> "Why did it pass reviews? Well, because of deadlines. And because there is simply so much code (and so much unparsable/misleading documentation) that it's simply impossible to review all of this. And because things move so fast that nobody understands the CI pipeline anymore, and the explanations of the agent are convincing enough that surely, it knows better than you?"
I have come to realize that AI is so "successful" because the system in which it is being deployed was designed to push product as fast and cheaply as possible from the start.
Humans are usually overworked and stretched to their breaking point, which I originally saw as the source of our broken software woes, which, like our streets in the US, just get a new layer of asphalt to cover up the crumbling bits each year instead of rebuilding the infrastructure with reliability and longevity in mind.
My former employer was using both Claude and Codex for firmware that was driving an over-burdened power circuit that itself was partially designed with ChatGPT. All of the individuals involved approach LLMs with god-fearing reverance because they do not understand _how_ the LLM works, just that it _does_ in a "good enough" way and they can offload their thinking, which is something we all wish we could do because thinking is hard, time-consuming and costly. I get it.
But like you mentioned, tests were being passed, not because the code was sound, but because the tests were altered to match the results. This is not necessarily the fault of the agent, either; it's just interpretting the prompt(s) - written by a flawed human, btw - with stochastic mechinations that seem to make a great deal of sense on the surface, but remain unable to be followed or repeated by the brains of (most of) its users.
As a rresult, I had to deal with product that work great in the field...at least at first, before it start literally catching fire, ruining its own powertrain because everything the agents touched became too complex with too many subtle cracks in the veneer to review properly. The system (read; capitalism) demanded viable product quickly to please investors, and the burnt-out humans who decided to try this AI thing ended up trusting it nearly completely, so any ideas of repeatable and complete testing, diagnostics and root cause failure analysis morphed into a sloppy "it works on the bench" checklist before being sold to a customer who had come to trust that their deceptively simple product would just work as advertised.
I'm going to die on the hill that AI as a replacement for our brains is precisely how we will make ourselves go extict, but I am old enough to already be regarded as a crufty dinosaur who is stuck in his ways, and I'm made peace with all of that. What I can't get my head around is watching people use this awesome tool (and it is, admittedly, awesome) to literally just speed up all the mistakes they were already making. Perhaps it is because I am aging, but slowing down and having a think seems more valuable to me now than it ever has, especially when creating something new. AI is powerful and, like any good tool, could be useful in the right hands, but more often than not I see it being used as an accelerant for all the worst parts of product development to appease a market that has suddenly been told they can now pick all three points on the Iron Triangle instead of just two. This makes about as much sense to me as taking a laxitive when you already are suffering diarrhea.
Eh, I miss video rental stores in the US, but I don't agree that they would solve this problem. I worked for a large chain of them as a teen, and know first-hand that people were spending just as much on rentals as they do with subscription services, especially when you add in game rentals and late fees.
More importantly, it does not solve the problem of ownership. We don't own a rented disc just as much as we don't own a DRM-controlled digital purchase that can only be accessed via the subscription portal it was purchased on.
I regularly snag DVDs for $1 USD or less at garage/yard/boot sales. I see tons of cheap used physical media in record shops when I travel, too. eBay and Facebook Marketplace are also great sources for used DVD.
I promote this because not only does it restore some mastery* over your media, it keeps these discs out of landfills and oceans, where they will linger in a shredded state.
*we are, of course, still subject to any anti-piracy measures and non-removeable advertisements on the disc, which admittedly seems quaint by today's DRM standards, but I ccannot rightly say "complete" mastery over your media, here.
I watch films on streaming rather than old dvds with unskippable nonsense like "don't pirate this dvd" (well obviously I didn't, that's why I have the dvd), or the pain of the menu systems to just press play.
I like how quickly this got dismissed as speculation as though we don't live in an age where election tampering and manipulation of public opinion for political reasons are so commonplace that incidents of it just blend in with the other forgettable global headlines.
Because it is speculation, with no special evidence. Could it be for just money? You can sell access to exploited systems in interesting companies for quite a bit of money. Or maybe it was for general use to twist public opinion in the future, not tied to those specific elections. Or just plain spying, We can't be sure, and the net was cast quite narrowly.
One could research where those repos are coming from, and do forensics on who controls the trojan network. But that wasn't done, so right now, it's all speculation. Something can be very worrying without us knowing exactly what the use cases for it will be
Indeed, and it troubles me that people don't know the difference between speculation (no matter how plausible) and analysis based on evidence. The anti-science movement (or impulse) is still pervasive, unfortunately.
that things happen doesnt mean you do not need evidence 0xEF.
