Pinned Tweet
Gemini 4 Argon made this, so I had to try it with Opus 5.5
Same idea: a periodic table synced to Tom Lehrer's "The Elements".
Here's what Opus built, sound on
Gemini 4 Argon built a 3D periodic table synced to Tom Lehrer’s song, “The Elements.”
Created in @Antigravity, it animates orbiting electrons, atomic data, and light spectra.
I still don't understand why Telegram hasn't added AI models or AI bots into its app.
It's one of the most widely used apps, but it missed the chance to be an app like Grok Bot.
Telegram bots and mini apps can accept only Stars (*Grams) for digital services.
Why? Apple and Google rules require Stars.
But users can buy Stars with Grams on Fragment.
VPN bots in authoritarian countries are exempt.
They do God’s work and can take any payment ☦️
19
GPT-6.1 Sol was my favorite, so 6.2 Sol is probably even better. Can't wait to test it 👀
Replying to @OpenAI
GPT-6.2 Astra—a frontier model built for your hardest projects. It sets new highs on evals across professional work and scientific research at less than a third of the cost per task of GPT‑6 Astra.
You’ll see the difference in things like slideshows, websites, and 3D designs.
18
Added a Cybertruck and a snowmobile to my winter-riding game, and it's way more fun to play now.
Also a good test for my new screen recorder app 😊
New game just dropped 🏂
Fresh snow, big mountains, and nothing but speed.
Built with Opus-5.5 and Argon. Play it and rate the result,
Link in the first reply.
1
2
48
Hey @grok, how many people are registered on X right now, and how many of them are verified?
1
2
68
The Claude Code team says Claude writes and reviews ~90% of their code. Here's how they check it 👇
1. Verifying means running the app end to end, not just passing tests
2. They keep a "verify" skill that boots and drives the app, and Claude runs it before every commit
3. End-to-end tests beat unit tests because they're hard to fool: the feature either works or it doesn't
4. If a check still fails after 3 fix attempts, Claude stops and calls a human
5. A change can never edit the rules that judge it (REVIEW.md is read from the main branch)
6. A 39-item reliability checklist, but Claude only checks the 3–10 items the diff actually touches
7. Flaky tests can't be retried or padded with sleeps. Fix the cause or delete the test and write down why
8. Every repeated miss becomes a rule, and the rules get cleaned up on a schedule
Biggest lesson for me: your job shifts from reading every diff to defining what "done" looks like.
Which one would you add to your own setup?
1
1
42
Full post from the Claude team: claude.dev/blog/how-the-clau…
16
25 years of progress, and the feature everyone wanted was the 2001 loading screen)
We’re bringing back the original XBOX boot-up animation for all XBOX Series X|S and XBOX One players, along with two additional XBOX 360 boot-up animations!
This update will be rolling out throughout the day and thank you for being part of 25 years of XBOX.
19
OBS did not always record my OS audio along with my game audio.
So I deleted OBS and built my own screen recorder. It has all the needed features.
2
2
36
Building new things is one of the biggest advantages of AI, you can build your own app when the existing ones don't solve your problem.
I build games, so I need to record only the game's audio.
I used OBS for that, but it started not always recording my system audio.
So I built my own screenshot and screen recorder tool. It's fast, and it has every feature I needed.
8
China is testing a new electric passenger plane. Hey @grok, how long can it fly on one charge, and could it ever be cheaper to run than a regular plane the same size?
1
3
36
My choice is to select all and buy 😅
PlayStation’s Black Friday Deals start Nov 16.
Discover what’s on offer: play.st/4zs3OxV
1
28
Always interesting to listen to @poteto. Taking some notes from this podcast.
At @SpaceXSI, Lauren Tan (@poteto) and Roman Ugarte (@romanugarte_) are on the core team that took Grok Bot from first line of code to launch in seven weeks.
It lets you spin up AI teammates with their own identities and logins, and they just keep working in the background. Their marketing team gave a bot its own ad budget. Lauren built an entire feature with her bot in one Slack thread without ever opening Cursor.
They joined us on the @LightconePod to talk about what it’s actually like working next to bots every day, and how the hard part is shifting from doing the work to figuring out what to do and why.
03:36 — Giving Bots Real Jobs
06:49 — One Bot or a Team of Specialists
10:39 — Sharing Bots Across the Company
14:46 — Identity and Permissions for AI Coworkers
18:37 — Bot-to-Bot Collaboration and Trust
22:35 — The Workflow Behind 4,000 PRs a Month
26:00 — Teaching Agents to Write Code You Trust
29:01 — Building a Personal Software Factory
33:44 — The Future of AI Coworkers
36:47 — Rethinking the AI-Native Company
40:13 — Teaching Agents to Check Their Own Work
43:56 — How Much Code Review Is Enough?
47:33 — Codebase Guardrails That Save Tokens
52:20 — When Product Judgment Becomes the Bottleneck
56:29 — From Finance to AI Engineering
1
31



















