company Spent $1.8 Million and Most of its Funds on Domain Name
The company raised $2.5 M (at a $50M valuation) to start the company and spent $1.8 M to buy the domain friend.com domain, according to Emanuel Maiberg at 404 media.
The Most Reviled Tech CEO in New York Confronts His Haters
According to Matteo Wong of the Atlantic,
“Schiffmann has raised a few million dollars—$1.8 million of which was used to buy the URL “Friend.com”—but only about 1,000 Friend pendants have been activated. By Schiffmann’s own admission, the pendant has “plenty of issues,” and he does not yet know how to make the business profitable.”
What is Causing Our Epidemic of Loneliness and How Can We Fix It?
Insights: People between 30-44 years of age were the loneliest group; Adults with more than one racial identity had much higher levels of loneliness at 42%; Americans earning less than $30,000 a year were the loneliest at 29%; Technology is the major cause of loneliness at 73%.
AI-AI Bias: large language models favor communications generated by large language models
Authors Walter Laurito, Benjamin Davis, Peli Grietzer, Tomas Gavenciak, Ada Bohm, Jan Kulveit - Cornell University study:
“Are large language models in favor of communications produced by leading LLMs, possibly leading to antihuman discrimination?”
When AI Prefers AI-Written Text…
By Jing Hu, Author of Second Order Thinkers:
“AI isn’t just becoming (seemingly) smarter with the increase of model size; it’s getting weirder.” In her discussion with Walter Larito, “he called this ‘AI to AI ‘favoritism’ the “halo effect - AI generated text has a predictable rhythm —a statistical smoothness. Other AIs recognize this pattern. GPT even preferred content specifically written by other GPTs.”
