dinvlad

13 ms·

121.
by dinvlad 3y ago
Yes, for redundancy it could be useful, but doesn’t change the affordability materially.
122.
by dinvlad 3y ago
Not absurdity, it was seriously considered on Nix Discourse and Matrix. Hetzner offers very attractive options like CX line of dedicated VMs, three of which would be sufficient for this project. Alas, this still requires more consideration
123.
by dinvlad 3y ago
(facepalm emoji)
124.
by dinvlad 3y ago
Oh! I haven't even tried Creative, since Balanced was so bad I thought it would be even worse. EDIT: Oh, it's bad! Just lies in more subtle ways.
125.
by dinvlad 3y ago
I've tried it in the last couple days and was very disappointed. Was it better at any point?
126.
by dinvlad 3y ago
I haven’t noticed that, it’s still hallucinating badly.
127.
by dinvlad 3y ago
They are both pretty bad. I ask about templates for CI/CD and they imagine parameters that don’t exist, and no amount of wrestling it around can suppress this. People like to cherry-pick examples where they work great and then proclaim
128.
by dinvlad 3y ago
I think even Amazon does that too still, such a shame
129.
by dinvlad 3y ago
I'm concerned which data these real human beings were forced to train it on. The images they had to review were highly disturbing and illegal in the US.
130.
by dinvlad 3y ago
With Bing chat, it takes me a couple polite prompts saying its answers don't work for it to end the convo on its own. I cannot challenge this "AI" anymore ')
131.
by dinvlad 3y ago
There’s a difference between violation of working conditions like working hours/underpayment, and forcing people to view the kinds of pictures they were forced to see ad nauseam
132.
by dinvlad 3y ago
Exactly, plus the kinds of things these people were forced to see are very illegal in the US. And eventually they had to get out of that job, despite not having good economic alternatives.
133.
by dinvlad 3y ago
It’s not very good at explaining stuff though. One example: ask it to explain a subtle joke, and it will keep failing in funny ways. This is not surprising though, as these kinds of models (LLMs) were specifically optimized for generation,
134.
by dinvlad 3y ago
I’m not sure I know more, but they like to and are incentivized to dream perhaps quite a bit more than needed.
135.
by dinvlad 3y ago
I don’t think they agreed to see the kinds of things they ended up seeing. As a result, they actually pushed to eventually excuse themselves from those jobs, despite not having a good alternative. That’s how bad it was.
136.
by dinvlad 3y ago
They are trained using input from an army of underpaid "ghost workers", i.e. people without many rights or economic freedom, and no consideration for their well-being. Nothing good can be grown on such ground.
137.
by dinvlad 3y ago
They were trained by cheap labor in African countries, using abusive practices with no consideration for ethical concerns. The worst part is they didn't even care about how this would be seen by the public, since the public is just so
138.
by dinvlad 3y ago
I'd even say that the social networks are a precursor to this. Everyone is constantly observed by everyone else there, and many use a fake persona to try to "fit in", or god forbid say something they will regret later. And th
139.
by dinvlad 3y ago
Reminds me of Ms. Casey's "wellness sessions" in Severance
140.
by dinvlad 3y ago
Right? I think it's just the Silicon Valley again and again trying to push its single-minded vision that computers should be used to "solve" all of our problems. They see everything as a nail with a "problem" to sol
141.
by dinvlad 3y ago
I'm kinda appalled that many professors have jumped on the GPT bandwagon so quickly, claiming it should be part of the standard curriculum and how we should have it as the foundation of academic programs, going forward. I don't re
142.
by dinvlad 3y ago
Sometimes people mistake convenience for something being good for them. I can conveniently do so many things in life without having to figure them out by myself, and they may indeed save me time and effort. Doesn’t mean it’s good for me or
143.
by dinvlad 3y ago
For me even GPT4 does that, super annoying
144.
by dinvlad 3y ago
OK let’s say if I don’t, how can I make sure what I get is correct, even after a few iterations?
145.
by dinvlad 3y ago
Give it a couple years, and everyone will forget it and move on to some new shiny
146.
by dinvlad 3y ago
I know right? To me it feels like cheating in an exam, where I ask a know-it-all who gives a straight answer without me understanding it, but then it turns out that the know-it-all didn’t know it very well either, but just appeared confiden
147.
by dinvlad 3y ago
Very well put, thanks for laying it out so clearly.
148.
by dinvlad 3y ago
Same here with GPT-4 via ChatGPT Plus
149.
by dinvlad 3y ago
A waiting list that gave preferential treatment to YC companies, that is. This is according to a recent lawsuit that was filed against OpenAI.
150.
by dinvlad 3y ago
I would say the same about GPT and LLMs in general. For me, it’s been nothing but garbage in, garbage out and a massive waste of my time, so far. Hopefully they’ll figure out how to do symbolic computation in the future so I can finally tru

More ›