Beyang

Building @ampcode, founder @sourcegraph

[Beyang @beyang]

A new UI for driving Amp from your desktop or, as I often do now, from your phone:

[Quinn Slack @sqs]

We rebuilt Amp's UI so you can watch and drive all your Amp agents on web, mobile, and CLI.

https://t.co/nlNhe1hAA3

[Quinn Slack @sqs]

I love when technological progress kills a feature.

The models are so good that (A) this feature is too risky and (B) it's not needed anymore to convince people that agents are good.

https://t.co/PlPosB3wYu

[Beyang @beyang]

@bygregorr Yeah it’s like Sourcegraph was founded by the same people who like humans or something :P

A good proxy for “major industries” is how they segment the major publicly traded companies: https://t.co/Y9w4YKHSz2

[Beyang @beyang]

If you're a builder who's tired of the AI discourse being about job loss, class warfare, and doomerism, come work with us as we figure out how to deploy AI to amplify the agency of humans working in major industries.

[Amp @AmpCode]

Amp Labs: small teams of the best software builders, working inside the most frontier-oriented company per industry and region.

https://t.co/IAI3YBqBWf

[Beyang @beyang]

There are plenty of coding tasks where you need to be mindful of your mental model and actively steering the agent to implement a clean architecture. For those, you want something fast and interactive but still smart and competent.

What we learned from the previous rush mode is that cheap tokens do not necessarily mean cheap in terms of total cost or time spent. If you haven’t used rush in awhile, try it again!

[Nicolay Gerold @nicolaygerold]

We made rush actually rush.

It is now GPT 5.5 with no reasoning tuned for smaller, bounded coding tasks where higher reasoning is overkill.

It’s almost twice as fast and outperforms the previous Haiku-based rush on our internal evals.

[Beyang @beyang]

The broader theme I'm seeing is that current agent UX makes the low friction path "LGTM". Software engineering has always had a verification problem in code review, but the rate at which agents generate plausible-looking code raises this to an imminent threat to both the mental bandwidth of humans and the integrity of the codebase.

[Beyang @beyang]

@urban_lobster Enterprise

[Beyang @beyang]

As coding agent usage caps continue to tighten, Amp's pricing remains the same.

The "Porsche of coding agents" is increasingly near price parity with alternatives. Rather than switch tools every 3 months, Amp users can use one opinionated harness that's built for the frontier.

[Beyang @beyang]

@alexellisuk hmm, my video was actually taken within tmux. Which terminal are you using?

[Beyang @beyang]

We hope the attention to detail in the new Amp TUI will be emblematic of the software you build with it.

Make it work, make it fast, make it 𝒔𝒉𝒊𝒎𝒎𝒆𝒓

[Beyang @beyang]

@alexellisuk Shoot an email to amp-devs at ampcode dot com. We're onboarding users gradually as we scale up our infra!

[Beyang @beyang]

@davidgasquez An internal mode we have for testing new stuff. Same as deep mode now.

[Beyang @beyang]

Wait til you hear how this was built

[Tim Culverhouse @rockorager]

Markdown rendering in Amp is now 2x faster end-to-end, thanks to a new parser we (ahem...Amp) wrote today. This parser is 100x faster in most cases, and in large files 200x faster. Uses struct-of-arrays and buffers for low allocations and lazy string materialization.

[Beyang @beyang]

@moinulmoin @benvargas @nicolaygerold @sqs @AmpCode Email amp-devs at ampcode dot com :)

[Beyang @beyang]

@benvargas @thorstenball The blog post is correct (low/medium/xhigh). Just pushed an update to the manual, thanks for reporting!

[Beyang @beyang]

@erd0xbc Yeah, there’s a protocol that ensures state is synced and broadcast across different thread clients

[Beyang @beyang]

A completely rebuilt-from-the-ground Amp TUI is out today!

[Thorsten Ball @thorstenball]

The best coding agent in the world is back.

https://t.co/BU96nS5IcC

[Beyang @beyang]

@sauercrowd ty, looking forward to hearing your feedback!

[Beyang @beyang]

@drnafizhamid @AmpCode This was a regression that went out today and will be fixed on the next version update. Sorry about the confusion and thanks for flagging!

[Beyang @beyang]

New model for Amp’s smart mode

[Thorsten Ball @thorstenball]

Amp's smart mode now uses Opus 4.7.

It's very, very good.

And we could remove three tools.

https://t.co/AWeAqdGw2G