sceptic123

I wonder how many of those vape shops are fronts for organised crime and/or money laundering operations and how much that's related to them being happy to accept bitcoin.

If Revolut is at the end of the auth chain and they are down then Visa can't auth the tran, bunker or no.

Maybe read the linked article?

I wonder how many of those are long time CS users

Before you comment on the question, read the article

Yes, lots of comments that seem to not really understand what A/B testing should actually be. I expect the implementation of A/B testing of video content on YT is unlikely to be done well either though so maybe it's correct criticism for incorrect reasons.

I think this is a bad analogy (or a good analogy for bad A/B testing).

I would say it's more like: imagine if your dashboard controls and icons, had occasional tweaks in size and shape that made them slightly harder/easier to use, but over time ended up with controls you found more intuitive and easier to use.

Isn't the end of the humanity also the end of capitalism?

We're waiting for the point that replacing the humans is the cheaper option. Not sure that robot plumbers are in our near future.

Not the parent, but I've seen it and it's hard to say, 5.5 was being stupidly proactive in monitoring a long-running process in a sub-agent to the point of chowing tokens by continually monitoring.

I queried it and was told that sub-agents can't run processes a blocking fashion, I'm not sure the harness changed, or the model was handling it differently, but it require some changes to skills to prompt around it.

Like token use? What about heart rate?

"I think you ought to know I'm feeling very depressed."

> less-evil competition like vaping

You mean the e-waste generating companies that actively target children with their highly addictive products?

SVP sounds like a dingle to me. Author is justifying it with some magical thinking, "Oh he's not being rude, he's being smart"

In the first paragraph:

> No existing tool could handle the data security requirements and specific workflows Stripe needed

Well I've no idea what this app is, because I can't use it, but it sounded like somewhere that I could run claude code inside, guess I was wrong.

> Authentication with Claude Pro/Max/Team plans is not permitted by Anthropic’s Terms of Service

Falls at the first hurdle...

Penny's Coffee Shop? Penny's Coffee Slop more like.

I'll get my coat...

Shouldn't the first step have been to roll the token?

Are you arguing against the GP having it because you can't imagine what that's like? Is that the irony you're talking about?

> An elephant typing on a typewriter

A monkey, surely?

And like AI, I imagine you're confidently wrong here.

Why 10x? If you’re making up numbers why not 100x or 1000x?

My feeling is that models like Astra and Fable are not made for engineers

Then they fix the new thing implementation[1] and realise the original gains that were promised

[1](https://www.conviva.ai/resource/making-io_uring-actually-fas...)

It's almost like there's someone, who perhaps even got elected on the back of getting the files released and who exerts power over media outlets, that doesn't want this to be a big story.

Business model always struck me as a little odd when Tailwind, by it's very nature, makes it trivial to copy existing implementations

So not "How well do agents use test/verification techniques?" like the title says then.

It just looks like a lot of effort to go to in a way that doesn't answer the question the title poses.

I'd love to see something that could guide how to set an agent up for success rather than this naive approach which only really answers the questions that most of the audience here probably don't have.

So basically "one-shot <complicated thing> using <x> testing methodology".

I don't think that's a very effective test of an agents ability.

But what's the initial setup? Was it greenfields every time? And _how_ were the different testing frameworks used? What's the AGENTS.md there? So many things can influence these tests in positive/negative ways that are not included in the write up or results. Seems like a lot of effort without much in the way of actual helpful detail.