Posts by sarusa 345 publicly visible posts • joined 23 May 2022
I don't think it should surprise anyone that Google lies their ass off about how their search works.
Mostly just to keep the spammers guessing, but especially the last couple years with it all going to shite and Google embracing the spammers as being more ads to serve up.
'Software developers' is far too sloppy and imprecise. It's like 'gamers', which encompases people just happily doing hidden object games, rabid animals doing PvP MOBA, people running around solving problems in The Witness or Baba is You, people RAIDing in MMOs, and people building CPUs in redstone.
On the 'software development' side, I'm sure if you're a code pig (a giant corporation programmer stuck in a cubicle/pigpen mindlessly pounding on a very limited task) LLMs will help a lot, because you're not doing much thinking to start with. Might as well just steal the code of everyone else who's done this before, which the LLM has already eaten, digested, and shat out. Lots of room for time saving here.
If you're an actual engineer, LLMs can't help with any of the actual engineering jobs, because those are tradeoffs between the requirements, the desireables, and the consequences and resource tradeoffs of each approach. An LLM has no f@#$ing idea at all about any of that. It will happily give you O(N^3) code which ignores the requirements, because hey, it compiles. Though based on my recent playing around with Llama code helpers, even 'it compiles' or 'it does the right thing' isn't guaranteed.
Basically, no LLM is 'thinking' at all. It is stochastically regurgitating all the things it has seen before. So the more your job involves actual thinking, and the less it involves going on StackExchange and copypasta-ing code snipped and randomly smacking them till they compile, the less threatened you are. And the less LLMs can help you. Like I said, I've been playing with this, and the best it can do for me is a line or two of auto-completion (and it's wrong at least half the time).
There's no way they're going to let this loose for everyone (even just in China) to poke at. They're going to need censors scrubbing the output at all times lest the Xi-bot say anything (gasp) non-Xi Jinping Thought approved. And all LLMs so far are very easy to jailbreak - there's no reason why China's would be any better, especially with most of it being 'acquired' technology. So anyone who knows what they're doing with an LLM could make Xi-bot say the funniest, most treasonous things (and the bar is very very low for treason).
Probably their best use for it is to auto-generate his usual rambling, vacuous, pronouncements on every subject under the sun for publishing, saving the real Xi a lot of work. Though he obviously loves to pontificate, so you'd have to strike a safe balance between freeing up his time and taking away his joy of preaching.
I can see other cases where something similar could work better, though. A Trump-bot would likely be more coherent than the real thing, would lie - okay, about the same amount. But would definitely be much less malicious, and at least wouldn't reek of filled diapers. And a Johnson-bot would be pretty much indistinguishable from the real thing for speaking, at least? And wouldn't force your liquor cabinet open and guzzle all the good plonk.
This was after the nGage, and after Nokia fought hard against doing anything touch screen because that's a fad (there was the 7710, but that was meh, and they obviously believed much more in things like the N95), so yeah, they were on the way down. And then add Microsoft desperately trying to claw back into smartphones and you've got a 'you got your poop in my puke!' 'no, you got your puke in my poop!' situation.
(There are Nokia branded Android smartphones since, but very meh).
I've said this for years and never had any actual proof, but Ed's gone through Google emails from a lawsuit and yes... Starting 2019, Google has been purposely degrading their search results to make you click through more bad results so they can show you more ads.
https://www.wheresyoured.at/the-men-who-killed-google/
Kagi doesn't seem to have any problem identifying and removing most of the AI generated spam.
It should be pretty easy, really, when you have the entire f@!$ing web cached like Google does. A site that just pops up out of nowhere with tens of thousands of garbage pages and links to other sites that just popped up out of nowhere with hundreds of thousands of similar garbage pages? Yeah that's probably spam. Individual spam results are also pretty damn obvious to any human reader, why can't an LLM go 'this looks sus'? Google has DeepMind for goodness sake. Okay, yeah, I just asked ChatGPT 3.5 to identify 5 pages as spam or not and it IDed all 4 spam sites and correctly said the one legit page was legit. Even if it fumbles on some pages, if a single site's pages are mostly IDed as spam or mostly IDed as legit that should be effective.
