Yousaf 🇵🇸

@WaqasQadar 🙌

Sialkot is such a model district in Pakistan. GDP per capita of city with one million inhabitants 18000 $ GDP per capita of entire district 4000 $ with 4.5 million inhabitants Literacy rate 80%

@jahooma How fast it though?

DeepSeek v4 **Flash** is absolutely insane. It costs almost nothing (~1/300th Opus), and yet performs among the best open source models. On our coding benchmark Flash does better(!) than Pro

@MDUmairKh How do you see the commoners benefit in the future Btw?

Pakistan records its lowest fiscal deficit in 27 years. However, the middle class has been crushed due to petroleum levies, high inflation and taxes. Either way, overall, this is good news for the economy, even though the common person has yet to benefit.

@wkoszek @richiemcilroy Doesn't seem that beefy. Which one are you talking about specifically?

@richiemcilroy Use your Macbook as dumb terminal. Get yourself beefy server at OVH for $20/mo and just SSH there.

a couple of months ago, I upgraded from a M1 Pro 32gb RAM, to a M4 Max 128gb RAM and honestly, it's made hardly any difference Rust builds still take a ridiculously long time battery dies within an hour or so with a couple of agents running and I'm still not running any local AI models

@productdevbook Great product dude 🙌 will buy the license at the end of the month

Your Mac knows which app is stealing your bandwidth. It just doesn’t tell you. ova does.

How did christians go from "hijab is oppressive" to "We did it FIRST" lol

@productdevbook Tf 😒

It reached 50% within 24 hours. No. 😪

@productdevbook ✌️🫠

@yousaf_sbr No, unfortunately. It's quiet all around. 😑

🍞A brand-new UX that makes finding a job and connecting with companies simpler and faster is coming soon. ( %90 )

@sarahmsachs Where does Kimi K2.6 stand wrt Opus 4.7?

DeepSeek v4 works fine, but it’s not the frontier-pressing moment we saw with Kimi 2.6. On Notion eval data, it’s similar performance to GPT 5.2, with understandable failings. Most interesting — it doesn’t scale well. It’s ridiculously slow. On multiple major, trusted, and performant US inference providers we see it 15x slower than GPT 5.2 and 2x slower than Opus 4.7, a problem Kimi never had. Curious if it’s a fundamental issue in architecture, or a matter of time til inference providers make it work. Doesn’t seem urgent either way, if Kimi can outperform. Cheaper maybe, but not groundbreaking.

@thewebtailor @diegohaz What's the best way to use them both? I personally plan & audit with GPT, & implement it via Opus, what's your setup?

@diegohaz Opus seems to be doing overly detailed planning, but it's missing incredibly simple things. ChatGPT is hasty, but much sharper in coding. Unfortunately, it's impossible to move forward without using both in parallel. we all’re basically sat down and made the two models paranoid.

I gave the same task to GPT-5.5 xhigh and Opus 4.7 max. – GPT took ~30 minutes. – Opus took ~2 hours. Then I asked them to review each other's work and give an honest verdict on which was better. – GPT said its own code was better. – Opus said there was no obvious winner. Then I asked them to learn from each other and apply the best parts they had learned to their own code. In the end, I asked them to review each other again and give an honest verdict after they had both improved. – GPT kept saying its own code was better. – Opus said GPT's code was better. What a journey.

@v_language This means two things, Either you're very good at the support or nobody uses V Lang.

V now has <50 open bugs! 10k+ bugs have been fixed. (We don't have a stale issue bot, each of those bugs has actually been fixed.) We're out of bugs, please report more :)

@Pakistanomy $ exchange factorisation where boy?

Oil Price above $100 vs 🇵🇰 Petrol Price 2008: Approx $147 🇵🇰 Petrol Price: Rs76 2011: Approx $126 🇵🇰 Petrol: Rs 100 2013: Approx $118 🇵🇰 Petrol: Rs 110 2014: Approx $115 🇵🇰 Petrol: Rs 111 2022: Approx $123 🇵🇰 Petrol: Rs150-Rs 235 2026: Approx $100–110 🇵🇰 Petrol: Rs 399.86

@theo Believe in myself or humanity ahh question

If >50% of people press the blue button, everyone survives Red button pressers always survive, but they’ll get a “red button presser” badge on their Twitter profile. What do you press?

@saltyAom Wow!

How is Elysia 2 this fast It's the "zero-copy" approach It's very easy to allocate a new array/object, eg. { ...a, ...b } or [...a] or a.concat(b) We basically reuse things as much as possible For example, when user provides a primitive validator with an empty option (eg. t.String()), instead of creating a new object, we reuse an existing one instead, marking it as "reference" Another example is when merging local and global lifecycle, merging multiple objects is an easy way out, but this means if you have a new 100k objects, user-provided and existing blueprint is there doing nothing after these new objects are created Instead, we use the object that the user provided as a base to mutate later, then store the snapshot of what is needed to process as a "reference", then "link" all objects and arrays when they are actually needed This means we basically reuse the input instead of creating a new one But this is hard to do in JavaScript because there's no explicit reference or pointer If you accidentally change the reference value, you will blow something up instantly and have no idea how or where because there's no explicit "reference", no indicator at all If you didn't properly design with this approach at the very start and constantly have this approach in mind, it's not going to work This means we have to rewrite the entire thing from the ground up, and it's extremely hard This is why I haven't been very active since the new year, I was busy playing around with this just to make sure it would work at this scale

@theo Exactly, that's why I use both, one for doing stuff(usually Opus) & the other for validation & audit of the stuff done by the other (usually GPT 5.4) & its working great. GPT surfaces real gaps, Opus Acknowledges them, proposes fixes, I hand them to GPT & proceed when both agree

It's so hard to describe the vibe difference between Opus 4.7 and GPT 5.5 (for coding) GPT is smarter and can unblock you, but it gets stuck in stupid ways and strangles itself with context sometimes. Opus will go down the most insane paths and refuse to acknowledge obvious answers, but it understands intent better and has more taste. Whenever I use one for more than an hour, I always reach to the other to "clean up". Best part? All of this changes every few weeks 🙃

Can you guess the people?

20 pe to humara AC chalta hai

@valigo Just escape it dude.

If you ever accidentally create a file named "~" in a local directory, PLEASE, do yourself a favor and delete it through a gui file manager. DO NOT try to be a hero. Don't ask me how I arrived to this conclusion

@shadcn Yeah, me too. Almost every time.

I see. How many of you do this? I’d like to get a sense of how common this is.