Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jeffybefffy519
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
61.
▲
by
jeffybefffy519
6mo ago
Codex is just as bad with this, i've received two ToS warnings for security research activities so far. I have also tried to appeal with zero response.
62.
▲
by
jeffybefffy519
7mo ago
Except this isnt using heavily quantised versions of the model thus reducing quality.
63.
▲
by
jeffybefffy519
7mo ago
Cortical labs have done this before, its their whole thing…
64.
▲
by
jeffybefffy519
8mo ago
Yup exactly, if this is the truth then put it on the terms/privacy policy etc... exec's say anything these days with zero consequences for lieing in a public forum.
65.
▲
by
jeffybefffy519
8mo ago
Someone needs to make an actual good benchmark for LLM's that matches real world expectations, theres more to benchmarks than accuracy against a dataset.
66.
▲
by
jeffybefffy519
8mo ago
I really think the key to addressing climate change should have started 20 years ago in a lot of primary schools, whereby the curriculum includes subjects that are more tailored to solving problems with capturing or converting C02 (eg scien
67.
▲
by
jeffybefffy519
8mo ago
The non dev vibe coders are probably a bigger group of users, and therefore equal more money. Change justified...
68.
▲
by
jeffybefffy519
8mo ago
Is this a privacy nightmare because it exposes graphs of people together publicly?
69.
▲
by
jeffybefffy519
9mo ago
Could this lead into how there are people who thing non ionising radiation sources affect them?
70.
▲
by
jeffybefffy519
9mo ago
I wonder if another waymo ahead could have seen that child earlier and told the main waymo. This would be pretty neat and have a large safety impact.
71.
▲
by
jeffybefffy519
9mo ago
I postulate 90% of the reason openai now has "variants" for different use cases is just to capture training data...
72.
▲
by
jeffybefffy519
9mo ago
Its funny, i think of tools like v0.dev, bolt etc as low code tools too. The definition in my mind is, does the person building the website understand or are they exposed to the underlying tech.
73.
▲
by
jeffybefffy519
1y ago
I still think nVidia has the most to loose in the AI race, optimisations like this will continue coupled with better ASIC's.
74.
▲
by
jeffybefffy519
1y ago
its insane to me that apple could feature "liquid glass"... the amount of devs that must of worked on this is probably not insignificant and could have been used on better things...
75.
▲
by
jeffybefffy519
1y ago
Didnt vista try and do a similar thing when it came out, lots of "widgets" and transparency crap that felt cool then tacky after 2 minutes...
76.
▲
by
jeffybefffy519
1y ago
Gemma3 is pretty useful on a long haul flight without internet
77.
▲
by
jeffybefffy519
1y ago
More ads and a huge push towards shorts, which is really annoying
78.
▲
by
jeffybefffy519
1y ago
Youtube has gotten so much worse in the last 6 months tho, introduction of shorts has devalued the platform terribly and it seems like all the good educational creators are moving off it anyway and now its just ripped crap that is often AI
79.
▲
by
jeffybefffy519
1y ago
Is mechanics of enforcement really a government thing tho?
80.
▲
by
jeffybefffy519
1y ago
Democracy is broken… its so hard to tell who did this.
81.
▲
by
jeffybefffy519
1y ago
what model do you use with cline?
82.
▲
by
jeffybefffy519
1y ago
Taking individuals circumstances into account isn't evidence based medicine though...
83.
▲
by
jeffybefffy519
1y ago
Because evidence based medicine is often about broader problems, medicine 3.0 is about solving problems for a specific person. It could be a rare type of cancer that a specific treatment has been crafted for that person - in this case there
84.
▲
by
jeffybefffy519
1y ago
If you think about it, Evidence Based Medicine is very much against Medicine 3.0 whereby individualized treatments/therapies are used.
85.
▲
by
jeffybefffy519
2y ago
I found it a bit concerning that this doesnt talk about safe dosages of any of the heavy metals.
86.
▲
by
jeffybefffy519
2y ago
Because iPad's
87.
▲
by
jeffybefffy519
2y ago
Arent all humans different and therefore N=1 actually ok in this scenario??
88.
▲
by
jeffybefffy519
2y ago
What is tf2 classic?
89.
▲
by
jeffybefffy519
2y ago
I feel like these guys are missing a pretty important point in their own analysis. I tried setting up a ollama LLM on a fly.io GPU machine and it was near impossible because of fly.io limitations such as: 1. Their infrastructure doesnt supp
90.
▲
by
jeffybefffy519
2y ago
The big mistake here is ownership of your apple devices is an illusion...
More ›