Sol (@solyatic) header
Sol (@solyatic) profile picture

AI engineer | latest products and tools https://xitter.gq/t.co/GVFSRmYlHf

~/.ssh
Joined January 2026
Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)Photo by Sol (@solyatic)
excuse me ‼️ hermes agent is free and open-source self improving AI agent ??? and can be run on $5 vps or any serverless infra that cost nearly nothing when idle!!! this gold thing is free 😭😭
1
2
101
I wonder what Einstein would have done with Fable 5.2 and no usage limits
1
4
64
Here's 5 prompts that would change your perspective of building software apps for the next 5 years: 🧵
2
2
80
4/ The Explainer "Explain this codebase to me like I'm a new dev joining tomorrow. Main flow, key files, and where I'd most likely break things." Onboarding in minutes, not days.
1
9
5/ The Minimalist "Solve this with the smallest possible change. No new dependencies, no refactors, no extra files. Show only the diff." Kills AI over-engineering.
6
that's very generous to go outside after weeks of copying someone else's work, then rebrand it and sell it under your name.
went outside after weeks of not touching grass this is what happened: - tried scrolling table like a trackpad - found 3 bugs in my mind - my eyessss got cooked - forgot people exist in real life (they do) - it was kinda nice tho need to try this experience again
1
4
138
most visits on upplaunch.com comes the USA, kinda nerves
1
2
81
This is insane, let me tell you what just happened. I'm not the one saying it, Google is! Hackers are using LLMs to continuously scan software updates to recreate the bugs that have been fixed, thereby preparing cyberattacks before teams even have time to fix them! There are NO HUMANS in the process. AI vs AI, with us at the end of the chain as very likely victims. @FournesMaxime everything just keeps getting confirmed week after week.
Hackers are now using LLMs to continuously scan software updates, spot newly fixed bugs, and build attacks before teams can apply the fix — all without human involvement. That’s according to the latest report from the Google Threat Intelligence Group. We're helping defenders counter automated threats with Google AI Threat Defense, which combines Gemini and other frontier models to find and fix flaws before attackers can exploit them.
1
3
130
I took an IQ test online. At the end, I had to pay €6.99 to find out my score. I think that was the test.
1
4
99
PewDiePie. a YouTuber with no tech background, built his own private AI that runs fully offline on his own hardware: > built a 10-GPU home rig > quantized a giant LLM to run locally > built ChatOS, his own local AI UI > added RAG/local memory > built a "council" of AI models > built a "swarm" of small models running in parallel for data collection > fine-tuned a Qwen 32B-based coding model > donated compute from his GPU rig to protein folding research > open-sourced Odysseus, a self-hosted AI workspace Let this sync in.
5
173
SIGNS YOU ARE A SUCCESSFUL INDIE HACKER: 1. You don't care about MRR
6
4
241
5. You pay at least $10 to charity every week
1
28
6. If a product that solves your problem charges you $5/m, you build a clone instantly to protect yourself from sharing you google account with strangers.
1
28
I'm looking to sell upplaunch.com 👀 Status: → 5,027 unique visitors → 320 registered users → 189 unique products → 91 repeat-launching makers → 194 live listings Asking $2,500. Strong SEO foundation already in place. Reason for selling: Focusing on my other apps. DMs are open.
1
6
243