HNHacker News
TopNewBestAskShowJobs

mirekrusin

5,755 karma · joined September 4, 2014

https://mirekrusin.com
submissionscomments
mirekrusin··on Alphabet Announces Third Quarter 2026 Results [pdf]
Why their shares tanked 10% then?
mirekrusin··on Artificial Analysis Intelligence Index v5
Open-weight models

Kimi K3.1 Pro: −9

DeepSeek V4.1 Flash 1017: −13

GLM-5.3: −13

MiMo-V2.6-Pro: −16

Proprietary models

GPT-6 Astra: +4

GPT-6.1 Sol: +3

GPT-6.1 Terra: +3

Claude Opus 5.5: 0

Claude Sonnet 5.5: −2

Claude Fable 5.1: −3

Gemini 4 Argon (high): −3

Muse Wonder Full: −6

Claude Haiku 5.5: −7

Muse Wonder: −9

Gemini 3.8 Flash (high): −10

Grok 4.7 (high): −11

Is there affiliation with OpenAI or coincidence?

mirekrusin··on Three new discoveries in mathematics, computer science, and physics
success and failure of anthropic are both good for open weight models
mirekrusin··on Anthropic S-1
It goes down because contracts expire.

AMD is missing, true but commitment schedule isn't known. Also missing estimated additional TPU purchase for 28-31, ~$125B-$150B.

- 2026: $97.08B = Broadcom $26.87B + SpaceX $30.13B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B

- 2027: $97.08B + AMD (TBD) = Broadcom $26.87B + SpaceX $30.13B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B + AMD (TBD)

- 2028: $128.33–134.58B + AMD (TBD) = Broadcom $26.87B + SpaceX $30.13B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B + additional TPUs $31.25–37.50B + AMD (TBD)

- 2029: $128.33–134.58B + AMD (TBD) = Broadcom $26.87B + SpaceX $30.13B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B + additional TPUs $31.25–37.50B + AMD (TBD)

- 2030: $98.21–104.46B + AMD (TBD) = Broadcom $26.87B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B + additional TPUs $31.25–37.50B + AMD (TBD)

- 2031: $98.21–104.46B + AMD (TBD) = Broadcom $26.87B + Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B + additional TPUs $31.25–37.50B + AMD (TBD)

- 2032: $40.09B = Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B

- 2033: $40.09B = Google $14.31B + Amazon $10B + Microsoft $5.61B + Nscale $5.58B + Lambda $4.59B

- 2034: $14.59B = Amazon $10B + Lambda $4.59B

- 2035: $10.00B = Amazon $10B

- 2036: $10.00B = Amazon $10B

Combined total: $782–807B+, including estimated additional TPU purchases and conditional AMD spending.

mirekrusin··on Three new discoveries in mathematics, computer science, and physics
1. it comes from VCs philanthropy, who cares?

2. it wouldn't happen so who cares?

3. no, but models will be better, it'll be cheaper to do

It's not a snark. The scale of VC philanthropy is unprecedented and it fuels itself. All of it will eventually, one way or another, become open. Long story short – there is nothing to complain about.

mirekrusin··on Anthropic S-1
from what you've written it's more like:

- 2026: $97.08B

- 2027: $97.08B

- 2028: $97.08B

- 2029: $97.08B

- 2030: $66.96B

- 2031: $66.96B

- 2032: $40.09B

- 2033: $40.09B

- 2034: $14.59B

- 2035: $10.00B

- 2036: $10.00B

Total: $637.0B

Considering everything that is cancellable, ie. hard commitment only, it's:

- 2026: $56.79B

- 2027: $56.79B

- 2028: $56.79B

- 2029: $56.79B

- 2030: $56.79B

- 2031: $56.79B

- 2032: $29.93B

- 2033: $29.93B

- 2034: $10.00B

- 2035: $10.00B

- 2036: $10.00B

Total: $431B

If they go bankrupt, there are some more cancellations available, also from the other side – if some providers don't deliver, it can be lower as well.

mirekrusin··on Kimi K3.1 Pro and K3.1 Pro Highspeed
I'm sure tools don't like annoying people either.
mirekrusin··on Kimi K3.1 Pro and K3.1 Pro Highspeed
I'm not sure if not doing RL capability gain on offensive cybersecurity can be really called "nerfing"?

They're not post training it to lose capabilities/censor them – they just don't spend training time on this topic, which sounds like a great idea.

It doesn't mean model can't do it, it just means it's less good at it; -10% on benchmark feels like a great result and overall approach to pace it.

mirekrusin··on Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware
It would be great if numbers weren't off, would include optimised versions (ie. speculative decoding), context size, include allowed parallel inference (single agents use parallel inference a lot) etc.

I don't think in general people care about long list of models, they just care about the best one.

