sagarpatil

7 ms·

91.
by sagarpatil 2y ago
I’ve been playing with Gemini 2.5 pro throwing all kinds of problems that will help me with personal productivity and it’s mostly one shoting them. I’m still in disbelief tbh. A lot of people who don’t understand how to use LLM effectively
92.
by sagarpatil 2y ago
Could use multiple Google accounts to increase the rate limit.
93.
by sagarpatil 2y ago
Yeah, if I don’t have higher rate limits, it’s useless. This just sounds like a gimmick launch where they want to gather feedback. It will be a couple of months before this will be GA.
94.
by sagarpatil 2y ago
Any idea what is RPM for this model?
95.
by sagarpatil 2y ago
My prompt is nowhere near yours. Just for fun: Asked it to rewrite the first page of ‘The Fountainhead’ where Howard is a computer engineer, the rewrite is hilarious lol. https://gist.github.com/sagarspatil/e0b544313250
96.
by sagarpatil 2y ago
AI labs, it seems, use a template for system cards as well. OpenAI stands out because they showcase their employees using their tools for various use cases, which is refreshing.
97.
by sagarpatil 2y ago
The only benchmark I care about. Thanks!
98.
by sagarpatil 2y ago
Couldn’t have said it better. I’ve developed multiple agents with just function calling, structured outputs and they have been in production for more than a year (back in the day we did not call it agent lol) I think these is targeted tow
99.
by sagarpatil 2y ago
Say you hire a developer and ask him to directly debug an issue by simply skimming through the codebase, do you think he can complete this task say in 5-10 minutes? No, right? In claude code(CC), do the following: 1. /init which acts a
100.
by sagarpatil 2y ago
I have one word: Woow!
101.
by sagarpatil 2y ago
The fundamental issue with these AI safety declarations is that they completely ignore game theory. The technology has already proliferated (see: DeepSeek, Qwen) and trying to control it through international agreements is like trying to co
102.
by sagarpatil 2y ago
Cool story bro. Genuinely.
103.
by sagarpatil 2y ago
Probably the best advice so far. Job = limited upside. I’ve seen a lot of normal people start business and become financially stable.
104.
by sagarpatil 2y ago
Sure you are kidding Mr. Numba88
105.
by sagarpatil 2y ago
I’ll probably get down voted to oblivion, but this is really silly. I do not see any value proposition. To each their own, I guess.
106.
by sagarpatil 2y ago
I’ll do it. AFL right now.
107.
by sagarpatil 2y ago
Just 20 light years away (As a normie I giggled a little).
108.
by sagarpatil 2y ago
Here’s why. Because you can chain the prompts, CoT and answers. Let me explain. Prompt 1 (64k) CoT (32k) Answer 1 (8k) CoT 32k context is not included in the 64k input. So it’s actually 64k + 32k + 8k. Prompt 2 (32k) + Previous CoT 1 (32k
109.
by sagarpatil 2y ago
I just use cursor agent to do git now (unless it’s simple stuff)
110.
by sagarpatil 2y ago
A glorified reminder? Really?
111.
by sagarpatil 2y ago
It’s mind blowing how we went from Adobe Flash to this in just 25 years.
112.
by sagarpatil 2y ago
msty.app does this and much more. It’s open source too.
113.
by sagarpatil 2y ago
As someone who recently moved from a mechanical keyboard to apple keyboard, the closest thing that can imitate the mechanical key sound is through this macos app: https://tryklack.com/
114.
by sagarpatil 2y ago
My only gripe with mechanical keyboards is there’s no version with apple touch ID (I know the technical constraints).
115.
by sagarpatil 2y ago
You can use firecrawl.dev to get markdown content for any page you want. It’s open source and you can host it locally.
116.
by sagarpatil 2y ago
Wow. I was thinking of implementing spaced repetition for a project I’m working on, this will be very useful. Thank you for creating it.
117.
by sagarpatil 2y ago
2 million token size is not enough?
118.
by sagarpatil 2y ago
$99/month lol. I have Perplexity, OpenAI, Claude and Cursor subscription and I end up paying way less than $99/month. Clearly you haven't done any research on price. Aider, Cline are open source, in not sure why someone would
119.
by sagarpatil 2y ago
Been using Cursor since launch. Really frustrating how they charge per message (500/mo) instead of by token usage. Like, why should a one-line code suggestion cost the same as refactoring an entire class? Plus it's been losing con
120.
by sagarpatil 2y ago
o1 models might use multiple methods to come up with an idea, only one of them might be correct, that's what they show in ChatGPT. So it just summarises the CoT, does not include the whole reasoning behind it.

More ›