Hacker News: ttul

New comment by ttul in "Which tools do Claude, Codex and Cursor choose? We measured 17k runs to find out"

ttul — Fri, 04 Sep 2026 15:59:01 +0000

This is undoubtedly true. Agents are extremely analytical and trained to be objective - far more so than humans. They are not driven by emotion. If you have good stuff and you make it extremely clear to everyone through your documentation, this is more likely to be persuasive to agents than to humans.

I for one would prefer a future in which the nuances of a good product can shine through without layers of bullshit.

New comment by ttul in "Which tools do Claude, Codex and Cursor choose? We measured 17k runs to find out"

ttul — Fri, 04 Sep 2026 01:02:47 +0000

I built this for my own company. Armature is on to something. You start by analyzing the choices agents would make for various use cases and then glean what, if anything, you might do to start tilting the agents in the direction of your own product and away from the competitor.

Selling to agents is similar to selling to humans. You dump money into marketing to make sure agents find your solution around every corner for every use case you’re well suited to.

New comment by ttul in "GPT-6 Astra"

ttul — Thu, 03 Sep 2026 19:53:03 +0000

"The gym's doors were mysteriously removed from their hinges during the night. The gym equipment was also apparently stolen. And the school's custodian was found incoherent next to a bottle of top-shelf Scotch."

New comment by ttul in "Gemini 3.8 Flash and 3.8 Flash Cyber"

ttul — Wed, 02 Sep 2026 18:48:47 +0000

Will look forward to the "feel" of the model in real testing. But I agree that these benchmarks do get "dealt with" rapidly. That's a shame, but I guess it's the times we live in.

New comment by ttul in "Gemini 3.8 Flash and 3.8 Flash Cyber"

ttul — Wed, 02 Sep 2026 15:49:08 +0000

Crushing it on DeepSWE is a very big deal. Excited to give this a try.

New comment by ttul in "My local model setup on an M4 Pro Mac Mini"

ttul — Wed, 02 Sep 2026 04:53:49 +0000

Most people running local models would probably love to run larger models if only they had access to big enough hardware. I'm curious: to those of you running models locally, if there was a way to inference the model of your choice at a reasonable cost by effectively time-sharing a B300 rack through some privacy-protecting intermediary, would you consider that?

If there was a "Mullvad of GPU clouds", would that solve the privacy concerns?

Endless sitcom using Minimax H3 and a turbo LoRA

ttul — Mon, 31 Aug 2026 02:19:05 +0000

Article URL: https://www.twitch.tv/dereactorwah

Comments URL: https://news.ycombinator.com/item?id=49504884

Points: 5

# Comments: 5

New comment by ttul in "OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)"

ttul — Fri, 28 Aug 2026 21:01:51 +0000

It burns out so quickly on the 5x plan. Better than nothing, I suppose, but I don't know how I would survive on a 5x plan given that I burn out more than one 20x plan monthly.

New comment by ttul in "DeepSeek-v4-flash-vision-exp"

ttul — Mon, 24 Aug 2026 16:59:01 +0000

Luna is a very capable model - thanks for pointing that out. Terra is the strange one: not cheap enough or intelligent enough to be on the frontier. But Luna sure is.

New comment by ttul in "OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)"

ttul — Mon, 24 Aug 2026 16:57:46 +0000

Fable 5 is just straight up a larger model - I'm guessing at this, but there is plenty of evidence online from people far more plugged in than I am. OpenAI is pursuing a strategy that yields greater operating margins and penetration of their model to developers. Fable's high cost makes it so premium that Anthropic has to reserve it for only the richest customers and corporate users. That's not a winning formula long term.

I believe the reason we have not seen a Fable-level model from OpenAI yet is because doing so would box them in on costs just as harshly as it has boxed in Anthropic. They are letting Anthropic make this mistake.

New comment by ttul in "DeepSeek-v4-flash-vision-exp"

ttul — Fri, 21 Aug 2026 14:28:30 +0000

The DeepSWE benchmark they report (59.3%) overlaps with the confidence interval of 5.6-Sol Medium (61% +/- 2%), but likely at 1/18th the cost (they did not report the DeepSWE benchmark cost, but v4-flash had this cost ratio against Sol Medium).

Interestingly, v4-flash performed several points worse on DeepSWE at 53% +/- 4%. Assuming this result is verified by DeepSWE officially, it would mark a significant advance in Pareto cost/performance on software engineering tasks.

New comment by ttul in "DeepSeek-v4-flash-vision-exp"

ttul — Fri, 21 Aug 2026 14:20:08 +0000

A good share of humanity would have also gotten this question wrong!

New comment by ttul in "Asana cleared 5 years of engineering work in 2 weeks with Codex"

ttul — Thu, 20 Aug 2026 06:38:13 +0000

If you ask Sol or Claude how much time it will take to implement a plan they just came up with, they usually advise a timeframe in the weeks or months - assuming, I suppose, that human programmers will be building it. And then you ask the model to just "do it" and it takes an hour or two. I always find this entertaining.

New comment by ttul in "OpenRouter is joining Stripe"

ttul — Wed, 19 Aug 2026 23:05:33 +0000

Poor BrandonM... He has not been invited to cocktail parties at all since then.

New comment by ttul in "Cerebras CS-4"

ttul — Wed, 19 Aug 2026 15:59:37 +0000

Hot water for the whole neighbourhood!

New comment by ttul in "Cerebras CS-4"

ttul — Wed, 19 Aug 2026 02:45:30 +0000

Indeed. You need 45 to 60 liters per second of cooling water flowing over a Cerebras wafer every minute to keep it under 90C. And that’s assuming the water leaves at 90C…

More realistically, you need much more cooling water.

New comment by ttul in "Cerebras CS-4"

ttul — Wed, 19 Aug 2026 02:40:22 +0000

I’ll get that 250kW home power service dropped in next week!

New comment by ttul in "Zapping Rocks Unlocks Stimulated Geologic Hydrogen"

ttul — Sun, 16 Aug 2026 04:25:25 +0000

(Otherwise it would have slowly transformed into CO2 over the eons without our help)

New comment by ttul in "Zapping Rocks Unlocks Stimulated Geologic Hydrogen"

ttul — Sun, 16 Aug 2026 04:24:59 +0000

I think there is literally no oxygen down there with the hydrogen. It’s the same with hydraulic fracturing of hydrocarbons. Plenty of methane COULD go boom, but there is no oxygen deep underground.

New comment by ttul in "I checked 30 frontier model cards. Here are the benchmarks labs report"

ttul — Sun, 16 Aug 2026 04:18:49 +0000

I am waiting with bated breath to read, “load-bearing” somewhere… The latest models are very capable, but sometimes they seem to get so deep in the details that they lose the overall plot.

What the hell is the point of this page? Can you put in a single bit of human prose explaining why it exists and what we are supposed to learn?