deepanwadhwa

5 ms·

1.
by deepanwadhwa 11mo ago
Not for any serious positions I bet. Only where they want to do dirty stuff like killing or stalking other humans. It's like recruiting for army- you get them before they learn how to use their brains.
2.
by deepanwadhwa 11mo ago
-> GPT 3.5 was awesome at chess I don't agree with this. I did try to play chess with GPT3.5 and it was horrible. Full of hallucinations.
3.
by deepanwadhwa 11mo ago
Does anyone know if there is anything like CRDT with end to end encryption?
4.
by deepanwadhwa 11mo ago
umm, I am not a fan of any of the recent new browsers but what's wrong with Chromium in itself? I think Chromium is pretty good, technologically mature, foss.
5.
by deepanwadhwa 11mo ago
I agree with you but is it scalable?
6.
by deepanwadhwa 11mo ago
Thank you for the responses. I understand the issue a bit more now.
7.
by deepanwadhwa 11mo ago
Are you directly sucking the cow? If yes, I'd support you drinking raw milk.
8.
by deepanwadhwa 11mo ago
Wait, can anyone help me understand how would they enforce this? All the AI detection tools I have reviewed failed miserably at detecting AI in text.
9.
by deepanwadhwa 11mo ago
Not sure if I follow really. Cooling from it's own generated heat? Are we even sure the system would get that hot in the first place? The temperatures can plunge up to -200 degrees. If needed, they'd cool it just like they keep t
10.
by deepanwadhwa 11mo ago
It doesn't sound strange at all. Good question.
11.
by deepanwadhwa 11mo ago
Nice find, there is a limited length for each stroke though :)
12.
by deepanwadhwa 11mo ago
Gemini2.5 Pro has assisted me better in every aspect of AI as compared to ChatGPT5. I hope they don't screw up Gemini 3 like OpenAI screwed ChatGPT with GPT5.
13.
by deepanwadhwa 11mo ago
How does it make us less human?
14.
Trap the Critters with Paint(deepanwadhwa.github.io)
60 points by deepanwadhwa 11mo ago|23 comments
15.
by deepanwadhwa 1y ago
I did not know GPU backed loans was a thing. It seems risky.
16.
by deepanwadhwa 1y ago
https://zulip.com/help/import-from-slack
17.
by deepanwadhwa 1y ago
Very cool.
18.
by deepanwadhwa 1y ago
What makes you so sure that the exploit is over? Maybe they wanted their secondary exploit to get caught to give everyone a sense of security? Their primary exploit might still be lurking somewhere in the code?
19.
by deepanwadhwa 1y ago
The apple example seems off, but literacy is still going to be a requirement for this kind of thing to work. First, you need to be able to understand the language(grammar, vocabulary etc) to communicate with this device. The guy in the demo
20.
by deepanwadhwa 1y ago
100% agreed. Using LLMs locally is the gold standard but not everyone has the infra for that (not yet at least).
21.
Add a privacy layer to your LLM app as AI companies' privacy policies evolve(medium.com)
3 points by deepanwadhwa 1y ago|2 comments
22.
by deepanwadhwa 1y ago
This is very cool. Is my understanding correct that you are trying to put a human in the loop? I would have loved if there was a demo video showing exactly how the approval process is for long documents. Is this something similar - https:&
23.
by deepanwadhwa 1y ago
I noticed it's privacy-first, which is very nice. Is the AI running locally?
24.
by deepanwadhwa 2y ago
Is ahocorasick in a different plane than this?