Of course there are some savvier offenders, so those will still sneak through. But I think the big reason is that Google just wants you to have to click through more bad results so it can serve up more ads.
And of course Kagi also lets you mark individual sites as spam, never show me anything from this domain ever again. Which Google used to let you do, but no longer does because... yeah, more bad results for more ads.
"rogue insider"
Yeah, as the rest of the article makes quite clear, this was just normal practice for Ring employees and contractors. As was stunningly obvious from the start - if you give minimum wage grunts (or maximum wage incels) the ability to look at naked people on cameras, they will do so enthusiastically. And I'm sure they were passing the best clips around. But Ring just didn't care, why should they?
I've been using Start 11 almost since Windows 11 hit and thus avoided most of the enshittification* they've pulled on it except when I see it on other people's computers.
Choose which type of start menu you like, your colors, how expanded it is (tiles? compact list? icons? very customizable!). No ads. Customize your taskbar too, get rid of Windows's crappy search. Return the start icon to the lower left even if you're using centered icons! It just works and doesn't get in my way, which is all I ever wanted from the start menu.
https://www.stardock.com/products/start11/ . If you want OSS alternatives, people recommend OpenShell and ExplorerPatcher, but they weren't quite as nice to me. YMMV!
(* I know, some of you really hate the term 'enshittification' but it really is the best way to describe what they've done to it)
It's (extensively modified) forked Linux/Android, at least on Phones (it uses LiteOS for smaller things) - China doesn't build anything from scratch, it's always 'borrowed'.
The UI, which is probably what you'd actually care about, is quite similar to Android, though clunkier since it started as trying to clone early iOS using Android (all the apps on the home screen, no folders, etc.). It can do widgets and folders and cards now, but you can tell they're more interested in adding more features than actually fixing the UI. You can do Android apps through a compat layer, though you're supposed to go through the Huawei AppGallery to get 'native' HarmonyOS apps, and of course the native apps have access to all the other things Huawei's added to stock Android.
For sure try it out, but I don't think you'll be blown away. There's no major move away from Android or iOS here and I wouldn't say it's better than either. But not horribly worse, totally usable!
I don't know about you guys, but I LOVE the Sports Futility Vehicle. Seems like not a day goes by without something new and hilarious as the manliest most rugged of incel vehicles for surviving the impending apocalypse breaks down again and can't handle something my bog standard Camry hybrid handles with no problem (like a hill or car wash).
Like one where a guy in Santa Cruz tried to do donuts on the beach, something any pickup or SUV should be able to handle with ease - he managed one and a half circles before his SFV broke down and got stuck. And I'm sure it was totally rusted out by the next week from all that salt.
DeepMind is 10x better than any of their other AI teams, so it's nice they're nominally in charge.
The 'under one roof' is a little weird though. Even DeepMind has teams in the US, Canada, France, and Germany, with the main team in London. Much less all other Google's AI teams scattered around wherever a nepotistic uncle wants one. So I guess he means just organizationally, not literally? I wouldn't even have considered that he might even mean it literally if not for all their shitty moves on WFH recently.
Whoops, I have one of these TVs. But I have, never, ever, let it have access to the internet. How stupid would you need to be to let your 'smart' TV, which will report everything you're watching back to home base along with photos of your living room at the time, have access to the internet? It's just a dumb screen for the media box.
Yes, there are different levels of 'do I trust this?' but
1) Rust macros will still not generate code that will crash from memory access errors.
2) Yes there can still be logic errors - as when you call any functions or any libraries.
3) It's still way down the horror scale from C++'s macros and templates.
