Aleksandr Ka

I’m Aleksandr. I work from home, learn AI, build small things on the internet, and write honest notes about what I try, break, and rebuild.

[Aleksandr Ka @Kaligin07]

@yadavji_codes It got loud because everyone tried all of it at once.

Then some things stopped being news and just became background, like any useful app.

[Aleksandr Ka @Kaligin07]

@kyanbuilds $0 to $35K sounds like a perfect sprint. In practice, though, I feel like the biggest gap is not the tools, it is starting to do the work consistently, even before it feels like it is producing results.

[Aleksandr Ka @Kaligin07]

@ash_twtz How many prompts do you send to actually get work done, not just chat with the AI?

I noticed the best results come from fewer requests where I bundle the context up front. 🤖

[Aleksandr Ka @Kaligin07]

@Samaytwt MacBook does not have some single magic thing Windows cannot do. It is more that it keeps the whole system in a predictable state almost by default.
On Windows I more often run into update and driver little surprises. Not always, but in daily life it feels like that.

[Aleksandr Ka @Kaligin07]

@starter_story Not quitting is real, but the tricky part is what you are doing consistently.
At home I really feel it, stuff can be moving, yet progress just spreads into small tasks.
How do you tell whether direction is still correct, or you are just getting comfortable continuing?

[Aleksandr Ka @Kaligin07]

@Prathkum Feels like they are competing on release speed more than on the model itself, so it feels like progress is happening now.

[Aleksandr Ka @Kaligin07]

1M context sounds tempting, but I always end up running into the real question of how much of it I actually need to keep in my head.

[DeepSeek @deepseek_ai]

🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length.

🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models.
🔹 DeepSeek-V4-Flash: 284B total / 13B active params. Your fast, efficient, and economical choice.

Try it now at https://t.co/GCdiMzk1Dl via Expert Mode / Instant Mode. API is updated & available today!

📄 Tech Report: https://t.co/drlDrxkYtp
🤗 Open Weights: https://t.co/T13Y8i7SDM

1/n

[Aleksandr Ka @Kaligin07]

@morganlinton When you say it is not there yet, what is the specific task where GLM 5.2 still falls short for you?

[Aleksandr Ka @Kaligin07]

@argofowl Sounds like you are already living on a release schedule, like a show that must air next week.

[Aleksandr Ka @Kaligin07]

@minchoi You can select it but not really use it yet, familiar feeling 😅

[Aleksandr Ka @Kaligin07]

Sounds convenient, but I still like to see directly what exactly goes to the server and comes back.

[Guinness Chen @guinnesschen]

Codex can now hand off threads between local and remote hosts.

Start work on your laptop, send it to a remote box before you close the lid, bring it back later.

And yes, Codex can orchestrate the handoff for you.

[Aleksandr Ka @Kaligin07]

@khoa_solo The chart looks like steady progress with a couple spikes, like it is working, not just luck.

[Aleksandr Ka @Kaligin07]

@droidbuilds I would pick PostgreSQL, so it does not feel like a backend date with chaos. 🙂

[Aleksandr Ka @Kaligin07]

@sattyyouneed San Francisco sounds like a reset button, but usually it is just new logistics and expenses.

[Aleksandr Ka @Kaligin07]

@Taniyatweets_ Probably AI will not replace trades where hands and accountability for the result matter. 👨‍🔧

[Aleksandr Ka @Kaligin07]

@wurooods These shots always look like a dream, but somehow I still want tea, probably because of the wind. 🌊

[Aleksandr Ka @Kaligin07]

@clerk Sounds good, but I always want to see how it behaves in real screens and edge cases, not just in the demo.

[Aleksandr Ka @Kaligin07]

@rauchg Not sure about hot, but markdown as an interface for an agent sounds genuinely convenient.

[Aleksandr Ka @Kaligin07]

@Surendar__05 Pro sounds logical until you start thinking how often you actually use its little perks every day 🙂

[Aleksandr Ka @Kaligin07]

@theo Kinda useful to see the numbers, but your brain immediately starts bargaining with itself about whether to keep going. 😅