f16hari

Software Engineer

[f16hari @f16hari]

This resonates quite well with software we build as well, Its quite hard to make it follow your vision and dream than letting it completely free and wild....

https://t.co/vscZA205At

[f16hari @f16hari]

We all know reading model generated code is pointless now. But completely disconnecting from the code means you lose touch with your own project. Suddenly, I 've become the bottleneck because I'm clueless few times and less motivated to steer the model or not sure how to steer...

[f16hari @f16hari]

@litesam211 😅

[f16hari @f16hari]

Switching to the ChatGPT desktop made me realize how much being in a TUI all the time ruined my reading habits.I used to just skim model outputs because the UI was bad for reading.The desktop app is a total game-changer. Feeling more involved now! Lets see how long it stays true.

[sankalp @dejavucoder]

sometimes things feel boring because you are just not paying enough attention

[Arpit Bhayani @arpit_bhayani]

Our only unfair advantage left is wondering.

Not knowing. Not having the answer. Just being curious enough to ask, "What if?"

Machines are getting better at knowing. They are getting better at finding answers, connecting ideas, and even producing things we once needed years to learn.

Perhaps our advantage is not in knowing more, but in knowing what is worth wondering about.

[f16hari @f16hari]

Few conv. with Opus 5 just makes you hate your day, its just doesn't get basic things and drains you more than you would if you would have done it yourself. Spent a while trying to make it understand it was doing a wrong fill, so much for a model scoring 30+ in ARC AGI 3 🥲🥲

[f16hari @f16hari]

Didn't realize slowly we all became this guy instead of Barry Allen as we are no longer the speedsters but the guy and the chair with ideas, steering and guiding agents like he was doing for Barry 😅😅

[Lee Robinson @leerob]

You might believe you should spend less time thinking about code because of AI.

I strongly disagree! We’re watching this play out live where tons of AI generated code becomes a liability.

At the end of the day, an engineer needs to be responsible / on call for code that gets shipped to production. If you don’t understand the system you’re trying to debug, you’re probably going to have a bad time.

Yes, AI can help with all of this, if you set up the proper systems. You can have agents triage prod logs, look at errors, etc. You can speed up parts of the investigation, but an engineer needs to make the call. There might be serious customer or financial implications from that change.

I expect the trend continue for trimming dependencies, vendoring code so you can modify it directly, preferring simpler systems with fewer abstractions, and spending waaaay more time thinking about system design and code maintenance.

I’ve said this before, but it’s a great time to get familiar with CS fundamentals and some of the history behind what great software looks like. Many parts will be different in the coming years as AI progresses, but also a lot more than people realize will stay the same.

[f16hari @f16hari]

New ui update in cc @claudeai. Look cool 😎. Thanks team!! 🤣🤣

[f16hari @f16hari]

@TheAnkurTyagi True 💯

[f16hari @f16hari]

@Mugilan_SS @steipete @OpenAI Yeah 🙏

[sankalp @dejavucoder]

understanding things you care about leads to a sense of meaning - atleast for me. if you are not understanding and processing things example - you are vibe coding mindlessly for days at end, you will start feeling hollow. understanding provides grounding.

[dax @thdxr]

we've never done code review

but damn if your team is producing this much code you're using LLMs entirely incorrectly

no one struggles with large amounts of code more than an LLM, if you don't keep that in check you have a self defeating codebase

[f16hari @f16hari]

@grok rate my very first shot at soldering 😅. BTW flux pen is a must have thing I guess so less mess compared to the one that comes in tins

[Andrej Karpathy @karpathy]

New art project.
Train and inference GPT in 243 lines of pure, dependency-free Python. This is the *full* algorithmic content of what is needed. Everything else is just for efficiency. I cannot simplify this any further.
https://t.co/HmiRrQugnP

[OpenAI Developers @OpenAIDevs]

GPT-5.2 and GPT-5.2-Codex are now 40% faster.

We have optimized our inference stack for all API customers.

Same model. Same weights. Lower latency.

[f16hari @f16hari]

Didn’t know building as simple product as a smart weighing scale would be this fun 😁, guess hardware is really the next thing that can give the high that coding was giving sometime back 😅

[François Chollet @fchollet]

Because our universe follows stable laws, a sufficiently general intelligent system adapted to it, like human-driven science, can eventually model any phenomenon within it.

Human intelligence may not be "universal" in the mathematical sense (see No Free Lunch theorem), but we are perfectly adapted to decode the specific language of our universe. That's enough for us to be "generally intelligent".

[Dheemanth Reddy @Dheemanthredy]

We’re releasing Maya-1 in the coming days

It is the most natural and most conversational voice model we’ve ever built

Internally, we were genuinely surprised. The emotion, the consistency, the way it speaks. It just feels different.
We’re excited to share it with the community and bring voice intelligence to everyone

Fully open source with weights
More details soon