sirk390

6 ms·

1.
ChatGPT's bug with scanned PDFs(medium.com)
1 points by sirk390 5mo ago|0 comments
2.
by sirk390 7mo ago
I would use this as a human. That npm output is crazy. Maybe a better variable would be "CONCISE=1". For LLMs, there are a few easier solutions, like outputing in a file (and then tail)., or running a subagent
3.
by sirk390 1y ago
All that complexity when you could simply do: <td class="<%= @teacher.available_places> 0 ? 'bg-colour-green' : 'bg-colour-red' %>">
4.
Using the OpenAI Reatime API in Python(medium.com)
2 points by sirk390 2y ago|1 comments
5.
by sirk390 2y ago
> France ("consistently invested in nuclear for the past half-century" ). Not really and this is currenctly causing a big problem. France stopped building new reactors after 2002. They only built 1 new generation EPR, which was
6.
Google (and others) please implement link analytics with "ping" or "sendBeacon"(medium.com)
6 points by sirk390 2y ago|4 comments
7.
by sirk390 2y ago
Ruby is an excellent language/platform and you can do a lot more than using Rails. But yes I don't know for NLP.
8.
Show HN: Street, a Local second hand item market app using AI(streetapp.com)
2 points by sirk390 2y ago|0 comments
9.
by sirk390 3y ago
I you hang them at 45 degree the depth will be reducted by sqrt(2). (about 0.7 x hanger length), and you will lose space on each side. And the more you increase the angle, the more you will lose space on the sides. With this technique, you
10.
by sirk390 3y ago
Better, but irrational numbers can have a predictable pattern like "0.110100100010000" ( E.g "0.1 10 100 1000 10000" )
11.
by sirk390 3y ago
For example you could set cookies before visiting another website. This is currently impossible in an iframe but possible in a browser. I've wanted to do this to automatically login users on some external websites.
12.
by sirk390 3y ago
I really though you were going to say "Chat-GPT detected" when quoting that message.
13.
by sirk390 3y ago
With compositions A uses B but B can never use A. With inheritance Child can use the Parent, but Parent will also call the Child (virtual methods) which in turn can call the Parent again etc.., so the code can become difficult to follow.
14.
by sirk390 3y ago
Humans are pretty bad at these questions. Even with the simplest questions like "Sally (a girl) has 3 brothers. Each brother has 2 sisters. How many sisters does Sally have?" I think that a lot of people will give an incorrect ans
15.
by sirk390 3y ago
> Whereas, in biological brains, the weights are updated continuously. My personal impression is that many "weights" are updated during sleep time. For example when training juggling, I will make no progress at all for hours of
16.
by sirk390 3y ago
In humans, sleep is also required fo learning, so it's not fully continuous. An AI that occasionnaly retrains using the new knowledge would still be very interesting.
17.
by sirk390 3y ago
It could be double speak, but the obvious thing you understand from this formulation it that he won because he is a philosopher. Not that he won, and by the way he is a philosopher.
18.
by sirk390 4y ago
lambdas are very useful for example for sorting data using a key function. Maybe the auther is saying that it should be called differently, or have a different syntax and the title is just clickbait.
19.
by sirk390 4y ago
Is outperforming GPT-3 still a good reference? It seems there are many models outperforming GPT-3 in the superglue benchmark: https://super.gluebenchmark.com/leaderboard/ GPT-3 is in position #21, with 71.8% score. The
20.
by sirk390 4y ago
And if every parameter is one byte, the minimum, it will take at least 70gb to save or share this model. So it's still way to big to package directly in a app.
21.
by sirk390 5y ago
I doubt that there are numbers to prove that. SMR might just be a buzzword. From the page below there don't seem too many operating or under construction (5 operating, and 4 under construction). https://www.world-nuclear.org
22.
by sirk390 5y ago
Crazy, but who thinks AI is not moving forward? I don't think anyone on HN.
23.
by sirk390 5y ago
Most of the time it's not compute or memory intensive so any computer will do. The most important thing then is too have a good chair, keyboard and mouse!
24.
by sirk390 5y ago
Just don't lookup your private key in there. This would be like giving it to the website
25.
by sirk390 5y ago
If you only look at the json format it's true. But it would be way worse if some decoders returned decimals and some returned floats.
26.
by sirk390 5y ago
I don't thing there is a more "correct" representation. Representing it as a string is equaly correct. Blockchain or banking often represents money as an integer with the smallest divisible unit (cent, or satoshi), but it is
27.
by sirk390 5y ago
timestamps and decimal are the two most useful additions compared to json. They would be nice to add to json if that is somehow possible.
28.
by sirk390 5y ago
But decimal are way more useful as they can represent currency amounts. It would be strange to show a currency amount like "3/4" or "11/12". Personally, the two datatypes I have always been adding manually to j
29.
by sirk390 5y ago
Indeed, I didn't know about the ping attribute. However, you will see they still do click tracking on the google image results for example. And if you check this article from 2013, you will see that they had it back then (I'm not
30.
by sirk390 5y ago
Https multiplied latency by three when opening websites compared to 10/15years ago where many websites were http. And also many websites include more tracking, for example Google search results used to directly send you to the result p

More ›