Hacker News: babelfish

New comment by babelfish in "Muse Spark – Meta Superintelligence Labs"

babelfish — Wed, 08 Apr 2026 16:23:36 +0000

Probably a better link: https://ai.meta.com/blog/introducing-muse-spark-msl/

New comment by babelfish in "System Card: Claude Mythos Preview [pdf]"

babelfish — Tue, 07 Apr 2026 18:56:33 +0000

Totally. Best-in-class for SWE work (until Mythos gets released, if ever, but I suspect the rumored "Spud" will be out by then too)

New comment by babelfish in "System Card: Claude Mythos Preview [pdf]"

babelfish — Tue, 07 Apr 2026 18:26:03 +0000

Combined results (Claude Mythos / Claude Opus 4.6 / GPT-5.4 / Gemini 3.1 Pro)

  SWE-bench Verified:        93.9% / 80.8% / —     / 80.6%
  SWE-bench Pro:             77.8% / 53.4% / 57.7% / 54.2%
  SWE-bench Multilingual:    87.3% / 77.8% / —     / —
  SWE-bench Multimodal:      59.0% / 27.1% / —     / —
  Terminal-Bench 2.0:        82.0% / 65.4% / 75.1% / 68.5%

  GPQA Diamond:              94.5% / 91.3% / 92.8% / 94.3%
  MMMLU:                     92.7% / 91.1% / —     / 92.6–93.6%
  USAMO:                     97.6% / 42.3% / 95.2% / 74.4%
  GraphWalks BFS 256K–1M:    80.0% / 38.7% / 21.4% / —

  HLE (no tools):            56.8% / 40.0% / 39.8% / 44.4%
  HLE (with tools):          64.7% / 53.1% / 52.1% / 51.4%

  CharXiv (no tools):        86.1% / 61.5% / —     / —
  CharXiv (with tools):      93.2% / 78.9% / —     / —

  OSWorld:                   79.6% / 72.7% / 75.0% / —

Better MoE model inference with warp decode

babelfish — Mon, 06 Apr 2026 19:33:09 +0000

Article URL: https://cursor.com/blog/warp-decode

Comments URL: https://news.ycombinator.com/item?id=47665795

Points: 2

# Comments: 0

New comment by babelfish in "The cult of vibe coding is dogfooding run amok"

babelfish — Mon, 06 Apr 2026 19:21:51 +0000

Sure, for humans. Not sure they'll be the primary readers of code going forward

New comment by babelfish in "Cursor 3"

babelfish — Thu, 02 Apr 2026 21:40:02 +0000

Wow, maybe something is wrong with my setup. In Cursor 3, I am clicking "New Agent" at the top left. My root repository is correctly listed on top of the composer, and I clicked the icon to the right of it and selected 'Worktree'. Then, I instruct the model to run `pwd` and tell me it's git status. It's always just on `main` in my root repository. I dug through the settings and couldn't find anything, and after finding this comment[0] on their forums gave up. Would you mind sharing a bit more about your setup/how it works?

[0] https://forum.cursor.com/t/working-with-worktrees-in-cursor/...

New comment by babelfish in "Cursor 3"

babelfish — Thu, 02 Apr 2026 20:24:10 +0000

No per-agent auto-worktree? This is the killer feature of Conductor, having to type `/worktree` into every new chat isn't really a resolution. Not even sure what selecting 'Worktree' for a new chat does

New comment by babelfish in "Cursor 3"

babelfish — Thu, 02 Apr 2026 18:22:20 +0000

different products - CLI vs apps

New comment by babelfish in "Cursor 3"

babelfish — Thu, 02 Apr 2026 18:20:16 +0000

Model independence

New comment by babelfish in "Google releases Gemma 4 open models"

babelfish — Thu, 02 Apr 2026 16:25:17 +0000

Wow, 30B parameters as capable as a 1T parameter model?

New comment by babelfish in "OpenAI closes funding round at an $852B valuation"

babelfish — Tue, 31 Mar 2026 20:31:36 +0000

Maybe? Previous valuation is $730B + $122B raised in this announcement = $852B valuation in this announcement (no actual increase in valuation)

New comment by babelfish in "DigitalOcean Seeks $800M in Funding"

babelfish — Mon, 30 Mar 2026 18:23:34 +0000

same

New comment by babelfish in "Jury finds Meta liable in case over child sexual exploitation on its platforms"

babelfish — Thu, 26 Mar 2026 16:15:58 +0000

Did I say that?

Judge dismisses lawsuit by Musk's X Corp accusing advertisers of illegal boycott

babelfish — Thu, 26 Mar 2026 16:10:17 +0000

Article URL: https://www.reuters.com/world/judge-dismisses-lawsuit-by-musks-x-corp-accusing-advertisers-illegal-boycott-2026-03-26/

Comments URL: https://news.ycombinator.com/item?id=47532275

Points: 14

# Comments: 0

New comment by babelfish in "Jury finds Meta liable in case over child sexual exploitation on its platforms"

babelfish — Wed, 25 Mar 2026 16:02:11 +0000

Yes! As Matt Levine says, “everything is securities fraud”

New comment by babelfish in "Jury finds Meta liable in case over child sexual exploitation on its platforms"

babelfish — Tue, 24 Mar 2026 23:43:41 +0000

Probably their auditors? Lying about this would be tantamount to (very serious) securities fraud. Not sure what you're basing on your allegations on besides "trust me bro"

How to Build a PMF Machine

babelfish — Tue, 24 Mar 2026 17:28:38 +0000

Article URL: https://speedrun.substack.com/p/how-to-build-a-pmf-machine

Comments URL: https://news.ycombinator.com/item?id=47506219

Points: 1

# Comments: 0

Powering Product Discovery in ChatGPT

babelfish — Tue, 24 Mar 2026 16:31:51 +0000

Article URL: https://openai.com/index/powering-product-discovery-in-chatgpt/

Comments URL: https://news.ycombinator.com/item?id=47505264

Points: 3

# Comments: 0

New comment by babelfish in "Finding all regex matches has always been O(n²)"

babelfish — Tue, 24 Mar 2026 00:06:49 +0000

Cursor just wrote a great blog post on this - "Fast regex search: indexing text for agent tools" https://cursor.com/blog/fast-regex-search

New comment by babelfish in "JPMorgan Offers Clients a New Way to Hedge AI Debt Risk"

babelfish — Mon, 23 Mar 2026 18:16:24 +0000

https://archive.ph/OPyKG