I tested ChatGPT against Copilot in Microsoft Word
I tested ChatGPT against Copilot in Microsoft Word

I tried ChatGPT and Copilot side by side in Microsoft Word, and I was surprised by what I found.

The fastest-adopted model on Vercel this year cannot write a word
The fastest-adopted model on Vercel this year cannot write a word

We have lightning in a bottle, and yet it is not useful. Diogo Almeida helped build ChatGPT. He worked on the instruction-following research behind it at OpenAI, and the four years since have left him disappointed by what it became, as he told TechCrun…

Anthropic and OpenAI ask Australia to ease its ban on AI training with local content
Anthropic and OpenAI ask Australia to ease its ban on AI training with local content

Anthropic and OpenAI have asked Australian lawmakers to reconsider rules that stop them from training AI models on the country’s creative work, Reuters reported on Tuesday, citing the companies’ submissions to a parliamentary inquiry into AI laws. In O…

I asked ChatGPT what it was least confident about, and its advice improved
I asked ChatGPT what it was least confident about, and its advice improved

I tried asking ChatGPT what it was least confident about and uncovered any weaknesses

OpenAI forms math advisory group as its AI resolves more than 100 open problems

The group won’t be given leeway to slow down or redirect OpenAI’s ongoing mathematical research.

OpenAI asks Washington to lead a global AI standards push
OpenAI asks Washington to lead a global AI standards push

OpenAI has called on the United States to lead an international effort on standards for advanced AI, beginning with rules on which incidents must be reported, and suggested routing the work through national AI safety institutes including the US Center …

People are using ChatGPT for much more than writing and research — these are the 7 ideas I want to steal
People are using ChatGPT for much more than writing and research — these are the 7 ideas I want to steal

Bored of ChatGPT? I went hunting through Reddit for the most useful, creative and unexpected ways people are actually using it.

A Microsoft scientist called AI training the largest theft of labor
A Microsoft scientist called AI training the largest theft of labor

A 92-page brief unsealed on Thursday opens by quoting the same Microsoft employee twice. Dr Brent Hecht, the company’s Director of Applied Science, wrote that people would come to see large models hoovering up their work as “an astonishing theft of unp…

A single git trick beat the safety lock on four AI coding agents
A single git trick beat the safety lock on four AI coding agents

The AI industry already knew attackers could poison plugin marketplaces. Its answer was to lock every plugin to one reviewed version of its code. That lock is what just broke. Researchers at Air Security published a vulnerability on Thursday that they …