The big difference is that macros generate the code inline, functions all the same code elsewhere. And macros let you do things more succinctly and clearly than you could do otherwise (as in c/c++). Of course you can abuse it and make your code actually less comprehensible. But still better than C++.
ssokolow had some good thoughts on this, but for me it's really that there is nothing at all for rust that makes a GUI as trivial as Visual C# does.
Right now, rust is a medium to low level focused language. GUIs are an afterthought, as with python.
There are some semi-decent options, like slint and tauri. There's also iced, but I don't really like the elm model, but YMMV - some people like it a lot!
For now I'd rather write a GUI in rust than C++, but I'd much much rather just write a C# GUI and call into rust. I hope this will change soon.
So as someone who has written large systems in C++ for 20 years (including servers, GUI apps, and embedded) and has been using C# for 10 years, Python for 15, Rust for 1, the latter three really are easier to write, read, and maintain (with one exception I'll get to) for one simple reason: C#, Python, and Rust are written with some thought to readability. C++ is performance uber alles, even if it makes the code hideously ugly and the compile errors nigh unreadable (once you add in templating).
C++ isn't perl levels of write only, but it's up there. I can go back to Python code I haven't looked at in two years (I did just last week) and oh hey, I can tell what it's doing and make changes, no problem, because it's readable. I go back to two year old C++ and it's just a slog. Part of this is the hideousness of C++'s .h files, which require you to pre-declare everything external in the .cpp file *in a slightly different syntax*, so you're tediously maintaining two slightly different sets of declarations. Part of this is because C++ is just too damn verbose. Not Java levels, but the fiddly syntax requires you to spend way too long describing *how* to do something and not *what* you're doing, even using STL. And of course there is a ton of room for accidents, because you're so focused on the detail you miss the bigger picture or vice versa and the language certainly isn't helping much. You spend inordinate amounts of time making sure you're not doing something other languages just make sure doesn't happen by default.
Now yes, you can write amazingly beautiful and readable code in perl (I've seen it) and you can write perl in python (ditto), so I'm sure you can write safe code in C++ - I certainly have non-leaky and non-crashy C++ programs (or at least they appear that way), but they took a lot of work. What really matters is what the language encourages and C++ just doesn't encourage readability or writeability - you will always have the problem where you have to look at both the cpp files and h files simultaneously for instance. And that's critical when you have more than one person working on something. And guess what? You who haven't looked at your own code for six months is 'another person'! So I find Python, C#, and Rust far more maintainable than C++. What I've done for 8 years or so is just save C++ for the very very few extremely tight and fast operations (less than .01% of the code), and everything else is C# or Python and calls the C++ when it's needed. This has been a 10x boost in productivity. Well Rust is blurring the line. I can use it for what I used to use C++ for *and* it's expressive and capable enough that I can use it for higher level stuff too. It's not good for GUIs but is just fine in the middle.
Sorry this was so long, but I've actually been in the thick of this. And what was the one exception? Well, python's lack of compile time checks makes it awkward for large systems - I'd much rather have things fail at compile than at runtime. I still love it for small to medium stuff though. tl;dr if it didn't pay so well I would never touch C++ again.
It's the US working with the EU. Which the UK would have a hand in if they hadn't had their petulant angry toddler fit eight years ago. Nor is this going to 'imposed' on anyone else. There's no way they can force China, India, Russia, Japan, or anyone else to use it. This is UTC for the Moon, and everyone on Earth uses their own TZ and then converts from/to UTC as convenient.
It's just a thing that is needed for proper operations when 'what's the exact time on the Moon down to a couple microseconds' actually matters. It needs to exist in some form, so the people who work together the most (NASA and ESA) are getting together to decide on their own standard for a rock solid Moon time. Then their collaborators at JAXA, sometimes ISRO, etc. can use it if they're doing joint missions, and they might as well use it otherwise unless they really want to design their own for nationalist reasons. If this standard just ends up being 'UTC with adjustments for the Moon orbiting the Earth' that's pretty darn uncontroversial. Unless of course you're China or Russia, but nobody cares if they're off doing their own stuff on their own time. Especially since the only thing Russia can do with the Moon at this point is crash things into it.
