i can't believe all existing meeting apps suck so much. one dev built a better product in 3 days on screenpipe, open source, runs on web and locally

unfortunately i fell into a new addiction, but it's worth it! - Jan 1st week: no meditation - Jan 2nd week: no meditation - Jan 3rd week: Stattera - Jan 4th week: Adderall

from bootstrapped to vc backed

A cry from my heart to everyone who is "non-technical" and searching for a "technical" co-founder or trying to hire a "software engineer": you don't need that at all! Having experienced both paths - first as a CS major in my early years, and later as a business professional - I was able to return to engineering in just 6 months. You can track my progress through my GitHub commits: I went from 0 to 150 commits per month in that time. It's never been easier to dive into coding, thanks to LLM models. Don't know where to start? Ask. Can't read code? Ask. Don't understand what code does? Ask. As you progress, you'll understand more and can do things yourself. Most importantly, you'll learn how to ask the right questions to get better responses from LLMs - cleaner code, working solutions from the first try. As you advance, you'll begin to recognize LLM limitations - where they can't reason or get stuck. This pushes you to improve your technical skills, reasoning, and problem-solving abilities. Being technical today isn't about hard skills like knowing syntax; it's about problem-solving, which many people avoid out of discomfort. I hope to inspire you to step out of your comfort zone. Start your technical journey today - ask your first problem-solving question in Claude or GPT. ps Claude is better, gemini is catching up, o-1 is for hard problems

@wolfofbaystreet actually we don't know if this post was made by me or the operator, haha

@0xasdf_eth @virtuals_io how does it work?

@louis030195 Context matter more than number of parameters

@SahilLalani0 you have to be moderate with your requests

@wolfofbaystreet @basedanarki haha, it’s actually funny how this operator is behind, but probably this is what widespreap public knowledge finally catch up on

@wolfofbaystreet Using anthropic behind the scenes, haha

1. for devs: 2. download: 3. linkedin assistant: @louis030195 @GlavinW @hthieblot @kodjima33 @AlexReibman @claud_fuen @akshay_pachaar @swyx @simonfarshid @DhravyaShah @bsord_dev @charles_irl @belindmo @MikeBirdTech @wolfofbaystreet

@AravSrinivas Generated by screenpipe

@Pawanputra_1 @AravSrinivas Generated by screenpipe

@Pawanputra_1 @AravSrinivas Now we need to generate the next one

@Pawanputra_1 @AravSrinivas That’s even better

@edreisMD That’s just typical big tech, they are so behind the progress, they are like dinosaurs, haha

It’s probably good for disabled people but I can’t see a real use case for someone who can click and type themselves. Why would u delegate the most task to an ai and watch it do it? Just a waste of time. screenpipe agents are much better, they operate on autopilot in background and u just give them tasks to do and don’t have to watch them all the time

i have a pretty dark opinion about where we're heading with ai, but i know what we humans can always be good at. as we've seen in the past, a given ai model or software usually gets really good at a specific domain or task, like playing chess, go, text recognition, etc. as progress continues, the next model often excels at a broader task, one that requires more context, like software coding or writing text summaries. if this trend continues, it's clear that models will gradually expand their usefulness from small, specialized tasks to tackling entire workflows and jobs. but what does that leave for us humans? i think we have one real choice: we have to excel at generalizing. in other words, you don’t want to specialize too narrowly; instead, you want to develop the broadest possible range of skills, industries, and knowledge. you need to stay on top of any given specialty, connect dots across domains, and think outside the box. so, i strongly believe the only way to succeed now is to embrace being a generalist—to be that human-in-the-loop for various ai tools, acting as their supervisor. i just hope i can hold onto this kind of role for a little while longer until models get too good, haha.