Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
petesergeant
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
91.
▲
by
petesergeant
1mo ago
Just want to underscore that this is the right answer.
92.
▲
by
petesergeant
1mo ago
It would be tempting to label this as regular grift, but, one of the few areas the guy’s been consistent is his Quixotic quest against windmills
93.
▲
by
petesergeant
1mo ago
I believe we are on autopilot towards essentially free renewable electricity, and so this doesn't feel like a big concern when held against the potential end-of-humanity potential that AI has.
94.
▲
by
petesergeant
1mo ago
> I feel this is merely stating the obvious. I think there is value in someone of Gates' stature stating the obvious, plainly.
95.
▲
by
petesergeant
1mo ago
I use all three every day, and I am absolutely not finding 5.3 to be better. Competent and in the same league as, sure, but not better.
96.
▲
by
petesergeant
1mo ago
As much as I love Claude/Fable, I have recently found Sol to be dramatically better at this kind of task
97.
▲
by
petesergeant
1mo ago
https://tosche.net/fonts/comic-code
98.
▲
by
petesergeant
1mo ago
I just bought a Nebula subscription, and hadn't heard of it before, so there's some value there
99.
▲
by
petesergeant
1mo ago
“when committed as part of a widespread or systematic attack directed against any civilian population” which I don’t think this really clears, as objectionable as his comments may be
100.
▲
by
petesergeant
1mo ago
I suspect his pitch here is that he knows UX and is technical enough to make it happens, rather than that this is a serious server OS.
101.
▲
by
petesergeant
1mo ago
The OP could well be referring to DHH's recent blog post comparing Roma in Copenhagen to a predatory species that needed to be culled.
102.
▲
by
petesergeant
1mo ago
I don’t think the credit/debit privacy split you’re imagining works like you think.
103.
▲
by
petesergeant
1mo ago
Embeddings are reasonably simple, but it’s a journey to get there, and I am very proud of the dog-heavy explainer I wrote on them: https://sgnt.ai/p/embeddings-explainer/
104.
▲
by
petesergeant
1mo ago
So: "Buy the NVIDIA DGX Spark, subscribe to Perplexity, and we'll turn it into a private local version of our Computer agent, while retaining optional access to our cloud"
105.
▲
by
petesergeant
1mo ago
I thought this was going to be a non-invasive sugar-level device, but it looks like in fact the innovation here is that a popular CGM line can now do ketones?
106.
▲
by
petesergeant
1mo ago
A cool project, but feels like fundamentally the wrong approach when what you do most likely leaves a string of structured digital footprints anyway. I’ve got an agent that fills out my timesheets by looking at git commits, agent history, S
107.
▲
by
petesergeant
1mo ago
Someone else having been too early on a prediction has little bearing on my prediction. A year ago almost nobody was using open models as daily drivers, today they are. When I run out of Fable and Sol credits in a week, I switch to GLM5.3,
108.
▲
by
petesergeant
1mo ago
I think within 12 months we’re going to see a frontier (inc open models) that’s so good at almost all human-directed tasks that which model you use just won’t matter. Only differences that remain will be in deep research or very long-range
109.
▲
by
petesergeant
1mo ago
Yet to find a model that cross-model review doesn’t find a bunch of things wrong with. I’m running simultaneous review with whichever of Grok4.6/GLM5.3/Fable/Sol didn’t write it, and each model tends to find items the others
110.
▲
by
petesergeant
1mo ago
These tests are not useless, but I’d agree (from personal experience) that it’s far from clear that everyone should take one. There’s a degenerative genetic disease that runs in my family that for various reasons we only found out about ten
111.
▲
by
petesergeant
2mo ago
> There is nothing as good as Fable, not even close. This is true, but only for certain tasks. Even as a Fable fanboi, Sol is much better at Fable for some non-programming tasks: Fable for life-planning tasks is miserable because it keep
112.
▲
by
petesergeant
2mo ago
I’m absolutely into agentic coding and you’re still right. A TUI can live easily in a tmux window on a machine I can SSH into from my phone. Also: yes, you can vibecode a UI, but the author’s point reduces to absurd conclusion that you shou
113.
▲
by
petesergeant
2mo ago
Mounjaro, TRT, Lexapro, statins if you need them, and much less alcohol.
114.
▲
by
petesergeant
2mo ago
> we dont have to give up our powers of seeing the truth The truth is that the constitution is interpreted by humans in a common law context, and enforced by the apparatus of state, which has the means to impose its will. Calling this do
115.
▲
by
petesergeant
2mo ago
43 years for me, and in addition to being notably smell-sensitive, I’ve also taken a probably close to 500 long haul flights
116.
▲
by
petesergeant
2mo ago
Which other aggregators that have anywhere near the model selection are cheaper?
117.
▲
by
petesergeant
2mo ago
I worked at an e-commerce clothing store in 2011 that made it abundantly clear you would get your ass handed to you if you went poking around in customer records you had no business looking at for curiosity's sake. I currently work wit
118.
▲
by
petesergeant
2mo ago
> We should stop paying this criminal organization Arguably much of the issue is that it's not a singular organization. It's 18,000 separate law-enforcement organizations. The police in the rest of the wealthy Anglosphere have
119.
▲
by
petesergeant
2mo ago
Fascinating. Now come and try that on the Arabian peninsula and see how you get on
120.
▲
by
petesergeant
2mo ago
OpenRouter’s “moat” is currently “why would I use anything else?”. Stripe has presumably only purchased them because they think there’s consumer-surplus to monetize here, which presumably will soon be giving me lots of reasons to use someth
More ›