Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
wasmainiac
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
91.
▲
by
wasmainiac
8mo ago
I don’t think so. I notice the same thing, but I just use it like google most of the time, a service that used to be good. I’m not getting a dopamine rush off this, it’s just part of my day.
92.
▲
by
wasmainiac
8mo ago
I don't think so. I am aware that large contexts impacts performance. In long chats an old topic will someone be brought up in new responses, and the direction of the mode is not as focused. Regardless I tend to use new chats often.
93.
▲
by
wasmainiac
8mo ago
Thanks for the response, I appreciate it. I do notice variation in quality throughout the day. I use it primarily for searching documentation since it’s faster than google in most case, often it is on point, but also it seems off at times,
94.
▲
by
wasmainiac
8mo ago
Mass-mass-mass-mass good comment. I mean. No I’m having an error - probably claud
95.
▲
by
wasmainiac
8mo ago
Jeez, read the writing on the wall. Don’t pander us, we’ll all got families to feed and things to do. We don’t have time for tech trillionairs puttin coals under our feed for a quick buck.
96.
▲
by
wasmainiac
8mo ago
Dumb question. Can these benchmarks be trusted when the model performance tends to vary depending on the hours and load on OpenAI’s servers? How do I know I’m not getting a severe penalty for chatting at the wrong time. Or even, are the mod
97.
▲
by
wasmainiac
8mo ago
I mean, he has a point it’s just not very eloquently written.
98.
▲
by
wasmainiac
8mo ago
> But if/when they pull it off, it won't really matter Are we in another Cold War Space Race? What matters here? Being better at science? Engineering? Space tourism?
99.
▲
by
wasmainiac
8mo ago
Not to mention… how do you repair it when components fail, especially sensitive electronics against cosmic radiation
100.
▲
by
wasmainiac
8mo ago
NT..? Good comment made me laugh.
101.
▲
by
wasmainiac
8mo ago
Nah, I think the implementation is just off. Graphics need HW acceleration for modern resolutions, but the whole thing should be fine in vanilla JS. Afaik wasm is just an abstraction on top of a jsvm
102.
▲
by
wasmainiac
8mo ago
Wow that is headline carnage. What does it mean? Edit: there is a paywall in my country. Why downvotes? I’m just curious.
103.
▲
by
wasmainiac
8mo ago
Sure, but not for a while since there is a lot of hardware to pay for and maintain.
104.
▲
by
wasmainiac
8mo ago
Awe Haskell! Happy to see it’s still used used in prod!
105.
▲
by
wasmainiac
8mo ago
Typo on “remove (US)” should be remote?
106.
▲
by
wasmainiac
9mo ago
> Zen coders Please don’t, it’s just my day job.
107.
▲
by
wasmainiac
9mo ago
I would and people do but no one respects the license because it is largely uninforceable
108.
▲
by
wasmainiac
9mo ago
This. I feel like the sentiment on HN is very binomial. For me my experience with LLMs is very much what you experience. Anything outside of generic tasks fails miserably. I’m really curious how people make it work so well.
109.
▲
by
wasmainiac
9mo ago
I used to use Dingtalk, it had a similar feature iirc. I couldn’t find a source, I can’t read mandarin
110.
▲
by
wasmainiac
9mo ago
Why not debate me rather than downvotes? Eh hallucinations break my workflow and end up costing me more time debugging then it’s worth.
111.
▲
by
wasmainiac
9mo ago
It’s no more useful than when google and stack overflow was at its peak! All I want is to find docs. The coding performance is lackluster, oversold and under delivered. Everything else gen AI is dystopian.
112.
▲
by
wasmainiac
9mo ago
> I think you're anthropomorphising the AI too much I don’t get it. Is that supped to be a gotchya? Have you tried maliciously messing with an LLM? You can get it into a state that resembles psychosis. I mean you give it a context t
113.
▲
by
wasmainiac
9mo ago
Have a source for that? I think you are anthropomorphising the AI too much. Imagination is inspired by reality, which AI does not have. Introducing a reality which the AI fully controls (looking beyond issues of vision and physics simulatio
114.
▲
by
wasmainiac
9mo ago
> Because large organisations get like this; Still it does not need to be this way. Large organisations used to actually get s** done generally in budget and on time. Now we can’t even do a simple tasks without mountains of paper work an
115.
▲
by
wasmainiac
9mo ago
> This is pretty normal for government procurement, though Why accept the status que? How many working lives of tax revenue did this bs consume?
116.
▲
by
wasmainiac
9mo ago
Possibly, but 1 I am concerned that the current LLM AI is not thinking critically, just auto completing in a way that looks like thinking. 2 current AI rollout is incentivised for market capture not honest work.
117.
▲
by
wasmainiac
9mo ago
If my kid gets addicted to fent I will get in shit, regardless that Purdue Pharma was found guilty. Point is Purdue Pharma is guilty for hooking people on an addicting substance.
118.
▲
by
wasmainiac
9mo ago
> I just don't really know where the line It is developed to be as addictive like a drug, but it’s not even fun. Just stupid mind numbing content.gambling does the same thing, and many jurisdictions have outlawed it for minors.
119.
▲
by
wasmainiac
9mo ago
The state of publishing in academic was already a dumpster fire, why lower the friction farther? It’s not like writing was the hard part. Give it two years max we will see hallucination citing hallucination, independent repeatability out th
120.
▲
by
wasmainiac
9mo ago
I don’t know how you ever get the opposite, in been working for 12 years now at 6 different jobs everyone has been fine at the job apart from a cheeky hack here or there.
More ›