No major new releases this week, but OpenAI did release GPT-5.5 Instant, which is a small bump rom GPT-5.3 Instant and a good upgrade for free users. They also released new realtime voice models in the API.
In other news this week, Anthropic was reported to have grown its revenue 80x in the last quarter (🤯). Mozilla also released a paper showing the huge number of bugs they fixed in the last month with the help of Anthropic’s Mythos model showing that there is a lot of substance behind the hype. And two Figure Helix robots made a bed 🤷♂️.
I was also disappointed on Saturday night to see a tweet from Sam Altman, that as an autistic person, I found offensive. He and OpenAI should be better than this, especially given how many neurodivergent people work in AI and technology.
Anthropic weighs deal for near $1tn valuation as revenue surges
This week it was reported that Anthropic’s revenue grew 80x in a single quarter, which is ridiculous levels of growth! To deal with all the new demand they’ve had to rent a whole data centre from xAI, bringing into question whether Elon Musk is pivoting the company to be a data centre builder, rather than a frontier AI company.
A lot have people have doubted the financial viability of the levels of investment that we’ve seen over the last 18 months into AI labs, data centres and chip manufacturing but I think there’s still a lot more demand to come and it’s unlikely we’re in full bubble territory yet. I might be wrong, but 80x revenue growth in one quarter is unprecedented and tells an interesting story.
How Anthropic’s Mythos has rewritten Firefox’s approach to cybersecurity
There was some skepticism that when Anthropic announced Mythos (their ‘too-powerful-to-release’ model) that it was just a clever marketing ploy. I didn’t buy that, and it’s good to now start getting an idea of the impact that it’s having for the limited number of organisations that have access to it.
Security researchers for Mozilla’s Firefox browser fixed 423 bugs in April, compared to the usual 20-30 bugs they fix every month. That’s astounding! Some of these bugs had lain dormant in the code for more than a decade.
And this isn’t automating or removing the need for very experienced and capable software engineers. Mozilla say that this is create more, higher value work, so they can improve the stability and quality of their web browser. It’s a great example of how AI is creating new opportunities for work and not just automating jobs away.
Helix 02 Bedroom Tidy
Good to see Figure getting their sales pitch in early that you need two of their robots and not just one 🤣.
In all seriousness, it’s a bit weird seeing two robots work together and pausing to look at each other before performing an action together. I guess it’s something we’re going to have to get used to seeing though!
Web 4.0
Image AI models now drive app growth, beating chatbot upgrades
OpenAI Enables Marketing Cookies by Default for Free ChatGPT Users
Anthropic and OpenAI are both launching joint ventures for enterprise AI services
AI Ethics News
I was disappointed to see this tweet from Sam Altman. As an autistic person, I found it pretty offensive, and I suspect many other people did too.
Describing someone as an “autistic genius with very strange taste” leans heavily on tired stereotypes about autistic people being eccentric savants or socially odd caricatures. Even if intended jokingly, it’s a strange thing for the leader of one of the world’s largest AI companies to post publicly, especially given how many neurodivergent people work across the technology industry.
Google developers significantly misstate carbon emissions of proposed UK datacentres
Anthropic Has Added Several More Religions on Its Quest to Inject Perfect Morals into Claude
Google DeepMind workers in UK vote to unionize amid deal with US military
AI firms should face ‘minimum wage for robots’ to limit job cuts, says tech boss
Canadian fiddler sues Google after AI Overview wrongly claimed he was a sex offender
“The future is already here, it’s just not evenly distributed.“
William Gibson