Yeah, the 600(e) was 'peak laptop' - tough as old boots, good keyboard, powerful (for its time), had decent support for drivers and whatnot so you could re-install Windows and get rid of the crapware IBM bundled with it, and you could run Linux too, if you wanted.

Since then, it's hard to find anything as good - although I've got to say my Apple Macbook Pro is pretty good. It too is pretty tough, doesn't run a bloated windows version, can run Linux, but is of course not upgradeable and is pretty much unrepairable by mortals, so there's that, I suppose.

Being a long time slashdotter, I'm not sure the fashion industry sees me as a 'target market', but...

When I look at a model wearing some clothes, I kind of need them to look a bit like me. I'm not really able to do the mental gymnastics to guessimate how it might look on me, even if it looks good on the model. As fabulously good looking as I am, I'm not 'idealised' in any way. As such, I'd really appreciate seeing 'ordinary people' modelling clothes, because then I could at least get a feeling for how it might look on me. Over the years, I've become so aware that I can't do this very well, that I now more or less avoid anything being worn by the young and cool that-don't-look-like-me, for fear of making a mistake.

I realise of course, they're not selling clothes per-se - they're selling the dream of looking super cool and having the opposite sex fawn all over you at every moment. But I'm too old to be swayed by that any more, and whilst I'd love for the young, taught and good looking to be fawning over me, it's not really likely.

So eventually then, I hope that if something good comes out of all this, it'll be that they can switch the model around depending on who's looking at it. If the young and stupid are looking, then sure, put the idealised super cool model in the clothes - do what you like. However, when I'm looking for clothes, that's no use to me, so show me how people-like-me might look in whatever clothes you want to push towards me. I might even buy them..

The other good thing that _could_ come out of this, but on present trajectory looks like it's never going to, and is being burned with fire at every opportunity is body positive advertising. If we over-idealise the models, to the point that models themselves can't even look that way, then we're just making trouble for ourselves.

I know there are a lot of big company CEOs saying this stuff, but I don't personally believe it. AI is doing *something*, but it's not as profound as all that.

I think there's just a general downturn, and it's hitting tech harder than most. The tech industry, and particularly the AI industry can't admit they're not peddling quite as much of their crap as they once were, so they're cutting their costs where they can - and are masking it as "efficiency improvements thanks to AI".

It's not all about the USA, but recent US Economic policy has been hit-and-miss at best. Whatever your idealogical thoughts on the matter, it's left most of the world holding its breath - and that means no new investment in anything, be it a big new plant in $country or even just hiring one extra person in your small company. No one wants any long term liabilities right now - just in case it all changes again, or indeed if we do have global recessions.

That all pre-supposes people have money to make long-term investments - which I suspect they don't. Inflation has hit just about every modern economy in the world, and there's a cost of living issue in most economies as a result. That means the masses just aren't buying as much, which means the mega corps aren't selling as much (and probably aren't advertising as much - hence the tech slowdown).

Most countries don't have a lot of spare money for investments right now either. My own country gave away all of its money to big companies during the pandemic. We didn't actually have piles of it before then, but we've sure got a lot more debt problems now. Even the government has to tighten its belt (or put up taxes, which it's been stupid enough to do already - rumours are it'll do it again soon too). So you're not going to see government stimulus doing too much for the foreseeable either - in my country, you might already argue the government is doing the opposite of stimulus, and that argument may get stronger if they make more policy mistakes. No AI was responsible for any of this.

All in all, yes, AI has its part to play in all this, but it's more likely the excuse for inaction, or the excuse for cutting costs more than it's the cause of any of it.

True - this works so long as the thing you're making doesn't get popular. If it does, then it'll be purchased by an agent, sent to China and copied by the end of the week.

So it works - for niche products. Mass-market will never be immune from some-other-country-doing-it-cheaper.

I was wondering the same. My very rusty, and beginner-at-best knowledge of lightning is that you get static charge in the cloud, some air ionisation at ground level, and then that makes a path for lighting to travel from cloud to ground. Lighting tends not to flash in exactly straight lines because of air movement and differences in the path it finds. I'm not clear how lightning, which can only be, what a couple of kilometers up could travel many times that distance sideways before finding ground.

Anyone know how it can happen?

PageRank doesn't really work on the Internet any more either - how often does anyone link to the things they're talking about? At best you might have some references at the bottom of a page, but most of the time, sites don't really have hyperlinks in the text. As such, a lot of the context for pagerank is lost, and all you have is "number of links", which is essentially link farming.

That said, Google do still have some notion of "reputation" for web sites - higher rep means higher results. I suspect that's how Reddit suddenly went from no-one-really-sees-them to suddenly on-just-about-every-search-you-ever-do - somewhere along the time, some money changed hands and away they went.

Having now tried to read some reddit rather than scrolling on by and finding something better, I've got to say, it's generally pretty low quality crap. If that's the basis for this awesome new search engine, then Google don't have much to worry about.

