jmtulloss

it appears that 3% is the threshold as 3.1% gets the nod the other way

It's because they had an AI say it instead of a human who's thinking deeply about what they're trying to say.

TBC I love AI and am totally bought in, but human taste and _severe_ editing is still required, just as it would be with any human.

Is your argument that $1500 / mo is too much? Why would the engineering team not be more rigorous in their model selection given a constraint?

Why do you think this? Confidential filings before an IPO are standard practice.

Exactly this. If there is some nuance in the article vs what Claude can tell you, then that's worthwhile. This article is just generated with a specific prompt on style but very little content editing. What's the point? It's like posting the results of a Google search. The prompt would have been more interesting.

It's not against the rules to post AI slop here, and I don't necessarily think it should be. But I do wonder how we value written content going forward. There's value to taste and style and editing and all the other human things... there's very little value in the actual words themselves. We'll figure it out.

The best stretch Github ever had was post-acquisition when Nat Friedman as CEO.

FWIW the whole group received permission to cross. The instructions were to "Truck 1 and company", not just Truck 1

The comments so far seem focused on taking a cheap shot, but as somebody working on using AI to help people with hard, long-term tasks, it's a valuable piece of writing.

- It's short and to the point

- It's actionable in the short term (make sure the tasks per session aren't too difficult) and useful for researchers in the long term

- It's informative on how these models work, informed by some of the best in the business

- It gives us a specific vector to look at, clearly defined ("coherence", or, more fun, "hot mess")

Yes, it’s not a hill I’m willing to die on. It’s a hill I’m willing to defend until the cause is lost.

This is what I'm saying. Chipmunks are not squirrels. I will do my best on this hill.

Obviously novel problems require novel solutions, but the vast majority of software solutions are remixes of existing methods. I don’t know your work so I may be wrong in this specific case, but there are a vanishingly small number of people pushing forward the envelope of human knowledge on a day-to-day basis.

The reason for this is Rivian and Tesla bet big on software defined platforms… ie every piece of hardware talks to a small number of central computers instead of many independent systems. This gives them a huge leg up in developing software than can actually take all the available input and use it to control all aspects of the vehicle.

Downside is all the buttons are on a screen. But I’ve grudgingly decided it’s worth it for software upgrades.

The current Gen 1s will start beeping at you if they can’t see the lines. If you don’t take over quickly it will start slowing down and beeping very insistently.

Not only is Rivian betting on an integrated platform being important for their own cars long term, they’ve also essentially sold that portion of their business to VW. They are investing in the software platform for a lot more cars than just the rivian branded ones.

Linear is a venture funded company

A lot of commenters are focusing on the legalities and likelihood of backpay, which is relevant but I tend to agree with you… it’ll get paid because it’s in the interest of both parties to pay their employees what they’re owed.

We’re staring down the barrel of two missed paychecks though. If you're living paycheck to paycheck you’re getting desperate. If you’re living with about 1 month of emergency buffer… that buffer is one paycheck away from gone. It’s a cash flow issue

And the republicans could just vote to change the rules of the senate.

The out of power party gets a little veto power here. The republicans know the day will come they want that, so they won’t change the rules even though they have the power to do so (theoretically… there are republicans that will never compromise on this). Unfortunately they can’t get on the same page with their lame duck leader

My interpretation of the parent comment was that they were loading specific curl calls into context so that Claude could properly exercise the endpoints after making changes.

It doesn’t have access to your repo when the agent is running (unless you give it internet access and credentials). The code is checked out into the sandbox before it’s let loose.

Good point. The environments I’ve set up have been pretty easy but I’ll admit that at first I was very annoyed that it couldn’t just use a pre-existing GitHub action workflow.

Edit: environment setup was also buggy when the product launched and still is from time to time. So, now that I have it set up I use it constantly, but they do need to make getting up and running a more delightful experience.

Why aren’t more folks using Codex cloud? Simon’s post mentions it, but the vast majority of comments are talking about parallel agents locally or getting distracted while agents are running.

Personally I’ve found that where AI agents aren’t up to the task, I better just write the code. For everything else, more parallelism is good. I can keep myself fully productive if many tasks are being worked on in parallel, and it’s very cheap to throw out the failures. Far preferable imo to watching an agent mess with my own machine.

I'm the same way but that makes the bulge even more annoying. They're designing it to put a case on it.

The thickness should be from the front to the back of the camera lens, not to the thinnest point they can find.

Seems really interesting! It would be great to see some evals to understand how well this works when stacked up against different approaches and use cases.

Both from the design scheme and the process it uses to go about its business, Jules seems very inspired by replit

https://blog.singleton.io/posts/2025-06-14-coding-agents-cro...

The former CTO of stripe, for one.

They show you the code they produce. Why wouldn’t you trust it after reading it?

Yes it does

Yes, but they’re over indexing on that imo. The smooth comfort they’re looking for is costing them multiples of how much time it will actually save

[deleted… my disagreements about time allocation aside, there’s no reason to argue on the internet about it]

It’s likely that the models don’t need to get much smarter, but rather the ux for providing needed context needs to improve drastically. This is a problem that we’ve worked on for decades with humans, but only single digit years for AI. It will get better and that tediousness will lessen (but isn’t “alignment” always tedious?)

This is not true at all.

Perhaps there are more qualifying statements that you meant to include? The certification and type rating requirements certainly differ between agencies, but in terms of raw number of flight hours it’s easy to find that this statement is false.