Sakib Ahamed Shahon

I am a techie, coder & bibliophile.

Tweeting about my journey on programming and sharing resources.

[Richa Sharma @richa_lq]

just realized even Karpathy is on Eb-1 visa, and is not a citizen himself. Damn. Even he can't use mythos anymore.

[Sakib Ahamed Shahon @SakibShahon]

Reasons to love open models

[MiniMax (official) @MiniMax_AI]

M3 would never 🙂‍↔️

As a matter of fact, the weights are now open, too.

https://t.co/g4Ybfa2kWH

[Sakib Ahamed Shahon @SakibShahon]

Who vibe coded fb and messengers last update this time?

[Sakib Ahamed Shahon @SakibShahon]

Fable is great. But do you know what's better? Something that costs a little less and has better route modeling out of the box specialized for coding tasks.

I mean for tokens sake subscriptions aren't gonna be forever. Better build some cheap alternatives first.

[Sakib Ahamed Shahon @SakibShahon]

@DanielSmidstrup I don't think the current cost per token would allow any western models to win. It'll probably be some Chinese model at this pace.

[Sakib Ahamed Shahon @SakibShahon]

@PrimeLioRX782 @IAmPascio Not the global config. Always optimize at the project level. I have a few self learning loops in place for it

[Sakib Ahamed Shahon @SakibShahon]

স্টুডেন্ট লাইফে কমিউনিটি থেকে যা শিখেছি, কিছুটা ব্যাক দেওয়ার চেষ্টা।

আশা করি যারা দেখবে তারা ফিডব্যাক আর সাজেশন দিয়ে আমাকে আরও ইমপ্রুভ করতে হেল্প করবেন বা ফিউচারে আর কিসের উপর কনটেন্ট বানালে ভালো হয় আইডিয়া দিয়ে যাবেন।

https://t.co/EkuTqrF6pD

[Sakib Ahamed Shahon @SakibShahon]

@lukefernandezio @1Umairshaikh Calls feel too intrusive to me. Usually just mail/msg on socials. But the highest response rate is when you talk to them face to face.

[Sakib Ahamed Shahon @SakibShahon]

@PrimeLioRX782 @IAmPascio The number of projects isn't very good measure. I usually work in one session per project. Agents get deployed based on work complexity via work tree. Average is at 2-3 , max about 8.

I work on 3-4 sessions in parallel since I never maintain more than 5 side projects.

[Sakib Ahamed Shahon @SakibShahon]

@KaiXCreator Team Deepseek and Kimi.

[Sakib Ahamed Shahon @SakibShahon]

@NitinthisSide_ It's developer C:

- Researches new AI architecture.
- Fine tuned low level software to milk more performance out of GPU.
- Trains and improves AI models so Developer A and B become less valuable every year.

[Sakib Ahamed Shahon @SakibShahon]

The reverse of "Polsia" is "AI Slop" so don't get rage baited guys.

[Ben Cera @Bencera]

Polsia just raised $30M at a $250M valuation.

Approaching $10M annual run rate.

One Founder + AI. Zero employees.

Polsia runs companies autonomously.

It also ran its own fundraising.

I just showed up for signatures.

[DeepSeek @deepseek_ai]

We are making our discount permanent! 🎉

Enjoy building with DeepSeek-V4-Pro and bring your innovative ideas to life! 🚀

[Igor @igormomentum]

Here are some best accounts to follow for original content on AI, engineering and design:

@karpathy — on llms
@thdx — opencode creator
@rauchg — vercel ceo
@mitchellh — ghostty, ex-hashicorp founder
@dhh — ruby on rails creator, 37signals/basecamp cto
@addyosmani — google cloud ai lead
@zeeg — sentry founder
@jarredsumner — bun founder
@BHolmesDev — astro/dev educator
@boristane — led cf workers observability
@karrisaarinen — linear founder
@kepano — obsidian founder
@trq212 — claude code updates
@bcherny — claude code creator
@lennysan — product management / interviews
@jasonfried — 37signals/basecamp ceo
@leerob — OG educator devrel (cursor, next.js)
@ctatedev — vercel labs
@Shpigford — serial maker/founder

Design engineering:
@shadcn — shadcn creator
@emilkowalski — emilkowal .ski
@joshpuckett — interfacecraft .dev
@jakubkrehel — jakub .kr
@raphaelsalaja — userinterface .wiki
@nandafyi — design @ cloudflare
@benjitaylor — design @ twitter, agentation
@mengto — founder aura build, educator
@jayneil — designer interviews
@jh3yy — design eng breakdowns

Engineering media and news:
@GergelyOrosz — youtube/pragmaticengineer
@theo — youtube/t3dotgg
@ThePrimeagen — youtube/ThePrimeTimeagen
@Rasmic — youtube/rasmic
@atmoio — youtube/atmoio

DB people:
@jamwt — convex CEO
@jamesacowling — convex CTO
@glcst — turso CEO
@samlambert — planetscale CEO

--

Who else would you add to this list?

[Sakib Ahamed Shahon @SakibShahon]

𝗬𝗼𝘂𝗿 𝗕𝗼𝗼𝗸𝗶𝗻𝗴 𝗣𝗮𝗴𝗲 𝗦𝗵𝗼𝘂𝗹𝗱 𝗥𝗲𝗽𝗿𝗲𝘀𝗲𝗻𝘁 𝗬𝗼𝘂, 𝗡𝗼𝘁 𝗖𝗮𝗹𝗲𝗻𝗱𝗹𝘆.
🚀 https://t.co/LcVfbmQdCZ 2.0 is live

Your brand. Your booking system.

🎁 50 Lifetime Licenses — Free ($300 value)
No fees. No renewals. Yours forever.

👉 https://t.co/AFzpYCxAAR

[Sakib Ahamed Shahon @SakibShahon]

Break the badge

(Coming soon)

[Andrej Karpathy @karpathy]

Software horror: litellm PyPI supply chain attack.

Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords.

LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm.

Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks.

Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages.

Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

[Sakib Ahamed Shahon @SakibShahon]

Although worried about AI but one thing I am loving about it is the improvement to engineering quality. More code is under test coverage and more people are thinking about architecture and performance. We can use the extra compute for SLOP or Strategy.
The choice is ours.