AI is proving to be something of a game-changer for search, but until it's better at citing sources and not spouting bullshit, there's still room for traditional search too. The problem (it seems to me) in AI search is that it relies more on the number of sites saying something, rather than the reputation of those sites. I'm sure they'll work on the reputation/pakerank sort of problem and then sort of "extra train" their models on higher reputation content, but it's fraught with difficulties. Exactly as Google seem to have decided Reddit's quite good now (when it probably isn't really), so too will the AI model makers have to make such decisions, and so it will pollute their models with skewed knowledge (or lack thereof). Then, even if it cites its sources, those sources will be so crap as to be worthless.

I don't know where the answer lies - so far, Google's results + AI summary might be the best (if they can fix up the AI, and frankly, fix up the results as well). AI on its own seems just too risky to me.

Yes indeed - and most people can't storyboard a 2 minute video terribly well, let alone an actual TV show. Youtube is already full of AI generated slop, and this will just make sure there's even more of it. Virtually unwatchable TV - can't wait.

In the UK we actually have some really terrific TV. $show might not be to everyone's taste, but in most cases they're well written and well made. I suspect it's because we have the BBC - love 'em or hate 'em, but they've kept the general level of quality pretty high. They consistently produce (or buy in) highly polished stuff, again, not to everyone's taste I'm sure - but well written, well made stuff. Other TV companies just can't produce crap and expect to get much viewership because there's better "for free" right on the BBC. It's kept general programming up, and it's also made advertising better too (weird, I know - but I guess if you're competing with an ad-free service, you'd better make sure your ads don't turn viewers off your show).

Contrast to the USA... there were always hundreds of channels when we only had 3-4, and from here, it looked like that was completely awesome (we imagined 100 channels like ours). Then we went on holiday and watched the TV to find that not only do the ads seem to take up most of the day, but the actual TV shows are absolute crap. Every once in a while something good comes up, but my goodness there's always been vast swathes of absolute dross.

So again, if you're competing with US television, then this AI generated stuff might actually stand a chance of at least being on-par. It's less likely to make the next Friends, or Big Bang Theory, but it'll cover the 99 channels of crap requirement quite adequately. How it'll fair against other countries is anyone's guess...

If it's free to the user, it'll be abused left right and centre. Even if you have to physically go to the library to get your VM allocated to you, the bot farmers will find a way to move their compute to it before you know it.

I agree trying to run everything yourself isn't the future. Hell, I'm a sysadmin, and I'm actively trying to get rid of as much stuff that I run for myself as I can. Just keeping up with the updates and the inevitable problems discovered by my family, and it's just not worth it. That said, Microsoft are making their 'family' offering so unattractive, maybe Libreoffice and my NAS aren't so bad after all!? ;-)

I don't know what the answer is. Co-location isn't really any better than running it in your garage, so then you could go to a managed services type model. You co-locate your stuff in a rack somewhere, and a couple of people look after it for you. The trouble is, no one wants to try to figure out what hardware to buy, and those people doing the looking after will probably start asking for specific hardware/software, which likely means your random project isn't going to be easy there. So then you're back to cloud hosting...

I wonder if there's a "guaranteed cloud" option. That is, you contract for (say) 3 years, and the cloud provider has to provide whatever service you have today for that long. They still have to do security updates and such like, but they can't go adding and removing features and moving stuff about for the sake of it. That sort of solves for a lot of the cloud vendor issues (except bankruptcy), but at the end of the 3 years you may have to go through a full server migration onto the (now) current version, or else "legacy support" could get expensive.

I'd also say that the Apple comes with updates for likely 5 years or more. Those updates are typically pretty simple to install too. Contrast to Windows, where somewhere in its life you can expect an update to complete screw your system up, and those updates are likely to cease after a few years because they'll want you onto Windows 12 by then.

I'd add that VC money goes on lots of companies "just in case" they make it big - not because they can necessarily guess the future. You see it all the time - a fad comes along, and VC seems to be falling over each other to invest in the hot new startups in that space (right now, it's AI).

VCs are also a different thing than most people think. That is, they "invest" $10million in your business. You have to pay back a monthly fee every month, regardless of anything else that's going on. That's effectively you paying the VC some percentage returns on their investment - even though you haven't necessarily made any money.

VCs will also press you to use their consultants to "really help your business", and you can expect those consultants to cost no less than $5k/day. Then of course they take dividends, likely before you get any (as they'll have preference shares, whereas you do not).

So whilst on the outside it looks like they've thrown down $10 million in the hope it might come back to them in 3-5 years, the truth is, they're making money every month regardless of the success (or lack thereof) of the business.

Then of course, if your business doesn't do so well, you can expect to be replaced - by one of those consultants. The VC will be extracting money from the business it's trying to save so that if it succeeds they do well, and if it fails, there'll be literally no money left in it for anyone else to take.

