Viewing profile — theptip
theptip
HN member- Joined
- Mon, Feb 16, 2015, 9:30 PM UTC
- HN karma
- 11,910
- Public activity
- 3,341 items
- HN profile
- View on Hacker News ↗
About theptip
No profile information was provided.
Recent public activity
-
comment
Comment #49216489
They are investing heavily in this, FWIW. https://about.fb.com/news/2025/10/instagram-teen-accounts-13... I guess you could still make the claim that this is just doing the math on…
-
comment
Comment #49212081
Yeah I can easily imagine the “AI for work” version never being ad supported and gradually capturing more of the value created. Thats a very different usage pattern. But 90% (99%?)…
-
comment
Comment #49211387
No, just report them with error bars. We know the 95th percentile revision size of the distribution. Moot point as the people who actually act on the data understand how to roll th…
-
comment
Comment #49206030
I bet they cache a lot more than OpenAI can though.
-
comment
Comment #49206026
OpenAI can very plausibly target ads far better than Google can. Email is a good window but lots of people talk in more depth with a chat bot. Search is a good “purchase intent mom…
-
comment
Comment #49197860
It’s true in the limit, but I think we are nowhere near that limit. A friend recently started a bio startup and automated parts of mouse experiments enabling higher throughput on i…
-
comment
Comment #49145549
I agree that the ansible isn’t hard sci-fi, it’s not defined rigorously IIRC. However I don’t think hard sci-fi has ever been strictly limited to known physics. It just requires co…
-
comment
Comment #49145438
Interesting, thanks - is the problem you’re describing solved by Rust’s macros (eg derive) or are there further issues you see even there?
-
comment
Comment #49140286
The interest rate is very low.
-
comment
Comment #49118280
Is there a broad benchmark like SWEBench or METR that’s credible for overall progress on dexterity tasks? I’m curious on the trendline for human parity.
-
comment
Comment #49098240
No… it’s really not. There is no “open this spreadsheet with macros turned off” button. You can say “don’t read other documents” but then the main usecase is voided. You can say “r…
-
comment
Comment #49078425
> If you truly believe that bans don't work, the same applies to hardware too. No, software and hardware are different. You can very plausibly prevent smuggling of physical objects…
- story
- story
-
comment
Comment #49060902
I have been using https://www.aihero.dev/skills-grill-me with great success. It’s a very nice small skill for lightweight up-front planning. And if you have that at the top of your…
-
comment
Comment #49059506
And yet demand outstrips supply; so they must be underpricing for the quality level they are hitting.
-
comment
Comment #49059376
Google X was building Loon at least a decade ago to solve internet connectivity with laser-networked high-altitude balloons. They clearly benefit structurally from expanding access…
-
comment
Comment #49052562
Go read Karpathy on this, his telling will be more nuanced than mine for the specific example. But the whole point of NanoGPT is that it’s distilling all of his wisdom and taste in…
-
comment
Comment #49051256
Agreed, all of the staff/principal engineers I work with are ~99%+ AI generated code, and increasing business value delivered as a result. This is on planet scale infra not CRUD ap…
-
comment
Comment #49049388
Thanks. I’ll concede the point and update accordingly.
-
comment
Comment #49048990
One of the key limitations of the last generation was that they biased towards inaction and gave up. (Hence Ralph-loops.) Seems pretty clear that most people wanted these models tu…
-
comment
Comment #49031520
I dunno man, the original announcement said things like: > To gain access, the models identified and exploited a zero-day vulnerability (which we’ve now responsibly disclosed to th…
-
comment
Comment #49030799
I don’t buy it. We are hearing a lot about Mythos capabilities now. If they existed a year ago they would have been trumpeted then. Look, a lot of energy from third parties goes in…
-
comment
Comment #49030768
The problem is, if you explicitly say “this is an eval” then you get different results. It can move the capabilities both up and down on the narrow metric depending on the scenario…
-
comment
Comment #49030319
If it’s true that year-old models could do this kind of thing, how come no one did and then wrote it up? I feel your ”with the right harness” may be doing too much work here. Certa…