Angus Bezzina

professional amateur

[Angus Bezzina @angus1192]

Neat way to find the best deal on a GPU rental

[GPU CLI @gpucli]

Searching for a faster way to find the best deal to rent that GPU you need?

https://t.co/8jdIIeP2Rl just shipped 🚢

Now you can now compare prices from @runpod, @vast_ai, @ThunderCompute & more in one place.

[Thariq @trq212]

been asking others at Anthropic how they stay in the loop with Claude and fully understand the work being done

this is one of my favorites from Suzanne:

[Chamath Palihapitiya @chamath]

Yup.

[James Lal @lightsofapollo2]

envious of the m5 for one reason: it fixes the exact thing my m4 can't.

the m4 gpu has no real matmul units, so prefill is compute-bound and long context crawls. the m5 put neural accelerators in every gpu core, and apple clocks ~4x faster time-to-first-token vs m4 (qwen3-14b).

token generation is only ~1.2x faster, because that part is bandwidth-bound, not compute. so the m5's real win isn't faster tokens, it's that prefill and long context stop being the wall.

[Angus Bezzina @angus1192]

This is a great solution to take the tedium out of training models

[James Lal @lightsofapollo2]

I train LoRAs almost every day now, for our other products. The jobs run for hours.

I haven't run one of those by hand in weeks. I drive them through Claude or Codex. The agent launches the run, watches it, and pings me when something needs attention.

https://t.co/1LjItdwz2o

[Guillermo Rauch @rauchg]

Claude Managed Agents 🤝 Vercel Sandbox
https://t.co/pW6iGH63MG

[Guillermo Rauch @rauchg]

You can just render images on the terminal btw:
▲ ~/ npx ai-cli image 'a vercel ai sdk diagram'

Run 𝚗𝚙𝚖 𝚒 -𝚐 𝚊𝚒-𝚌𝚕𝚒 and access every image, video & text model from @vercel AI Gateway instantly

[Boris Cherny @bcherny]

The best way to level up from 1 agent => many agents. No more cycling between terminal tabs

[Thariq @trq212]

HTML is the new markdown.

I've stopped writing markdown files for almost everything and switched to using Claude Code to generate HTML for me. This is why.

[Justin Skycak @justinskycak]

Never underestimate how much time and effort you can waste by trying to automate a process you do not understand manually.

[Angus Bezzina @angus1192]

https://t.co/KDkum6KTl6

[Angus Bezzina @angus1192]

Amphetamine is the best tool I've found for preventing agentic sessions from ending early on a macbook

[Naval @naval]

New podcast on vibe coding - A Return to Code.

A Return to Coding 00:20

The Personal App Store 03:17

Vibe Coding Is a Video Game with Real-World Rewards 06:22

Pure Software Is Uninvestable 10:33

A Place for Each Model 14:22

AI Is Eager to Please 17:57

Why Math and Coding? 22:10

The Beginning of the End of Apple’s Dominance 24:17

Coding Agents As Customer Service Reps 27:55

[Brendan Falk @BrendanFalk]

To check if your Google Workspace has been compromised by the same tool that compromised Vercel:

1. Go to https://t.co/TpuIOW5Fwg
- This is Google Admin Console > Security > Access and Data Control > API Controls > Manage app access > Accessed Apps
2. Filter by ID = https://t.co/uqJnCqp5Ah
- This is the ID of the compromised OAuth app

If you see an app after filtering, you have potentially been compromised

[Vercel @vercel]

We’ve identified a security incident that involved unauthorized access to certain internal Vercel systems, impacting a limited subset of customers. Please see our security bulletin:
https://t.co/0S939n3qHC

[Angus Bezzina @angus1192]

This article by @MattPaige32 might be the cleanest article/walkthrough I've found for @karpathy's LLM Wiki or "Second Brain".

Really enjoying working with the setup so far

https://t.co/xGzPZRh4bG

[Angus Bezzina @angus1192]

Now you're up and running!
For what it's worth, I recommend adding an alias for the script above to make life easier.

[Angus Bezzina @angus1192]

Playing with #TurboQuant KV cache compression is easier than you think. Here's how to get started with an Apple Silicon mac in 30mins or less

[Angus Bezzina @angus1192]

Finally, run the model