No evidence was presented, so its speculation. It is _much_ more common to do other things than election tampering as desired effect of cyber activities. The amount of election tampering activities is statistically insignificant to other activities conducted in this domain.
Why would they steal credentials when governments already have fake accounts for this exact purpose (see UK’s JTRIG from the Snowden documents)
… you also have to remember that the JTRIG leaked docs were about a decade before LLMs, so you could imagine tooling these days is 100x a they used to have
Parties were not called out and a large amount of ensuing Othering is happening anyway. Arguably, that proves that the EFF was sound in their decision to mitigate that by not calling out the parties/politicians in hopes to keep the focus on the bill itself, doesn't it? I've long suspected that we humans tend to lose the plot so often because we want to immediately sort everyone into buckets as though compartmentalizing them brings about complete understanding of the issue on the table.
Can you give us an example of a new idea that is not derivative of something that already exists? Should only take about a minute.
Snark aside (and apologies), there's absolutely nothing wrong with the "no new ideas" take and nobody should think there is. Humans tend to work collectively, try as we might to do or appear otherwise, and often come to the same conclusions through reasoning and logic. No one-person truly invented the light bulb, etc, when really all inventive thought is branches of derivative thought as we build our collective knowledgebase. A better question would be how many novel ideas are the logical conclusion of branches of derivative thought and how many are tangential brought about by the injection of our irrationally.
So I guess it was a dig on OP not just giving himself equal, but top billing somehow, over his girlfriend on creating a child.
Wow those 6 seconds you contributed were what made the difference, big guy. Not the 9 months of gestation, by any means. let's hear it for OP and his splooge everyone, and so on. You get the picture.
I'm not sure what's going on exactly with Gen Z males, it's an interesting phenomenon. I wouldn't expect that kind of dialog from GenX or Boomers even.
"Tech giant Microsoft announced today it plans to build a useful quantum computer in just 3 years"
Leading with this dubious claim is fine comedy. Microsoft can barely build a useful conventional computer that runs its own platform and associated products. The Surface Book 3's being used at my job all launch Teams, Excel and Outlook like they're perpetually struggling to get out of bed. The Surface Pros have display issues minutes out of the box. This foolish Dash to Quantum will result in a lot of disappointment as unreliable rushed-to-market hardware is tauted by companies like Microsoft that are historically fond of doing just that.
And no, I don't think quantum computing is all hype. There is definitely meat on the bone, even if it only returns us to the somewhat annoying days of time sharing that most HN readers are probably too young to remember. But these idiots are fueling the hype train for the sake of quick-buck valuation, puffing for an audience that is increasingly tired of over-promising and under-delivering to the point where we will finally have an actual quantum computer and nobody will give a shit, especially after this whole AI circus has continued to prove itself an expensive pagent full of mummers and gimmicks.
I've been to a lot of countries (and thus through a lot of customs agents), the most they ever ask me to do, if anything at all, is turn the laptop on. I think the point is they want to make sure it's an actual laptop and not just a shell hiding something else. I've never had an agent touch my machine or show any interest in doing so, and I say that as someone who gets the extra searches often because I carry a lot of odd looking parts and small tools for work. Just pointing that out because I think the paranoia about what customs agents are allowed to do is a bit overblown unless you're suspected of smuggling or transporting something nefarious. They're not interested in what's on your laptop until you give them a reason to be.
I almost got denied boarding for a EU -> US flight ~13 years ago because the TSA agent at the gate noticed my 2011 MBP had 2 screws missing on the bottom panel (I've opened it up a bunch of times and lost some screws in the process). It didn't convince them that I turned it on and logged in etc. They still had doubts because, apparently, missing screws on a macbook was unheard of.. in the end, they held up the plane for ~10 mins due to waiting for a go/no-go decision via phone from some decision maker at the airline (as the final call was apparently theirs to make for some reason). Luckily, they were OK with missing screws and I was let on board.
I think it probably depends where you're going. We have relatives in a country where it might be a bit more of a concern, and we did briefly research taking a trip there to visit them, which is when all of this came up. In the end, for a variety of reasons, we decided it was going to be too risky to take that trip unless and until conditions change.
There are many countries where I wouldn't be at all worried about that, but I'd still be concerned about the possibility of theft (which, let's be real, can happen anywhere: I went on a trip to Switzerland once - generally considered very safe and low crime - where somebody had their laptop stolen from their room).