The real money makers are the companies that make it - and they can expect maybe 1 in 30 to be like that. More than 1 in 30 will be a reasonable going-concern, but every once in a while, you find a company that either gets sold or IPOs for 100 times the value it was after the VCs put their money in. That big pay out pays for dozens of failures, and an awful lot of VC bonuses and company perks.

So whilst the average mid-level VC might not look too bright on the outside, they're likely making their employer plenty of money in the long run. They're probably making a decent amount in the short term too. Maybe not bright enough to build their own VC business, but bright enough to make plenty of money.

If you've ever been to a Russian airport, you might not be able to spot the difference between "disruption" and "normal service". Having vast swathes of people all standing about waiting for something or other is Normal Operations for Russian airports.

Carefully organised lines and areas isn't really their thing. I don't know if you need to, er, "pay for the express service" or what, but normal people look like they spend hours or days at the airport at a time. One wonders if it's not quicker to walk wherever you're going than it is to fly there.

I'm not surprised Anker are leaving the market either. I've got to say, some of the new breed of printers are really pretty awesome. There's not much room for profit any more.

I bought a Bambu A1 - £300 or so (call it $450?). It prints in super-high quality faster than my old printer could print draft. It's also really well calibrated, so a 50mm part really is 50mm, circles really are round, etc etc. Hell, this thing's got a mini LIDAR in it to check on the extrusion. If you buy their more expensive printers, they've got cameras watching for defects in the printing process.

When printing with PLA, apart from a few tall and skinny parts, it's printed perfectly without any help from me - pretty much the same as our laser printer, but for 3d parts. There's almost no post-processing required on the finished article either. I have tried some more exotic materials, and yeah, this entry level printer isn't so good at them - it can do it, but it takes a bit of love.

I didn't buy the filament switcher thing, but if you do, you can have multi-colour prints, or use something soluble for support. It changes filaments automatically and you don't get 'running' from the old colour into the new one.

Honestly, the new breed of printers really are incredible. The old Prusas, Replicators, Ultimakers and such like really don't compare (those vendors may have new models that compete - I haven't checked). I couldn't find any that could compete on price with the Bambulabs ones, and honestly, if they stop making parts for it, then it goes in the bin and I buy someone else's. The days are gone when an entry level printer would be £1000 upwards.

It's worse than that - even if you have a local-everything device, don't ever do firmware updates either. That's what happened here - it didn't *need* the cloud until they did an update, then all of a sudden it did, and for that you needed a subscription.

Honestly, *any* complex product is potentially vulnerable to this sort of move. I can't imagine how you could research before purchasing to mitigate against it either.

(Home Assistant is theoretically just as vulnerable, but being open source, you'd imagine it would get forked and moved elsewhere pretty promptly)

These guys could connect to all sorts of stuff, and I don't think cloud access was mandatory. So you could have local-everything if you wanted.

The vendor then sent out a firmware update which made a subscription mandatory - which presumably needs a cloud connection to verify it's paid for. They also took away the local access API, so you *had* to use the cloud.

Honestly, this is a really horrible move, and if it happened to me, would have me moving off that platform with immediate effect. You can't expect to fundamentally change the nature of the product and keep your customers (or at least, not me). I'm not sure how any of their customers could ever have researched or otherwise mitigated for this though - back in the day, it had all the 'freedom' you could want. It just didn't have a very solid business.

I personally use Home Assistant (having used OpenHab for a few years before that). It's "local everything", and it's great. However, in theory it could go the same way as FutureHome - one little software update and it'll all turn to dust. HA is a lot more open-source and not for profit though, so if they ever did do something this stupid, you can bet it would all be forked and reanimated somewhere else in short order. Not possible with FutureHome though.

The problem is, you can't buy any complex products that don't connect to their maker somehow. You can only buy fairly dumb devices (eg. a smart switch or thermostat). The hub is, by its nature far more complex and so needs to get updates from its maker.

The problem, it seems to me, is 'old school' product people like to make boxed products and sell them. That's fine, but for every one sold there's a burden of providing updates to it - and that's fine too, right up until you make a new model. Now you have to support the old model which you're not even developing on and the new one - you've got a lot of extra work and no one paying for it.

It feels like a lot of people starting companies still don't get this. Ultimately, *all* complex products will need to have a subscription* They don't *have* to have cloud access and cloud control, but they will need some sort of subscription to pay for the updates (especially after 2-3 years since the model was introduced). Either that, or the product has "planned obsolescence" and they cut it loose after 3 years of making it (plus maybe another year or two for the late purchasers).

* It's possible you could make products without subscriptions, but you'd have to take an Apple style approach to doing so. You need to build in a lot of profit up front when the product is sold to pay for the updates for however long you plan to offer them. That's okay for Apple, but it's going to be tough for industries where price is the deciding purchasing factor (ie. most electronics).