Open AI

Monday

John Mac Ghlionn

My son was murdered after whistleblowing on OpenAI

From our US edition

When Tucker Carlson sat down with OpenAI founder Sam Altman in an interview aired last week, the conversation took a dark and frosty turn when Carlson raised the death of a former OpenAI researcher. Suchir Balaji, who exposed the company’s systematic theft of copyrighted work, was found dead in his San Francisco apartment last November. Altman called it “suicide.” Clearly unconvinced, Carlson asserted that Balaji was "definitely murdered.” Altman was offended by his insinuation and described the death as a "great tragedy," saying he was "really shaken" by it. But Balaji’s grieving mother, Poornima Rao, is very much in agreement with Carlson.

Sam Altman openai

Tuesday

Rachel Tyrell

The race to superintelligence

From our US edition

This summer, two of the leading contenders in the great AI race have suddenly, alarmingly, declared that the endgame is in sight and that they’re now spending vast amounts of time and money to try to ensure that their own AIs beat the others. What does winning mean? It means that their models (you know them perhaps as GPT, Claude and Gemini) reach first AGI (human-level intelligence), then superintelligence. No one quite knows what superintelligence will do (we’re not smart enough) but it’s clear that whoever owns the winning model will wield unimaginable power. They’ll dominate the world. A new Alexander the Great. The first to show his hand was Sam Altman, the chief executive and founder of OpenAI, a company he once shared with his former friend Elon Musk.

superintelligence

Tuesday

Cockburn

British journalist talks America’s ‘authoritarian culture’ with Jon Stewart

From our US edition

Cockburn is not a regular viewer of The Daily Show. It is no longer as epoch-defining as it was in Jon Stewart’s heyday. But he did take an interest in Stewart’s segment last night with Carole Cadwalladr. For the uninitiated, Cadwalladr is a former Guardian and Observer columnist from the UK most prominent for her reporting on Cambridge Analytica. CA is the political consulting firm known for its contentious use of Facebook data in the 2016 US election and Brexit referendum. After Brexit came what Spiked’s Brendan O’Neill dubbed “the middle-class meltdown to end all middle-class meltdowns.” “And at the heart of it all,” wrote O’Neill, “was a writer for the Observer called Carole Cadwalladr.

Tuesday

Cockburn

Meet CuomoGPT

From our US edition

Former governor uses AI to co-author housing plan If you were worried about disgraced former governor Andrew Cuomo being too “hands-on” during his run for New York mayor, fear not: the Love Gov’s campaign has an impersonal touch to it. Local news site Hell Gate exposed how the 29-page housing plan released by the Cuomo campaign bore the hallmarks of being partially put together using ChatGPT. One particularly unparsable passage: Nevertheless, several candidates for mayor this year have either called directly for a rent increase or for other measures that would tilt the scale toward lower rent increases. This is a politically convenient posture, but to be in. Victory if landlords — small landlords in particular — are simply unable to maintain their buildings.

cuomo

Wednesday

James Ball

Conspiracy theories are as old as witch hunts 

To millions of people across America, Hillary Clinton sits atop a global network of satanic child-traffickers and is battling an underground resistance led by Donald Trump to maintain her malign influence. That is the core tenet of the QAnon movement, a conspiracy theory that originated in obscure corners of the internet before being seized upon by members of the Republican elite for their political advantage. Did the Clintons get rich while undertaking a life of public service? Absolutely QAnon is at the heart of Gabriel Gatehouse’s The Coming Storm. But the book begins with a conspiracy theory from centuries earlier, about witches in medieval Europe.

Friday

Ross Anderson

Could AI ruin the election?

From our US edition

The artificial intelligence space is strange. Significantly overfunded, overhyped and overcovered — in part because AI can easily produce bad, generic copywriting, which is how many journalists presently earn their livelihoods. Though AI tools have rapidly advanced over the past year, few look to be truly society-morphing, and it’s fairly obvious when something is a product of AI, or it hasn’t mattered. Do you really care if a cliché-ridden cover letter was produced by an unimaginative human mind or a chatbot? But Thursday’s announcement of OpenAI’s new video-creating tool, Sora, is something different.

openai elections

Wednesday

The Spectator

Cuomosexual conversion therapy

From our US edition

Why apologize when you can just wait and hope people forget what you did wrong? As we enter the season of goodwill and gratitude, that’s the question posed by disgraced former New York governor Andrew Cuomo, who this year is thankful for the complicity of his allies as he attempts to stage a comeback.While New York City mayor Eric Adams chokes on a different kind of Turkey, Politico writes that Cuomo has “begun in recent days to gauge the viability of a potential mayoral bid.”Cuomo resigned as governor in ignominy back in August 2021 after an investigation by New York attorney general Leticia James claimed that he had sexually harassed as many as eleven women.

Monday

Ross Anderson

Why is Sarah Silverman suing artificial intelligence?

From our US edition

Crypto was a wonderful Wild West of anarchic financial innovation, absurd idiocy and scamming. Lots of scamming. Then regulators came along and made everything a lot more sensible and boring. Given how fast Generative AI has developed — from computer science theory to high school cheating scandals in but a few years — it was inevitable that the lawsuits would quickly follow. On Friday, the comedian Sarah Silverman joined authors Christopher Golden and Richard Kadrey in class action copyright lawsuits, claiming OpenAI and Meta had stolen material from their books to train their Large Language Models (LLMs). They allege the LLMs were trained on their books through pirated online libraries, such as Library Genesis and Z-Library. (No, I haven’t used them for years, don’t ask.