Viewing profile — jryan49
jryan49
HN member- Joined
- Thu, Aug 02, 2012, 9:12 PM UTC
- HN karma
- 1,286
- Public activity
- 586 items
- HN profile
- View on Hacker News ↗
About jryan49
Recent public activity
-
story
Ask HN: Anyone else getting token limited at work now?
We used to have infinite tokens but now we are being severely limited. Devs are running out of tokens in 1-2 days which is supposed to last the whole month. Just curious what other…
-
comment
Comment #49149661
It's sad because the ai one is at least one of the worse of the set...
-
comment
Comment #49115366
All these hot takes that LLMs are -10x, 100x, 2x are kind of silly. It really depends on: What you are using it for? What language, what framework, how many LOC, what domain, how m…
-
comment
Comment #49110142
It was a frontier model from a few months ago. It hasn't done it recently. But it has happened. I imagine if there isn't a lot of training data for an idea it probably has a higher…
-
comment
Comment #49109323
I mean ai has generated copyright headers verbatim in code I asked it to make. I had a tutorial open from the web and asked the AI to generate an example for me and it was the exam…
- story
-
comment
Comment #49104897
AI makes mistakes like humans. We used to review human's code because humans make mistakes. What the hell has changed? I mean it's pretty effective to have another agent review the…
-
comment
Comment #48945755
Since every token costs money and I dont think they are conscious, never.
-
comment
Comment #48939394
Squinting is surprisingly effective for me for seeing the hidden text. That's really cool!
-
comment
Comment #48911900
Stocks are long term investments, 10yr+ So you should expect the possibility of a sideways market.
-
comment
Comment #48844376
A tractor is not exactly bleeding edge technology...
-
comment
Comment #48777038
Use a better model like Opus.
-
comment
Comment #48769618
Then why are you reading LinkedIn...
-
comment
Comment #48763529
What is it?
-
comment
Comment #48678095
Rather not waste tokens. Saying thank you literally costs you money.
-
comment
Comment #48671979
So they can train on everyone's copyrighted works to create their model, but when someone trains a model off their model it's not okay? Seems kind of hypocritical.
-
comment
Comment #48594514
I would think modern optimized binaries would be too complicated for current LLMs to handle. LLMs currently get overwhelmed with large code bases, imagine turning that all into bin…
-
comment
Comment #48593736
I mean is your place modest compared to the average place in a third world country?
-
comment
Comment #48593334
If you're made of money sure and then the cases it can't fix I guess you're just stuck.
-
comment
Comment #48593323
Can AI actually read binary code and change it?
-
comment
Comment #48541577
What about AI changes things though? Why didn't you just ignore code quality from humans writing code too?
-
comment
Comment #48391531
Honestly? I don't think we really understand consciousness. So it's kind of hard to say something is or isn't conscious.
-
comment
Comment #48383931
Could anything be more ironic, the employees that work to track every person in the world are now being tracked themselves :)
-
comment
Comment #48383919
I've been thinking about it a lot. I've been looking into becoming an electrician for maybe like 6 years before I retire.
-
comment
Comment #48314672
Kinda buggy, but impressively nonetheless. How long did it take?