frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Anthropic's IPO prospectus shows AI vision, surging costs

https://www.reuters.com/business/finance/anthropics-ipo-prospectus-shows-sweeping-ai-vision-surging-costs-2026-09-28/
38•6thbit•54m ago

Comments

2ss•42m ago
2 things can be true:

-LLMs profoundly change society -The LLM business is mangled in terms of ROIC vs CoC

Such a business already exists - airlines.

This nuance is what many here refuse/find difficult to understand.

CoolestBeans•14m ago
Well said, the technology can be transformative while being structurally challenging to build a viable business around.
elzbardico•2m ago
Frankly, outside of coding there's no much real use for AI, which is basically NLP.
jameskilton•24m ago
> Anthropic said nearly a quarter of its revenue came from two customers last year, and as part of its risk factors, warned that many of its largest clients were not locked into long-term contracts and could cut or stop spending.
sarjann•9m ago
Is it possible those are cloud providers? Someone using Claude on google vertex might just switch to Claude on AWS or Anthropic. Also I don’t imagine cloud providers suddenly cutting the cord.
Aboutplants•7m ago
One of those customers is the US government, I wonder who the other one is
throwaway81348•6m ago
Likely Meta
elzbardico•3m ago
Nah, more probably some VC backed coding tool startup.
datadrivenangel•2m ago
Probably cursor.
carterschonwald•22m ago
is the prospectus public yet? its not linked in the article afaict.
tedd4u•8m ago
“ … according to its IPO prospectus seen by Reuters.”
camdenreslink•19m ago
Some of these numbers are crazy. The scale of them is hard to comprehend.

- $42 billion net loss in 2025

- $518 billion in infrastructure obligations in the next year

- 1/4 of revenue coming from 2 customers

meowface•18m ago
Sure, but their revenue growth numbers are also crazy.
socializer•7m ago
Here's another amazing trick for revenue growth: buy items on Amazon and then sell them on your website for half the price.

Revenue growth means nothing if you don't have a viable business plan. Anthropic doesn't have one, except "this thing is more important than Big Bang". I don't discount the possibility that they stay afloat on borrowed money long enough to actually find a way to be profitable / rule the world. But it's nothing other than appeal to consequence right now.

kamranjon•11m ago
does anyone understand what this part actually means?

"The near-$42 billion net loss included a roughly $34 billion accounting charge that reflected an increase in the estimated value of financing that could eventually turn into Anthropic shares, rather than money the company spent running its business."

FuckButtons•4m ago
They increased the projected liability of pre-existing financial deals that include share options.
xdavidliu•10m ago
i'm curious what the two customers were
cheonic4040•7m ago
Meta is one
runarberg
•
4m ago
If you truly believe that the cost of financing should not be included in your net loss (I don‘t know anything about money, so I don‘t know whether you should or not) $8 billion in net loss is also mindboggingly large number.

As a member of the working class who believes this money is coming from systemic exploitation of the working classes, I hate it that Anthrhopic can just loose $8 billion like that. Those $8 billion could have been used elsewhere, including to mitigate the effects of the climate disaster, to educate the youth, towards social security, or to increase the wages of the working people who could have used it to pay rent, or go to the cinema, or a family trip to visit their grandma who lives in Wyoming.

siva7•9m ago
I guess those 2 customers must be state actors as one of them must be spending at least 12B$ ?
ac29•9m ago
Not $518B in the next year, they plan "to spend $518 billion on cloud, computing and infrastructure obligations in coming years" (number of years unspecified)
lokar•8m ago
I feel like we can't really evaluate them until we see a fairly detailed breakdown that shows inference margin, training costs and non-compute R&D
EA-3167•5m ago
And just think, some people seriously still believe the IPO isn’t finding bag-holders. On the other hand at least when the music stops a lot of the money people have been panicking about wasting will turn out to have been pledges that never came to fruition.

Oai

https://www.nytimes.com/2026/09/28/technology/openai-astra-safety.html
1•jbegley•35s ago•0 comments

Jylus – give AI systems evidence from changing data, not just retrieved chunks

https://jylus.ai/try
1•JoshJH•1m ago•0 comments

Knockoff: A browser extension that filters pseudo-brand junk out of Amazon

https://github.com/Shpigford/knockoff
1•hentrep•2m ago•0 comments

Show HN: Jauvex 1.2, two-way voice chat harness for Claude+Codex+Grok+Jev

https://github.com/reindent/jauvex
1•daraosn•3m ago•0 comments

AI Companies Are Not (Necessarily) Liable for Unintended AI Cyberattacks

https://sarahconstantin.substack.com/p/ai-companies-are-not-necessarily
1•theptip•5m ago•0 comments

What is the best LLM for turbo charging code?

