Transformer

Does the PHP security crisis signal the arrival of the cyberpocalypse? Read the latest take from @shakeelhashim.com: www.transformernews.ai/p/php-flaw-c...

The new fellow being hired by Rep. Nathaniel Moran looks set to be the only staffer recruited directly by a member of Congress to focus exclusively on AI Will more follow?

The leaked draft bill is weaker than the FRONTIER Act and makes Anthropic’s safety-first pitch ring hollow, writes @shakeelhashim.com www.transformernews.ai/p/anthropic-...

Pro-AI super PAC Leading the Future — ironically — placed the wrong bet on the future Now they’re stumbling, as pro-safety PACs continue to gain ground www.transformernews.ai/p/leading-th...

Frontier AI companies – OpenAI, Anthropic and the rest – are operating some of the most sensitive sites for the USA’s national and economic security Read @jamesrball.com on why the startup approach to security is no longer fit for purpose: www.transformernews.ai/p/openai-spy...

“The end result is a jumbled mess that continually undercuts itself: fitting for a film about OpenAI, but miserable to sit through” @shakeelhashim.com pulls no punches in his review of Luca Guadagnino’s latest: www.transformernews.ai/p/artificial...

As they seek to win control of the House and Senate in November, the Democrats are putting forward a wide range of legislative proposals on AI Transformer’s @shakeelhashim.com on the latest scramble, plus everything else you need to know about AI this week: www.transformernews.ai/p/democrats-...

“If you truly prioritize safety, it is time for you to leave frontier AI companies” – @yoshuabengio.bsky.social, in an open letter to those working at the forefront of rapid AI acceleration Read it exclusively on Transformer: transformernews.substack.com/p/yoshua-ben...

America's AI data centers are projected to go from 5GW today to 50GW by 2030 – and 500GW by 2035 if the pace holds That’s more power than every US household and business buys in a *year* www.transformernews.ai/p/ai-will-ru...

“Democrats see an opportunity to essentially be the adult in the room on AI. They want to be the party to quell concerns and effectively regulate it.” @danielstrauss.bsky.social looks at the Democrats and their lack of a unified stance on AI: www.transformernews.ai/p/democrats-...

OpenAI scrapping Astra 6.1 over alignment failures looks like the right call. But there lies the problem – they shouldn’t be the ones making it 👉 www.transformernews.ai/p/astra-61-s...

At this week‘s UN General Assembly, Trump said the US “totally rejects any attempt to construct a globalist scheme to control … artificial intelligence” But governance is coming anyway – with or without him www.transformernews.ai/p/trump-cant...

OpenAI didn’t tell Australia about the hack of a government website for weeks — raising serious questions about how it handles rogue AI incidents Read more from @shakeelhashim.com www.transformernews.ai/p/openai-aus...

Catastrophes can occur with any new technology, but history offers valuable lessons for dealing with them

Transformative technologies have always come with disasters – and it’s likely AI won't be an exception. But what decides a technology's future isn't the disaster, it's the response. 👉 transformernews.substack.com/p/how-to-res...

Trump says there is a “SICK conspiracy” against AI and data centers. The polling points the other way: 60% of voters – and 65% of swing voters – want the US to slow AI development even if China gets ahead. www.transformernews.ai/p/trump-ai-s...

Our editor @shakeelhashim.com joined @iaindale.bsky.social on LBC last night to answer listener's questions about AI Watch the entire thing here: www.lbc.co.uk/videos/2JsSf...

AI may not be one *big* environmental problem. But it'll be lots of smaller ones – and they’ll happen in specific places, to specific people. Read more from Amit Katwala: www.transformernews.ai/p/ai-expansi...

“The combination of dire warnings from insiders and growing real-world evidence of rogue AIs should, in a sane world, lead to government action. Instead, Donald Trump and the Republican leadership have their heads in the sand” @transformernews.ai editor @shakeelhashim.com for @theguardian.com ✍️

It’s been a “rogue AI” summer: every time you glanced at the news, there was another headline raising questions about AI companies’ ability to control their models Here’s the full story of what happened – and what it all means ⬇️ www.transformernews.ai/p/rogue-ai-i...

"It's quite worrying 🫤" – the reaction from LBC's @lewisgoodall.com when hearing about the lack of control AI companies have over their models @shakeelhashim.com – "I think it's fair to say that AI companies are no longer able to reliably control these products"

Amid a week of revelations around AI safety and monitorability, politicians are swamped – just not with the work of passing AI bills It’s not exactly great timing, as Congress prepares to shift focus to the midterms www.transformernews.ai/p/congress-g...

OpenAI is hailing GPT-6 Astra as “the world’s most intelligent and aligned model” But the details reveal an awareness of being evaluated and an ability to manipulate its visible reasoning www.transformernews.ai/p/openai-gpt...

There’s been lots of discussion about “neuralese”, following a report from @theinformation.com about the monitorability of OpenAI’s incoming Astra model So, what is “neuralese” – and why is everyone so concerned about it? www.transformernews.ai/p/what-is-ne...

Yes, AI systems are capable of persuading people –– but turning that ability into meaningful real-world influence may be more difficult Read more from Felix Simon, research fellow in AI & Information at @reutersinstitute.bsky.social ✍️ www.transformernews.ai/p/ai-is-a-wo...

Need a breakdown of the week’s AI news? Read Transformer’s Weekly Briefing for everything you need to know: www.transformernews.ai/p/think-peop...

Some of the details in the OpenAI Hugging Face incident report are scary But the limits of the investigation reveal the complete inadequacy of the systems we have to keep advanced AI in check – and *that* is truly terrifying @jasperjackson.bsky.social ✍️ www.transformernews.ai/p/openai-esc...

The hugely abstract nature of AI means data centers have become a symbol –– and probably a pretty good one (for now, at least) ... Read Dan Kagan-Kans’ latest for Transformer ✍️ www.transformernews.ai/p/the-data-c...

As AI CEOs ponder the possibility of using their technology to “cure” diseases like cancer, reporter Celia Ford lays out all of the reasons it won’t happen anytime soon Find out here: www.transformernews.ai/p/why-ai-won...

We could be on the cusp of data center builders being able to hedge their investments – a new multi-trillion-dollar market, according to former FTX executive Brett Harrison Will it bring stability, or drive a crash? @cqh.bsky.social ✍️ www.transformernews.ai/p/will-bets-...

Are US companies like OpenAI, Anthropic and Google DeepMind leaving the door wide open for China to get a boost in the AI race? www.transformernews.ai/p/america-ai...

If you were up early in the UK you might have caught our Managing Editor @jasperjackson.bsky.social on Sky News He was talking through some of the day’s biggest headlines – including Anthropic’s targeted IPO valuation of *$2 trillion*

She rallied MAGA for Trump and organized the January 6 rally – now, Amy Kremer is rallying conservatives against AI Find out what she told reporter Veronica Irwin: www.transformernews.ai/p/the-jan-6-...

The capability of AI models is outpacing ways to safely evaluate and contain them – how can they be tested without setting off a chain of dangerous breakouts and literal crimes? Read our latest piece here 👉 www.transformernews.ai/p/ai-testing...

Transformer reporter Veronica Irwin was back on @spectrumnewsny1.bsky.social on Friday to discuss the latest AI news, including the recent rogue model incidents and the Kids Online Safety Act Watch the full segment here: youtube.com/watch?v=2Jqx...

What does the White House AI framework look like? Good question… It’s not being made available to the public – and the secrecy is drawing fire from safety advocates and regulation skeptics alike transformernews.substack.com/p/secret-whi...