ryanisnan

I wish you great luck! I want this to succeed.

Why does the system work like that? Is the cache local, or on Claude's servers?

Why not store the prompt cache to disk when it goes cold for a certain period of time, and then when a long-lived, cold conversation gets re-initiated, you can re-hydrate the cache from disk. Purge the cached prompts from disk after X days of inactivity, and tell users they cannot resume conversations over X days without burning budget.

This is also an oversimplification. If I understand the issue correctly, the notification with the message contents was what was cashed locally and then accessed. This same vulnerability would exist with Signal if you had the notifications configured to display the full message contents. In this case, it has nothing to do with either Apple or Signal.

I think Martin isn't wrong here, but I've first hand seen AI produce "lazy" code, where the answer was actually more code.

A concrete example, I had a set of python models that defined a database schema for a given set of logical concepts.

I added a new logical concept to the system, very analogous to the existing logical set. Claude decided that it should just re-use the existing model set, which worked in theory, but caused the consumers to have to do all sorts of gymnastics to do type inference at runtime. It "worked", but it was definitely the wrong layer of abstraction.

Anecdotally, I've always been reasonably pleased with their products. I think I've owned a couple of powerbanks, and a USB/HDMI hub. Of the <Insert_random_smattering_of_letters> brand names on Amazon, I do tend to lean towards them a bit more.

edit: having said all of that, relating to this article, I don't want AI anywhere near the products of theirs I'm currently buying.

I think it's a bit of a failure state if we're all creating our own orchestration layers, to be honest.

I've been digging into some of the history of modern cryptography, as I seek to better explain how an app of mine works under the hood. It has been fascinating. I didn't expect AES-256-GCM's history to involve a duel where someone literally died. I'm hoping to write a few part series digging into more specifically, the history of the cryptography space.

Convenience is our Achilles heel, as a society.

We'll keep dangerous devices like the SuperBox in our homes, if it helps us get access to free movies and tv.

We'll use single-use plastics, even if we know they're bad for the environment, because they're just so damn easy.

We'll let AI run that thing for us, because it's just too easy.

A whole generation has grown up without knowing what it was like to infect your computer with AIDS trying to download an MP3, and it shows. That caution will come back, just at a terrible cost.

Wow, thank you for sharing. If I could upvote your comment twice, I would. I'm going to enjoy this series.

This is a great idea, but respectfully, if you're going to get traction you need to be the one instigating getting people to talk to you. Have a pitch, have an explicit ask, and be willing to put effort into making it happen.

Fantastic idea though, you should do it.

Easy... $<INSERT_VERY_LARGE_RANDOM_NUMBER_HERE>

Wow, thanks for that lesson! I owe that guy my career.

Dang this is so good. Well done.

I think they mean developers have no way of requesting or specifying order.

Are you serious? When you lower the cost of killing, nobody wins.

For reals. I love the general premise behind the article, but to me how you publish it, and how others access it, is the sauce. Creating static sites is hardly the problem.

I'm not totally clear, but it probably involves the ol' stranger.

Great question. People answered already, but, yeah, basically what they said.

For hobby sites, you could argue (I think the argument is still weak), about the MITM threat being low enough to not be worth doing something, but this is a security blog.

Security blogs matter, because people follow their guidance. This makes them a potentially attractive target for some groups of people.

Quality absolutely matters, but it's hyper context dependent.

Not everything needs to, or should have the same quality standards applied to them. For the purposes of the Cursor post, it doesn't bother me that most of the commits produced failed builds. I assume, from their post, that at some points, it was capable of building, and rendering the pages shown in the video on the post. That alone, is the thing that I think is interesting.

Would I use this browser? Absolutely not. Do I trust the code? Not a chance in hell. Is that the point? No.

My apologies - my point(s) were more about the original submission for the Cursor blog post, not your post itself.

I did read your post, and agree with what you're saying. It would be great if they pushed the agents to favour reliability or reproducibility, instead of just marching forwards.

I think for the point of the article, it appeared to, at some point, render homepages for select well known sites. I certainly did not expect this to be a serious browser, with any reliability or legs. I don’t think that is dishonest.

The amount of negativity in the original post was astounding.

People were making all sorts of statements like: - “I cloned it and there were loads of compiler warnings” - “the commit build success rate was a joke” - “it used 3rd party libs” - “it is AI slop”

What they all seem to be just glossing over is how the project unfolded: without human intervention, using computers, in an exceptionally accelerated time frame, working 24hr/day.

If you are hung up on commit build quality, or code quality, you are completely missing the point, and I fear for your job prospects. These things will get better; they will get safer as the workflows get tuned; they will scale well beyond any of us.

Don’t look at where the tech is. Look where it’s going.

This looks awesome - my son loves gears, and my wife and I have been talking about buying him a 3D printer soon. Thank you!

You are a great writer - thanks for putting this together!

I think the average human would do a far worse job at predicting what the HN homepage will look like in 10 years.

That one got me as well - some pretty wild stuff about prompting the compiler, starship on the moon, and then there's SQLite 4.0

thanks! love the app, it's really fun, and surprisingly engaging, despite knowing that it's all AI nonsense

Needs a dang archetype, who merges similar posts.

What an awesome piece of technology. I've been wanting to create something similar, just on the technical merits. We have some pretty amazingly capable technology these days, but so much of it relies on IP infrastructure, which is fine when things work and you are either aligned with your government, or live in a society where there are strong checks and balances on government overreach.

I think you're misreading the situation. As far as I can tell, Russia has every reason to want to continue engaging in heavy cyber-criminal activities. I don't think this is the virtuous Kremlin turning a blind eye. This is a classic case of deception. Look at my left hand, so you don't see what my right is doing.