Vinit

Computing things | UWaterloo CS '21

[Vinit @fnetisma]

C4-

[@jason @Jason]

B41

[Bill Gurley @bgurley]

I disagree, I don't think this has to be "game-over" for Google. I do think they have only one play left now - to pull from their own Android/Kubernetes playbook and fully embrace open-models. It's the obvious move in this situation.

Details here:
https://t.co/P8rbKhIjzH

[Vinit @fnetisma]

@nickscamara_ @firecrawl Do I use this or @reductoai ?

[Vinit @fnetisma]

@tech_optimist Onwards and upwards! Can’t wait to see what you do next, you’re a great value add to any org you join (especially if it is AI adjacent ;) )

[Vinit @fnetisma]

One of these makes money in the ai era either above surface or below. Your choice

[Vinit @fnetisma]

@chamath @kingtrillackman What about Distribution as a bottleneck Chamath?

“Customer proximity” as your post implies seems like the limiting reagent

[Vinit @fnetisma]

@jasonkneen Yes dude of course. If you’re not a billionaire then niche down. Big fish small pond > small fish big pond

[Vinit @fnetisma]

@OpenAIDevs I wonder if there’s an App Store moment somewhere where we can actually extend the Codex app with our own vibe coded apps that we use on our computer. Does Codex emit or publish events which we can subscribe to with our applications?

[Vinit @fnetisma]

Everyone is worried about the ban of OS Chinese models but they don’t know @elonmusk and team have put Grok 4.5 right up there. How am I so confident?

We are running our fintech AI Agentic workflow demos live for $10B AUM cos and we literally hear “wow this is what we need”

[Parker Conrad @parkerconrad]

A set of politically-connected growth investors are mobilizing to protect their bag in Anthropic. They will secure bans on chinese open source models, when no crime has been proven--or even alleged--and without any act of congress. All to protect a set of wealthy investors.

[Vinit @fnetisma]

@MTSlive @DavidSacks can you help prevent this? We need market competition to collapse model prices

[Vinit @fnetisma]

@thsottiaux Can you make it easier to do multiple features? Better worktree management. Right now I have to manually manage backend and fronted worktrees

[Kimi.ai @Kimi_Moonshot]

Introducing Kimi K3: Open Frontier Intelligence

🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost
🔹 Built for long-horizon agentic coding and self-evolving workflows

Kimi K3 is now live on on https://t.co/zrk6zZxZUo, Kimi Work, Kimi Code, and the Kimi API.
Open Weights by July 27, 2026.

🔗 API: https://t.co/XCrgjXAqMw
🔗 Tech blog: https://t.co/YTfiMSNM1f

[Vinit @fnetisma]

@sama This is true but you should see if it’s possible to access the files uploaded, etc by the voice. it’s unusable if I have to talk about some context

[Vinit @fnetisma]

Kimi K3 is the new SOTA. I’m so excited. Need more downward price pressure from Grok 4.5 and co

[Vinit @fnetisma]

It’s back up :)

[Vinit @fnetisma]

Chatgpt is down :(

[Vinit @fnetisma]

Touching grass

[Vinit @fnetisma]

@elonmusk @SpaceXAI Thank you for having DeepSWE front and center. You know your audience

[Vinit @fnetisma]

I love this. The pricing is great, the models are great, I can’t wait to use these models in my harness

Idk if it was an Elon decision but thanks for having DeepSWE front and center. I knew within 10 seconds of reading the page these models are legit and no gaming is going on

[SpaceXAI @SpaceXAI]

Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.

https://t.co/i8HpU7w64k