The Surface Book or whatever is going to be your best option because you want the 2-in-1 features. We had a few at my job before I switched to an XPS 13 since I never used it as a tablet and it was a weighty thing to have in my bag. Didn't hate using it like a laptop, though. Unfortunately, the price tag is also going to reflect the branding, so it won't be cheap. Same thing with a Lenovo Yoga or X13. That kind of functionality with good hardware is almost always going be pricy, I guess.
Can I ask why you want 2-in-1? I've personally never found the convert-to-tablet useful, and I have to imagine only visual artists might. I bought a nice case with a keyboard for my iPad Mini thinking I'd use it as a tiny laptop on the go, but in all honesty, I forgot the keyboard existed until I started typing this.
Not knocking your needs, just curious what kind of user those are for since I am obviously not the market
2in1s make a laptop immensely more versatile and useful:
- Tent mode is a much better to watch movies on or play games (via controller)
- In tent mode you can position a keyboard how you like, and you can put a secondary screen more how you would on a proper desktop. This way you can create a comfortable full desktop work environment on every desk.
I wouldn't even care much about the touchscreen otherwise, although it's a nice way to read articles on a train.
Didn't really think about tenting it, thanks for explaining. I don't game or watch much media beyond the occasional how-to on YouTube, so the usefulness was lost on me. Appreciate the perspective, since I frequently get asked "which computer should I buy for X" IRL because I'm the local crufty computer guy, I guess.
As someone who always favors the smaller laptops that don't require me to gear up an entire backpack just to do a bit of work on the go, I'd argue that the difference between a 10" and 13" screen is not nearly as much as it sounds. I've found the Dell XPS 13's to be an excellent choice for stowing in my service bag so I have a small-but-functional machine on a job site. That and the Dell XPS 13 just has better hardware all around, when stood up against the Chuwi.
15", sure, that's a bit big, but smaller models are available.
The thing about a diagonal measurement is it doesn't tell you if it's going to fit on a shitty airline tray table or not. Some laptops with a larger diagonal measurement are not too deep. Others are way too deep.
> "Why did it pass reviews? Well, because of deadlines. And because there is simply so much code (and so much unparsable/misleading documentation) that it's simply impossible to review all of this. And because things move so fast that nobody understands the CI pipeline anymore, and the explanations of the agent are convincing enough that surely, it knows better than you?"
I have come to realize that AI is so "successful" because the system in which it is being deployed was designed to push product as fast and cheaply as possible from the start.
Humans are usually overworked and stretched to their breaking point, which I originally saw as the source of our broken software woes, which, like our streets in the US, just get a new layer of asphalt to cover up the crumbling bits each year instead of rebuilding the infrastructure with reliability and longevity in mind.
My former employer was using both Claude and Codex for firmware that was driving an over-burdened power circuit that itself was partially designed with ChatGPT. All of the individuals involved approach LLMs with god-fearing reverance because they do not understand _how_ the LLM works, just that it _does_ in a "good enough" way and they can offload their thinking, which is something we all wish we could do because thinking is hard, time-consuming and costly. I get it.
But like you mentioned, tests were being passed, not because the code was sound, but because the tests were altered to match the results. This is not necessarily the fault of the agent, either; it's just interpretting the prompt(s) - written by a flawed human, btw - with stochastic mechinations that seem to make a great deal of sense on the surface, but remain unable to be followed or repeated by the brains of (most of) its users.
As a rresult, I had to deal with product that work great in the field...at least at first, before it start literally catching fire, ruining its own powertrain because everything the agents touched became too complex with too many subtle cracks in the veneer to review properly. The system (read; capitalism) demanded viable product quickly to please investors, and the burnt-out humans who decided to try this AI thing ended up trusting it nearly completely, so any ideas of repeatable and complete testing, diagnostics and root cause failure analysis morphed into a sloppy "it works on the bench" checklist before being sold to a customer who had come to trust that their deceptively simple product would just work as advertised.
I'm going to die on the hill that AI as a replacement for our brains is precisely how we will make ourselves go extict, but I am old enough to already be regarded as a crufty dinosaur who is stuck in his ways, and I'm made peace with all of that. What I can't get my head around is watching people use this awesome tool (and it is, admittedly, awesome) to literally just speed up all the mistakes they were already making. Perhaps it is because I am aging, but slowing down and having a think seems more valuable to me now than it ever has, especially when creating something new. AI is powerful and, like any good tool, could be useful in the right hands, but more often than not I see it being used as an accelerant for all the worst parts of product development to appease a market that has suddenly been told they can now pick all three points on the Iron Triangle instead of just two. This makes about as much sense to me as taking a laxitive when you already are suffering diarrhea.
reply