mnk47

9 ms·

61.
by mnk47 2y ago
>the guy who quit/lost his job at OpenAI because he didn't agree with their corporate shift and departure from the original non-profit vision There is no evidence of this being true. He is one of the biggest proponents of keepi
62.
by mnk47 2y ago
>It sounds like an overly dramatic person pretending to be a robot That's precisely what it was ordered to do.
63.
by mnk47 2y ago
You can tell it to talk in a robotic, unrealistic way and it will do so. Here is a demo from their presentation: https://youtu.be/D9byh4MAsUQ
64.
Chain of Thoughtlessness: An Analysis of Cot in Planning(arxiv.org)
2 points by mnk47 2y ago|0 comments
65.
by mnk47 2y ago
FWIW, that one guy on twitter is spot on more often than not in his leaks and predictions.
66.
by mnk47 2y ago
>You're posting under a thread where many seniors are discussing how they don't want this because it doesn't work. Many artists and illustrators thought AI art would never threaten their livelihood because it did not under
67.
by mnk47 2y ago
Most people don't love their jobs. Switching careers isn't easy, and we all have bills to pay.
68.
by mnk47 2y ago
Yi Tay's response (chief scientist at Reka AI, ex-Google Brain researcher): https://twitter.com/YiTayML/status/1783273130087289021 >not true, especially for language. if you trained a large & deep MLP
69.
by mnk47 2y ago
Can you elaborate on why this is self-serving and manipulative? I disagree with him on this, but I think he's being honest. He left his cushy job at Google Brain due to ethical concerns.
70.
GPT-4 users, how are you using it? Is ChatGPT Plus still worth it?
4 points by mnk47 2y ago|5 comments
71.
by mnk47 2y ago
>at least with the AI chatbot, there's not a person being sexually objectified because they need to pay rent next month. I would love to be sexually objectified for a living if I had the looks and the bravery to put myself out the
72.
by mnk47 2y ago
>This means that software engineers will need to be more like software designers. Do you have any suggestions on how to improve my software/system design skills? Any book or course recommendations would be appreciated.
73.
by mnk47 2y ago
Note that this isn't new research, it's essentially just a fancy blogpost by a researcher, but I found it very interesting. For what it's worth, Yann Lecun retweeted it earlier today.
74.
Can Large Language Models Reason and Plan?(arxiv.org)
4 points by mnk47 2y ago|1 comments
75.
Wu's Method Can Boost AlphaGeometry to Outperform Gold Medalists at IMO Geometry(arxiv.org)
7 points by mnk47 2y ago|1 comments
76.
by mnk47 2y ago
full title: Wu's Method can Boost Symbolic AI to Rival Silver Medalists and AlphaGeometry to Outperform Gold Medalists at IMO Geometry
77.
by mnk47 2y ago
>I work at OpenAI I know there's a lot you can't talk about. I'm not going to ask for a leak or anything like that. I'd just like to know, what do you think programming will look like by 2025? What do you think will h
78.
by mnk47 3y ago
I watched it on Prime Video (US). Pretty sure it's free to watch on Viki as well. Apparently it's a very faithful adaptation. It's not perfect. Sometimes the music/audio quality is subpar. and there are filler scenes lat
79.
by mnk47 3y ago
I watched a talk by the OpenAI Sora team [1] yesterday. They achieved amazing results with what they called "the GPT-1 of video", making a huge leap from the ugly messy low quality GIFs we were getting before Sora. It understands
80.
Ask HN: Usefulness of formal verification (Coq) and formal specification (TLA+)?
3 points by mnk47 3y ago|1 comments
81.
by mnk47 3y ago
>On the other hand, if we can expect lower power costs and higher trust in the output (i.e. needing less human intervention) then it makes sense to start finding places where it can fit into the business and grow over time. So far, simpl
82.
by mnk47 3y ago
Yeah, I think we're just at an awkward point where you still have to do some prompting tricks to make it work as expected. Hopefully this all gets smoothed out soon and makes hacky solutions like "prompt engineering" less imp
83.
After 2 Weeks of Testing, What Do Developers Think About Claude 3?(favtutor.com)
3 points by mnk47 3y ago|0 comments
84.
Ask HN: Anyone here using Gemini Pro 1.5? How does it compare to Claude Opus?
1 points by mnk47 3y ago|1 comments
85.
by mnk47 3y ago
Well, I did see that, or at least people who claimed to be seniors online, plus some mid-level friends and yes, a junior friend. Admittedly, it's just anecdotes, but it's very difficult to get hard facts, even from perusing bls.go
86.
by mnk47 3y ago
It took me much longer than usual to get past the depression phase. The main reason is that everyone online kept telling me I was dumb, too inexperienced or extremely incompetent if I thought AI was a serious threat to me, a junior develope
87.
by mnk47 3y ago
>Which is something AI companies are doing at every level. The message I keep hearing time and time again is "Our aim is not to replace but to empower". The same companies with this message are also saying that their goal is to
88.
by mnk47 3y ago
To be fair, they mostly faked the near instantaneous, real-time flow of the conversations. The answers were, as far as I know, legit. But I still agree that we should be skeptical.
89.
by mnk47 3y ago
I have gastroparesis, which often gives me bloating/pain that lasts a long time, and if I'm too hungry at night, I have to choose between trying to ignore hunger and not sleeping for the next ~3 hours (even a snack or a cup of wat
90.
by mnk47 3y ago
His views, biased as they may be, match the scientific consensus and are backed up by decades of research. Many of his Youtube videos are just him going over studies and meta analyses, and explaining the findings in layman's terms. I h

More ›