Also dual gpu is common setup, there should be ie. 2x 4090.

mirekrusin··on How the Claude Code team does code review
It's conducting/steering/oversight/supervision, not reviewing.
mirekrusin··on Redesigning source control at Anthropic
I think the consensus is to just keep calm and carry on, next model release will sort some of it out.
mirekrusin··on Decisions API is in public beta
copilot is not the best example as it's a bit like openrouter for enterprises.
mirekrusin··on Beam: Reflection's 501B open-weight model
It takes time / few iterations to get it right (and it's moving target), but yes, expensive trial, my personal feeling is that they went a bit too high, at the same time who knows, maybe good move – as they're saying RL didn't plateau. It feels like they had something like $100M budget for it?
mirekrusin··on Beam: Reflection's 501B open-weight model
Who's funding this?
mirekrusin··on Beam: Reflection's 501B open-weight model
Multiple independent approaches are cool and all but fully open source model training (datasets, pipeline, checkpoints) should be taking advantage of being open and share runs/budget between different entities.
mirekrusin··on Huawei and Qualcomm Announce Broad Patent License Agreement
Patents serve cartels. Good luck for innovative startups trying to enter this rigged game.
mirekrusin··on Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
27b runs perfectly fine on 2x 24GB at ~100 t/s (4090) with speculative decoding on 8 bit quants
mirekrusin··on Ember-1
It's an example number, which doesn't matter much but it comes from widely reported $2.30 spent on ops for every $1 of revenue in spring of 2025 for Anthropic.
mirekrusin··on Ember-1
I think people make mistake here, google’s approach is not to spend $2.3 on every $1.0 earned, they’re riding on serving to masses “luna”, they absolutely have way more powerful models internally but they don’t clutter their infrastructure with fragile and costly intelligence-of-size inference frontier. I think “underdog” perception is illusory/temporary, not stupidity - calculated, conscious, longer term bet.
mirekrusin··on DeepSeek Elastic Compute (DSec)
Let's wait a bit and see how your US frontrunners will age with their operating costs above revenue, model pricing collapsing faster than compute efficiency improvements and cost pressure from DeepSeek and friends.
mirekrusin··on DeepSeek Elastic Compute (DSec)
They are as “underdog” as Linux is to Windows.
mirekrusin··on Claude discovers a novel enzyme system with CRISPR-like repeats
Given his background (biophysics PhD, postdoc at Stanford School of Medicine) and that medicine was the core of Machines of Loving Grace back in 2024 it reads less like PR and more like a long-held goal, no? Personally I'm actually surprised it took so long.

Agree trials won't compress much with AI in the near future. But they're starting with basic discovery rather than therapeutics – that part can move fast.

I'd also judge it less by what result is and more by the rate of change – even a year ago ~1k agents running ~1d on single prompt producing wet-lab-verifiable leads wasn't really a thing.

mirekrusin··on AX – Google’s Open Agentic Orchestrator
k8s, but for agents

must look cool for people who want to solve every problem with k8s

it starts with interesting misnomers like "Task" which is not a work item but a sandbox.

"billions of tasks" is a "solution" to problem nobody has (maybe some RL labs? but they solve it other way and with orders of magnitude better optimizations).

freezes design too early – unless they'll actually focus on developing it and make tons of breaking changes it looks shit.

shared state in the same workspace, identity, authority, etc – stuff like that needs to be solved

mirekrusin··on AX – Google’s Open Agentic Orchestrator
yes, has a bad smell of k8s
mirekrusin··on Ask HN: What are you working on? (September 2026)
Thank you, it was inspired by caveman skill hence the name :) It's like formalized caveman. I think there is still some work to make it better/more intuitive/easier.
mirekrusin··on Ask HN: What are you working on? (September 2026)
https://mirekrusin.com/cave

Palantir like system – data, ontology, logic, actions, security, operational workflows – packaged as single cli, local first, git friendly text .cave files and sqlite3 instead of k8s-style complexity blowup.

Working on the weekends.

Next is probably extension of the language to integrate something like https://github.com/mirek/ast for arbitrary ast traversals/modification/generation/better integration with external world - other file types, but ast concept goes further than that, ie. filesystem is also ast, traversal is ast/xpath like walk.

Public domain on all my recent work.

mirekrusin··on Astra for Coding: Why Are We Doing This Again?
After 5d 14h I gave up, not astra https://github.com/mirek/cave/pull/218
mirekrusin··on Astra for Coding: Why Are We Doing This Again?
It's public domain [0] I'll reply here with link to PR/cost/stats once it's done.

[0] https://github.com/mirek/cave

mirekrusin··on Astra for Coding: Why Are We Doing This Again?
There is something odd, I've got single astra session that's now running for... 4d 13h 10m and still going.
mirekrusin··on The Microeconomics of Artificial Intelligence (2025)
you should have answered – "...yes, if you consider your answer an auto complete".
← PreviousPage 2 of 34Next →