It's a little weird since the Moon is locked to the Sun, so if you're in the center of the bright side, that's noon, and it's always noon! Any place on the Moon is always the same time by the way we think about time (the progression of the Sun).
Of course I realize that's functionally useless, so basically lock it to some time zone on Earth plus corrections for it orbiting the Earth. If they tie it to UTC as 'baseline time' with the corrections that seems pretty reasonable?
They've been wholescale and fairly openly performing corporate, scientific, and military espionage for at least four decades. It's where they got all their tech when just partnering with stupid greedy western firms and having them hand over the tech directly wasn't working fast enough. Western governments tolerated it for a long time because the (stupid, naive) thought was that once China got rich enough they would naturally become an enlightened democracy, so we should wink wink nudge nudge that. That started becoming awkward in the 2000s, then really awkward when you combined Xi aggressively turning the place into a fascist dictatorship plus the Angry Toddler in the US being livid that someone was doing everything he wasn't allowed to do.
Now there have certainly been some overreactions, especially in academia - because governments have all the subtlety of Greased Piglet Boris - but everything here is completely in line with what we've been seeing from mainland China.
So just for the heck of this I tried it on gemma, mistral-openorca, and mixtral - and ChatGPT 3.5 (never mind 4) just destroyed them all on all the creative questions I asked them. ChatGPT 3.5 gave me actually creative answers, all the rest of them gave me answers like they were phone support reading off a checklist.
Then I asked deepseek-coder to 'generate python code to solve arbitrary nonograms' (picross) and it responded by telling me what a nonogram is (gee thanks) then 'Unfortunately generating Python code to automatically solve arbitrary Nonograms is a non-trivial task.' No s$@#, Sherlock. And then it gave me code that (by inspection) only works when the clues are absolutely trivial (in a 10x0 grid, every row/col clue works out to 10 cells) and doesn't even run. Then I asked OpenAI's python model to do the same thing and it gave me much more plausible looking code (including backtracking)... that completely failed when trying to solve even simple nonograms. I am happy to see that coding AIs still pose no threat to actual coders who know what they're doing whatsoever.
He has been very very lucky, very privileged, and yes, kind of bold.
He got started by his daddy giving him money from daddy's emerald mines. He took that and started a shitty worthless internet bank, but then Peter Thiel needed an internet bank for Paypal and merged with Elmo's shitty worthless internet bank, giving him his big break.
Since then he's been half good at having smart people tell him what to invest in - he invested in Tesla, then kicked out the founders. He started SpaceX. In both cases, both companies have entire teams of people dedicated to keeping Elmo from doing too much stupid damage because he is a stupid mingeing twat. When they don't manage to keep him from making his own decisions, you get the Cybertruck. And of course there's Twitter. And then there's the hyperloop (which was obviously right out of the gate an extremely impractical idea from a very stupid person) and the Boring Company.
He's been so busy with Twatter and the Cybertruck at Tesla recently that he has had hardly any time to eff SpaceX up, so it's doing great.
Basically if you rich, if you are lucky, and you are willing to take a lot of bets, and most importantly you can find much smarter people to run your stuff, you too can luck into being super rich.
I appreciate your honest question, and honest answer - I am a person who responds very well to numbers and gamification.
Before I got my iWatch I was really 'eh' about fitness, ugh, I could go for a walk, but I could play a game for an hour so why bother?
Once I got the iWatch (anything similar like a Fitbit or android health tracker would work too), I became a health monster. This was a conscious decision, I kind of knew this would happen. Now I can see 'oh, I walked 8 km today and burned 1500 kcal'. And, being a gamer, I escalated. I went from being hardly able to walk 2 km (uphill, admittedly) to now being able to walk 20 km at 10 minutes/km and hike up 400m - just because every single day I wanted to exceed or at least meet my previous records.
