HackerTrans
トップ新着トレンドコメント過去質問紹介求人

simianwords

1,304 カルマ登録 昨年
if you wanna be friends [email protected]

投稿

Can AI do fact-checking?

wired.com
4 ポイント·投稿者 simianwords·6 日前·0 コメント

Use AI for reviewing code especially when the diff is huge

simianwords.bearblog.dev
7 ポイント·投稿者 simianwords·19 日前·8 コメント

Ask HN: How are thinking efforts implemented?

104 ポイント·投稿者 simianwords·先月·31 コメント

MCP Needs an Approval Button

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·先月·1 コメント

Why Deep Learning Works Even Though It Shouldn't

moultano.wordpress.com
4 ポイント·投稿者 simianwords·先月·0 コメント

Investigating the hidden moat behind all the LLM apps

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·2 か月前·1 コメント

The state of AI voice assistants is bad but there's a clear winner

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·2 か月前·0 コメント

The generation vs. verification delta explains why LLM's are useful

simianwords.bearblog.dev
1 ポイント·投稿者 simianwords·2 か月前·0 コメント

Omni 1

ai.google.dev
5 ポイント·投稿者 simianwords·2 か月前·1 コメント

AI can fix the fragmented online public transport space

simianwords.bearblog.dev
1 ポイント·投稿者 simianwords·2 か月前·0 コメント

A new wave of software efficiency is coming

simianwords.bearblog.dev
1 ポイント·投稿者 simianwords·2 か月前·0 コメント

Redesigning Agent Skills – two missing parts

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·2 か月前·0 コメント

Grok 4.3

docs.x.ai
405 ポイント·投稿者 simianwords·2 か月前·529 コメント

[untitled]

1 ポイント·投稿者 simianwords·2 か月前·0 コメント

The LLM costs are not going up

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·3 か月前·0 コメント

The LLM costs are not going up

simianwords.bearblog.dev
3 ポイント·投稿者 simianwords·3 か月前·0 コメント

[untitled]

1 ポイント·投稿者 simianwords·3 か月前·0 コメント

Redesigning Agent Skills – two missing parts

simianwords.bearblog.dev
1 ポイント·投稿者 simianwords·3 か月前·0 コメント

Improving Agent Skills user experience by adding hooks

simianwords.bearblog.dev
2 ポイント·投稿者 simianwords·3 か月前·0 コメント

Under these conditions, LLM's basically never hallucinate

simianwords.bearblog.dev
1 ポイント·投稿者 simianwords·3 か月前·2 コメント

コメント

simianwords
·6 分前·議論
Books that haven’t been touched or opened in years have far more use as training fodder than a nice museum piece
simianwords
·12 時間前·議論
You have zero theory of mind if you think the poster is a libertarian instead of making a tongue in cheek comment
simianwords
·13 時間前·議論
why would you use one over the other?
simianwords
·13 時間前·議論
what's the difference between Sol Ultra and Sol pro? is pro a thing of the past now
simianwords
·13 時間前·議論
they can now https://www.youtube.com/watch?v=8vvWTz6N7Qg
simianwords
·14 時間前·議論
Temporal awareness with GPT-Live

https://www.youtube.com/watch?v=8vvWTz6N7Qg
simianwords
·16 時間前·議論
I’m still not exactly at the point where I don’t need the IDE. I still like Idea products for their indexing and navigation - I’m just used to them.

Codex doesn’t yet do a good job showing diffs nicely. Nor with navigation.

I agree I reach out to the IDE less and less now.
simianwords
·16 時間前·議論
I said basically the same thing in my comment which is buried. The phenotype of developers using Vim is completely different from the ones using AI.
simianwords
·19 時間前·議論
This article tries to sneak in class hierarchy concepts while it is really forced. Hierarchy is when subscription tiers exist? lol?
simianwords
·19 時間前·議論
The intersection of people who like vim and ai is basically nil
simianwords
·19 時間前·議論
Recent reports that OpenAI found optimisations would explain that these are legit
simianwords
·一昨日·議論
GPT Terra is 50% cheaper than 5.5 while being more performant. So it’s like a straight up 50% reduction in cost!

That leads me to a question. Why wouldn’t they just default to terra in ChatGPT in the last few months? If they didn’t then they burnt money for no reason by giving a shittier model at a higher price
simianwords
·一昨日·議論
Very interesting. My prediction is that Mythos would outperform Sol.

Also what does this tell about Yann LeCuns whole world model theory? Bro has been going on and on about it. He has made multiple wrong predictions on the trajectory of LLMs.

At some point his claim should be fully falsified no?
simianwords
·一昨日·議論
Downvoted comment but I did find this comparison aggressive and tacky.
simianwords
·一昨日·議論
> SWEBench Pro, which is a pretty high-quality benchmark

No, doesn't seem like it

https://openai.com/index/separating-signal-from-noise-coding...
simianwords
·一昨日·議論
> On the Artificial Analysis Coding Agent Index, GPT‑5.6 Sol with max reasoning sets a new state of the art at 80, 2.8 points above Fable 5, while using less than half the output tokens, taking less than half the time, and costing about one-third less.

> That advantage extends across the family: Terra performs just above Fable 5, while Luna outperforms Opus 4.8; each does so in roughly one-third of the time, with about half as many output tokens, and at approximately one-quarter the estimated cost.

Wow. I don't believe it. Every indication and twitter post told me that Fable is much more intelligent than Sol and here we are told that even Terra outperforms Fable?

Not only that, Sol doesn't even come with run time classifiers. So it is even more suspicious.

What's even stranger is that OpenAI is directly referencing a competitor in this direct way.
simianwords
·一昨日·議論
Codex UI is way way way better than Claude Code

- codex UI is much more responsive

- i get feedback about the progress easily

- the tool calls and results are very legible, I can click them and see the progress

- no one talks about this but the tool call and response notification are handled much more elegantly in Codex. In Claude Code, it is handled in a clunky way using loops which always causes some delay

- you can steer the conversation midway in Codex

- /side is underrated (/btw is the equivalent and is much worse in Claude Code)

- I have to admit subagents are handled better in Claude Code
simianwords
·一昨日·議論
I notice something more interesting. This post shows Andrew to not only personally criticise Ben but also clearly shows an ideological stance against AI. I can see it from multiple angles - refusing AI PR's, refusing Anthropic's donation and multiple other things.

Either this ideology helps Zig position itself as a hand crafted language. Or this ideology is self defeating.
simianwords
·一昨日·議論
What if other competing companies adopt it aggressively and push out features faster and get more market share?

What if pushing features faster brings more money because users like the features?
simianwords
·一昨日·議論
50 years ago we had religion for people to have something common to believe in. Now we have this. I think most people want a common enemy/scapegoat and an in-group presence. I doubt its about morals but more about self identity.

A poster who reactively posts "but Elon is a Nazi!!1!" does it not do out of care for Jews but more for establishing their own self identity. Within the mini-group, the poster aims to get the moral high-ground and status by speaking out against Nazi salute. It is absolutely not a moral thing because Natanyahu himself wrote a tweet vindicating Elon.

The same type of people show excessive concern towards AI's climate impact, "big tech bad" etc.

Ultimately its a new religion replacing the old one.