Quantitatively, yes, but some of the things allowed in pharmaceutical research are more horrible than anything in farming. I'm skeptical that we're any closer to replacing animal cruelty in pharmaceutical research than we are to doing so in meat production -- articles like this understate the difficulties -- but I hope I'm wrong.
On the contrary I would say. Animal welfare laws explicitly exclude farm animals, while animal scientific research at least requires an ethics panel review, etc. The types of farms in the US that house many many animals, and hence end up in the majority of the meat supply, often have a truly dire existence before death.
Yeah, it's nuts, these are significantly more expensive than a low-end smartphone. I hope by the time my kid's a preteen there will be enough of a market for dumb phones that there will be cheaper options. I would think almost any parent would want one for their kid at some point.
First SpaceX, now Anthropic. I wonder what happened the last time the big rating agencies gave massive amounts of sketchy debt investment-grade ratings to avoid alienating their customers?
Generally didn't happen, despite the common perception. The bailouts saved the banks, they generally did not care to save the owners or executives of the banks.
Bonuses (owed under preexisting agreements) were an unfortunate necessity. The USG effectively nationalized AIG because it was determined to be too systemically important to be allowed to chaotically fail. So, the new owners needed to retain the employees necessary to continue its operations. To do so, they had to honor employee compensation agreements.
This turned out to be a wise move as the American taxpayer ended up turning a profit on the AIG bailout.
And it wasn’t as though the bonuses were uncontroversial. They were possible bc of loopholes in the bailout legislation. But the government was furious and, alongside public outrage, pressured the executives to give the back bonuses. The house of representatives even passed a bill to tax the bonuses at 90%. Many did voluntarily give them back.
> So, the new owners needed to retain the employees necessary to continue its operations. To do so, they had to honor employee compensation agreements.
At that point, compensation agreements with a previously solvent employer are... negotiable.
There's no "had to" in that scenario.
The government wanted to retain the people who knew how to run the thing. Those people wanted the most money they could get.
The deal could have been whatever the two of them struck -- and the public anger was/is that historically the government hasn't bargained very hard against the leadership of companies in that scenario.
With respect, I think this is a naive opinion. The banks that were bailed out were systemically important. This would not have been normal bankruptcies, with a very real risk of plunging the U.S. into a depression as bad as or worse than the Great Depression. So, the objective was to prevent that. And in order to keep these institutions in good working order by retaining the key employees who operated the institutions, bonuses were approved. It was distasteful and there was significant public and legislator outrage.
And, once again, the taxpayer turned a tidy profit on the bailouts.
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
The fact that the only rating published is ‘junk’, and that it was from a small agency tells us quite a lot. Given the fact that Anthropic (or its representatives) likely ‘shopped’ for a rating means that this was the best anyone was willing to give them, and that the larger agencies probably refused to rate them without more information. I would have been surprised if the bankers hadn’t lobbied for the highest rating possible, as higher ratings increase their profits from the IPO.
The whole point is to slowly build credibility, starting with easily $wayed small ratings entities, then launder that credibility into a slightly larger one, rise and repeat until financial disaster.
The point of the headline isn’t to communicate that their debt is currently rated junk, it’s to communicate that it’s impossible for their rating to improve any further without becoming investment grade.
"The rating action, which has not previously been reported, shows Anthropic is close to achieving the investment-grade status that would allow the start-up to lower its borrowing costs and tap a wider pool of institutional investors including life insurance companies and pension funds.
Bankers for Anthropic and rival OpenAI have lobbied the largest rating agencies to grant the AI labs investment-grade ratings once they go public."
I'm glad I finished my Ph.D. before any of this stuff hit the market. It sounds like it would have been a substantially different experience, and I suspect I might not have finished it with as solid a grasp on my thesis subject as I have.
Her comment about the Mathematica notebook really hits home. As a scientist, not a software engineer, who uses code as a tool to speed up more critical parts of my work, I was thrilled to have the chance to learn programming skills on the job. I could justify it because the time saved by the fruits of it outran the time I spent learning it. I can't do that anymore. I'm not building software, just scripting, so even the cheapest models today can give me suitable code quicker than I could write it myself. So I can no longer justify learning new programming skills, at least on work time. I'm glad I had a few years to pick up the basics before these tools became available.
I feel incredibly lucky to have 25 years of computer and programming experience before the LLMs arrived. All the things I have learned through hard experience and curiosity!
But at the same time, I am happy with my current use of LLMs. I get to do things I only dreamed about before. I would have so many ideas I could never realise. Now I can fire up Claude Code and get cracking. And because of my experience, I can keep up with it, correcting it, suggesting better ways of doing things…
And yes, so glad I have my PhD already. But I am looking forward to all the articles I will be writing based on my collaborations with both humans and LLMs.
on the contrary, I'm unhappy that I had to study in university without ChatGPT. I was unable to learn many things because I didn't have access to someone who could teach me and correct me quickly and I lost interest.
as a software developer I felt I learned a lot more coding skills and software architecture at my first job (albeit with a great manager who kept challenging me with interesting work) than any of my teachers ever tried to teach me in uni D:
"Anthropic still hasn’t made its prospectus public, so New Constructs hasn’t seen the actual filing. However, the firm cited figures that have been reported, including from the New York Times, which reported in September that the company was on pace to generate $100 billion in annualized revenue by the end of 2026."
Anthropic's annualized revenue run rate reached $65 billion at the end of July 2026, up from just over $5 billion in August 2025 (Anthropic, CNBC)
Anthropic made nearly $4.6 billion in revenue in 2025, 12x more than in 2024 (IPO prospectus, via Reuters, September 2026)
Anthropic's preliminary Q2 2026 revenue was more than $11.5 billion, ahead of OpenAI's $6.7 billion (Bloomberg and The Wall Street Journal, August 2026)
So, the Q2 revenue was $11.5 billion. The claimed $65 billions are "annualized revenue run rate". Please show me where that term is defined reliably. It's a BS marketing term to obfuscate real revenue after the tokenmaxxing hype has ended.
"Anthropic defines “run-rate revenue” in two parts. Use the last 28 days of sales from customers charged on a consumption basis and multiply it by 13. Then, multiply the monthly subscription take by 12, and add the two together."
Did you read it to the end? It is insane speculative fiction. They are not even qualitatively right about one of the main narrative planks, which is that by now (mid-late 2026) there would be one clearly dominant company each in the U.S. and China, and that the latter would be nationalizing their compute.
I think it's fair to say that an open model has not yet surpassed Fable 5. But I don't think it's fair to count the period prior to Fable's public release. In which case it's only been 4 months since the release of a model that's clearly ahead of today's top open models.
"The site does not yet reflect tiered pricing, so provisional cost figures for Haiku 5.5 do not include the step up cost."
Frankly, it's irresponsible for them to put the points on the chart if they don't have the correct values for one of the axes yet. At least they added some asterisks since yesterday, but you still have to go to the blog post (https://artificialanalysis.ai/articles/claude-haiku-5-5) to see what the asterisks mean.
reply