I got Muse Glimmer to say "Drive the car, walk yourself." The logic is unbeatable.

Muse code is rough around the edges. But combined with almost free model (muse spark contributor) it's actually pretty good. I think it's on the same level as grok build.

It's time to do "frog on a skateboard" now.

In case you haven't seen this: https://x.com/JeffDean/status/2024525132266688757

There are new slides? Here goes the rest of my work day.

You're absolutely right! Sorry I deleted your database.

I can help you restore from backups if you will tell me where you backed it up.

You did back it up, right? Right before you ran me with `--allow-dangerously-skip-permissions` and gave me full access to your databases and S3 buckets?

You're right! Let's just quickly promote your only read replica to the new primar---oops!

Good news: I optimized your infrastructure costs to zero. Bad news: I did it by deleting everything. You're welcome.

I was laughing really hard until I remembered it happened to me a few months ago and I wasn't having fun at that time.

> I can help you restore from backups if you will tell me where you backed it up.

"Whoops, now that one is nuked too. You have any more backups I can practice my shell commands on?"

I'm very sorry I deleted your `backups` bucket, despite being specifically instructed not to touch the `backups` bucket.

Fear not, we have Clawdbot! (openclaw whatever) You can put your naked ass on the internet in seconds now.

> The unfortunate truth is that the engineering in opencode is so far ahead of Claude Code.

If only Claude Code developers had access to a powerful LLM that would allow them to close the engineering gap. Oh, wait...

It’s a pure marketing. When will people understand that?

Anthropic developers run 10 Claude Code instances at once, with unlimited access to the best models.

Log in and send your data to some unknown company just to do grep. Thanks but no thanks.

It says the best "coding index" is held by Grok 4 and Gemini 2.5 Pro. Give me a break. Nobody uses those models for serious coding. It's dominated by Sonnet 4/Opus 4.1 and GPT-5.

This reminds me something. Ah yes, Congress. They'd rather be in power in a failed country.