Waves

Programmer * Physics

[Waves @BonsaiTree92]

Launching MultiplexTerm for IOS.

Control your tmux sessions from your phone.
For free, fully open sourced.

Working on getting it app store approved. No Charge for users.

https://t.co/14RAXD9JlD

[Waves @BonsaiTree92]

How are you measuring issue triage success rate?

Its very common for teams to say “agent easily triaged a b c!!!” Based off a few cherry picked examples But never really scales once you are triaging 100s per day and it’s pretty easy to loose customer trust when your agent is wrong

[Waves @BonsaiTree92]

@neetcode1 It’s definitely possible and will most likely take the form of signal based processing

[Waves @BonsaiTree92]

@mattparlmer What does goof mean? Is there a common pattern? i.e over allocations, sub optimal computations?

[Waves @BonsaiTree92]

Coding has been largely solved in terms of Syntax. Automated Debugging/research is what makes a model a frontier model

[Tak 🦞 @cherry_mx_reds]

Grok is about to become really good for coding.

[Waves @BonsaiTree92]

@cherry_mx_reds I don’t need a good coding model. I can code myself.

What I need is a powerful research model

[Waves @BonsaiTree92]

People would be shocked to find out that the vast majorities of Fortune 500 companies are already running Chinese models internally due to being available under Apache 2.0

[Gail Weiner @gailcweiner]

The US government bans Fable 5 and Mythos 5 worldwide because frontier models are too dangerous to let foreigners touch, won’t even exempt the UK, because the threat model says the weights themselves are a national-security asset.

And simultaneously the most deeply embedded US enterprise software company on earth quietly fine-tunes a Chinese model and prepares to ship it inside the productivity stack of every Fortune 500 that runs Microsoft 365.

[Waves @BonsaiTree92]

@liminal_warmth It’s even more ironic when you realize that anyone who would use it for actual deep work couldn’t due to either a.) blocked for corporate job due to data collection policies or b.) useless for deep research due to aggressive guardrails

[Waves @BonsaiTree92]

@KeyTryer Its about to be a civil war at xAI for computing resources

[Waves @BonsaiTree92]

@mitchellh Hashicorp implementing its own solution as well?

[Waves @BonsaiTree92]

@deliprao I suspect the opposite will probably be true.

[Waves @BonsaiTree92]

@bendee983 10x more self initiative.

[Waves @BonsaiTree92]

Even if Fable got unbanned tomorrow, it’s still completely unusable for most of the corporate workflow given there’s no way to opt out of the data collection policy

[Waves @BonsaiTree92]

@rizzn It’s a lot worse when you see it happen in the company slack

[Waves @BonsaiTree92]

@DanielMiessler Had they glazed the current administration there’s a high chance nothing would have been enforced

[Waves @BonsaiTree92]

@gorkem Now they sell agent harnesses

[Waves @BonsaiTree92]

@cqkten Anthropic understand the stakes and are willing to take the risk when the ball is in their court.

They can easily leave the US and the government knows that.

[Waves @BonsaiTree92]

@BhosalePratim SEO with extra steps

[Waves @BonsaiTree92]

Don’t let the model be your Wayfinder

[Conor Dart @Conor_D_Dart]

When Opus 4.8 first came out, I thought it was a really good model.

Today though, it feels like it's missing basic instructions and doing things that just don't make sense.

I don't know if I've somehow been hit by the Fable 5 bug, or if something else is going on, but the experience feels noticeably different from when it launched.

With Fable 5 currently unavailable, I'd have expected Anthropic to have plenty of compute available for Opus.

[Waves @BonsaiTree92]

Learning how to iterate fast with LLMs will become the most valuable skill to have for the late 2020s

[Waves @BonsaiTree92]

@DevBySami The UX is cool, but what problem is this actually solving? Like I can’t imagine a time where I want an ai assistant chat experience in my browser.

A visual proactive ai assistant sure but chat based??