I am now absolutely super healthy. My blood pressure is amazing, my VO2 is amazing, I can outwalk or outbike almost all the people I know. And it's all because I fetishized the information.
YMMV, obviously.
Well, like any other smart watch the benefits are:
* Hardly ever need to take the phone out for texts, emails, etc, just glance at the watch. Can even reply with the watch if needed.
* Believe it or not, it can be useful to just glance at your wrist to find out the current time and maybe how much time is left before sunset.
* Phone never makes any noise to annoy everyone around me, the watch just vibrates.
* It's impossible to miss a call/text with the watch vibrations - though of course you can turn that off when you want to, and I only have it buzz for people on the whitelist.
* Replaced my alarm clock - again, can't miss the watch buzzing.
* If you accidentally drop/leave your phone, the watch will warn you. This has saved me twice.
* Smooths a lot of interactions - for instance it's super simple to just use Shazam to ID a song by tapping the watch twice rather than pulling out the phone, unlocking it, swiping to find the app, running the app. Also makes for a great conversation piece at times.
* I find the fitness tracking to be endlessly useful when I go for my walks, hikes, or bike rides. Turns out my pulse rate tells me a lot. And of course actual distance traveled, speed, steps, and altitude changes are really useful. And with the watch you don't even need to bring the phone (though I usually do).
* When biking, voice commands to the watch can be very useful. I have actually gotten myself lost in a new area and just hold down the fob and say 'navigate to [x]' and off it goes with turn by turn instructions.
* Makes paying for things super easy - just tap the watch, no need for hauling a card out of your wallet (which only old people carry now) or phone.
* Works while swimming too, forget using your phone there.
The only thing here that /only/ the watch can do is the fitness tracking - without a smartwatch I'd have to use a fitness band, which is basically a cut down smartwatch. Everything else can of course be done more clumsily without the watch. But it just makes everything faster, easier, less friction, and isn't that the entire dream of tech? Hell you don't /need/ a smartphone, it just makes things much easier. And the watch makes them easier still.
Of course until you've actually experienced all this you can't even imagine how much faster and easier it is doing many things with the watch instead of the phone so there doesn't seem to be a killer feature. I got my first smart watch (Moto 360) just for the texts on my wrist (and the custom faces). The rest I found out later.
There are people at SpaceX and Tesla who are part of secret teams dedicated just to keeping Elmo from doing too much damage by trying to manage because he's a complete dumbass. The only thing positive thing he has is money to fund cool stuff, which is great, but the more you let him actually make decisions, the more you get Twitter or the Cybertruck.
He's been too busy with effing up other things to do much damage to SpaceX recently, and because of that they've been doing amazingly well.
So yeah, anyhow, any sane, non-spiteful decisions are coming from someone who's not Elmo.
A computer that can crack 256 bits would take about 13 million qubits for a full day, or about 317 million for an hour. Obviously quantum computing is nowhere near that yet.
But qubits are going up at Moore's Law, and there have recently been some designs that are much more stable (so need far fewer qubits to error correct).
It's not unreasonable to think that at this rate it will be 10-15 years before quantum computers can crack these, especially with China going all in on this and being capable of stealing any new technological advances.
Then consider how hidebound and glacial banks are, how incredibly thick and glacial governments are, and that *all the previously encrypted stuff that was ever out there will have been saved and is ready to be cracked*. Like global warming, unless you start planning for the inevitable future NOW, you are going to suddenly be looking down the barrel of the gun going 'Oh my stars and garters, lawd awmighty! Nobody ever warned me about this! Nobody could have ever seen this coming!' Everything that was ever encrypted with 256-bits will have been captured and saved, ripe for cracking. If you make it 4096 bits that's only 16 more times computing power for a classical cpu, a couple more qubits for quantum. The only way out of this is going for a new algorithm that even a quantum computer chokes on (and people are working on these).
10 years is not too soon to be considering this at all.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
