Elisha Silas

The significance of the Internet today is hard to exaggerate 🗿

When the mountain can't come to you, go to the mountain. Nicely done @JeremyKasirye Keep walking.

Inside Diplomacy turns 1. Our team hosted a public dialogue with @EUinUG ambassador @JanSadek at @CavendishUg. He passionately highlighted youth inclusion as actors in foreign policy , Uganda - EU relations and shared insights about how to get into Diplomacy. #EUandUganda50

@JeremyKasirye @CenterKampala @CPSS_Ug @USEmbassyUganda @UDalumni @ugyouthsenate @EAYNOfficial @UgandaMFA @diplomatical05 @eusufrancis @EACYParliament 🦁🦁🦁

Embracing The Blue Hen Spirit . A journey it’s been and lots of new friends along the way . “Welcome to Foreign Policy” a phrase that redefined my views and experience in this field of International Relations .Much Gratitude @CenterKampala @CPSS_Ug @USEmbassyUganda @UDalumni

@andrewbrown Here for the roll call

100K views in three days. nice nice.

@andrewbrown Thank you

My office is so cold I am making this Claude Architect course with mitts on. 🥶

@georgediano For parked vehicles

Advice from a qualified mechanic: When the floods are over, drain your engine oil and replace it before starting the engine.

@DanKulkov were you anticipating to print money from it, and if so what was the anticipated ROI.

token money i spent to set up openclaw: $100 money i earned with openclaw so far: $0

@emil_priver a problem identified and solved, but in all still human skill and input is inevitable

Talked to a friend who writes code for economics. Their leadership really wanted the company to use LLMs for calculations, so they tried everything from those tiny, cheap open-source models to the expensive ones like Opus and GPT 5.2. They got total garbage results, both in terms of the explanations and the actual numbers being wrong. So, they went back to building the logic in code and just built a dedicated ML model for the tiny part of the algorithm that actually needed AI. It’s way more efficient, it’s cheaper, and it gets a much better result. The gap between what people expect from LLMs and what they can actually do is just massive.

@mischavdburg True faith in the goal 🙏🏾

Numbing yourself with chemical substances does not bring you closer to God or reality. The goal is to get as close to reality as possible and to experience it as clearly as possible. Any chemical substance introduces levels of delusion that prevent you from doing this.

@tomhacks what could be the reason behind that?

Have you noticed with all that AI you are slacking off more often while still being productive?

@milan_milanovic and that's when friends turn to be enemies

Most change is unwelcome, even if it leads to improvement

@kaaaash____ Left, but it would have been better if you were to pair it with a keyboard.

Be honest Left or Right ?

@madalinabalazs tend to find it hard in creating tangible mockups with these AI tools, the output seems to be too generic probably it could be a skill issue 😞

@Sayyidalijufri @dodopayments Let's connect if you don't mind

@esilas_ @dodopayments thanks Elisha

My little SaaS made $239 in 7 days 🥹

@neetcode1 Awesome 😎

NeetCode now supports Rust i better see some usage

@cetorres holy grail

Meet 🦐 PicoClaw: The ultra-efficient AI assistant built in Go. 🚀 Why switch from OpenClaw? • 99% less memory (<10MB RAM) • 400x faster startup (<1s boot) • Runs on $10 hardware (LicheeRV Nano) ✨ Key Features: • Single self-contained binary • Multi-platform (RISC-V, ARM, x86) • Built-in Web Search & Cron • Sandboxed security 🐹 The Go Advantage: • Native performance with minimal footprint • Rock-solid concurrency for multi-agent tasks Repo: github.com/sipeed/picoclaw Site: picoclaw.io #PicoClaw #AI #Golang #OpenSource

@steipete you got this 🙌🏾

omw to London for an english language test Hope I’ll make it!! 🇬🇧

@_trish_xD round trips querying

most performance problems are database problems. not language problems