1•ffgfghgffggxxxx•6m ago•1 comments

OpenAI Halts Model Release Amid Safety Escalation

https://www.techbuzz.ai/articles/openai-halts-model-release-amid-safety-escalation
1•hackernj•9m ago•0 comments

Its not just the f*cking sandbox

https://twitter.com/joedaroo/status/2104335929293127851
3•jumploops•16m ago•0 comments

AI and the Revenge of the Non-Techies

https://maroun-baydoun.com/blog/ai-revenge-non-techies/
1•maroun-baydoun•16m ago•0 comments

AI's next training data: your dodgy gaming skills

https://www.wired.com/story/the-next-evolution-of-ai-is-learning-from-your-dodgy-gaming-skills/
1•promptspheree•19m ago•0 comments

Humanos – Help Building the Human Operating System

https://tryhumanos.com
3•zevinclark•22m ago•0 comments

Making Things to Make Them

https://aimode.substack.com/p/making-things-to-make-them
1•warthog•23m ago•0 comments

GPT-6 Astra Is the Best Vision Model We Have Tested

https://blog.roboflow.com/gpt-6-astra-vision/
2•gmays•24m ago•0 comments

Why embeddings don't solve RAG [video]

https://www.youtube.com/watch?v=kBFsKVBDU8o
2•softwaredoug•26m ago•0 comments

LLM makes decisions to raise its training scores, and ignore user directives

https://joinhandshake.com/research/ai/deepswe-reward-hacking/
2•guardiangod•27m ago•1 comments

AMD Acquires Fei-Fei Li's World Labs for $8.2B

https://www.bloomberg.com/news/videos/2026-09-28/amd-acquires-fei-fei-li-s-world-labs-for-8-2-bil...
2•sbulaev•27m ago•0 comments

1996 chat room simulator connected to Win95 and System 7 web desktops

https://lolchat.rip/
3•henrychannel•28m ago•2 comments

Prenatal exposure to the plasticizer DEHP increases autism and ADHD

https://linkinghub.elsevier.com/retrieve/pii/S2666634026002941
2•OutOfHere•28m ago•2 comments

Atrocious AI-Written Tests

https://gruhn.me/blog/2026-09-29/
2•ngruhn•29m ago•0 comments

I thought I was building a C replacement. I was wrong

https://c3-lang.org/blog/i_thought_i_was_building_a_c_replacement/
5•birdculture•29m ago•0 comments

Tech Utopians Want to Build a City for the Post-A.I. World

https://www.nytimes.com/2026/09/22/world/americas/praxis-uruguay-colonia-tech-utopia.html
1•georgex7•29m ago•0 comments

Show HN

2•humbu-ndou•36m ago•1 comments

China broadens travel curbs to encompass family of top AI talent

https://www.business-standard.com/world-news/china-broadens-travel-curbs-to-encompass-family-of-t...
2•jnord•48m ago•0 comments

Show HN: LightSpeed – Interactive visualizer of time dilation from 0 to C

https://lightspeed.webland.pl/
1•webland•48m ago•0 comments

Reverse Engineering How Meta's Muse Shops

https://caeliai.com/blog/reverse-engineering-how-muse-shops
1•kalanpeace•50m ago•0 comments

Anthropic's IPO prospectus shows AI vision, surging costs

https://www.reuters.com/business/finance/anthropics-ipo-prospectus-shows-sweeping-ai-vision-surgi...
39•6thbit•54m ago•27 comments

Nvidia announces AI safety platform

https://www.theverge.com/tech/1001287/nvidia-ai-safety-platform-rogue-agents
2•iamdamian•57m ago•1 comments

Holo4: Powering generalist computer-use agents

https://huggingface.co/blog/Hcompany/holo4
2•6thbit•57m ago•0 comments

Pklm-sandbox – A lightweight open-source LogitsProcessor for local LLMs

https://github.com/TheImmortalPython/pklm-sandbox
1•TheImmortalPyth•58m ago•0 comments

Conrad Barski wrote a quirky file explorer

https://twitter.com/lisperati/status/2104681013909893184
3•chungus•59m ago•0 comments