Grow up? And where, exactly, are the profits to be found in that?
Posts by vtcodger
2026 publicly visible posts • joined 13 Sep 2017
Page:
Palantir designed to 'power the West to its obvious innate superiority,' says CEO
China's DeepSeek just emitted a free challenger to OpenAI's o1 – here's how to use it on your PC

Re: "How many "R"s are in the word strawberry?"
Indeed. Who knows what fate might befall us if the Chinese were to find out how many "R"s there really are in "STRAWBERRRY"?
==========================
Seriously, One has to expect that any information sent to publicly accessible servers anywhere on the planet is probably an open book to the national intelligence agencies of most developed countries if they choose to look for it. It probably matters not at all if information is sent to Beijing, Buenos Aires or Topeka. If someone with resources wants it, and it's on the Internet anywhere, they can probably get it. As can the CIA, MI5, IRGC (Iran), Mossad, etc, etc, etc. Not to mention Google, Meta, etc, etc, etc. And millions (probably) of hackers.
Welcome to a world where everyone on the planet is your creepy next door neighbor.
Tool touted as 'first AI software engineer' is bad at its job, testers claim
Clock ticking for TikTok as US Supreme Court upholds ban

Inquiring minds want to know
"It includes the website."
And your plan for enforcing THAT is ... what? Install a national firewall in order to protect the children? Remove TiKTok from every DNS provider on the planet? And what's to prevent users from simply using one of the site's IP address to access it? I expect that TikTok or friends could come up with new IP addresses a lot faster than ISPs and such can install blocks,

Re: Impossible to ban?
Maybe.
OTOH, anyone who has ever worked with teen agers will tell you that they are just as clever/smart as adults and are prone to regard rules as challenges rather than imperatives. All it takes is for just one of them to find an easy way around this ruling and 60% of the teen age and young adult population will quite possibly know how to stay on TikTok within 72 hours. A great many, and not just teens, might do exactly that.
Not all teen age mentalities are in the 13-19 year age bracket. One need look no further than the typical politician to recognize how many folks never develop beyond the maturity and social attitudes of a confused 13 year old.
AI datacenters putting zero emissions promises out of reach

"Building nuclear power stations is a terrible option."
Why yes, it is. However, contrary to your assumptions, It's probably the least bad of the available options. Hydro is or soon will be pretty much maxed out. Geothermal and tidal have very few appropriate sites. That's really true of wind as well. Solar is possibly OK in the tropics. But, when properly assessed, it's far more expensive than its advocates believe. It needs a huge amount of back up storage. And, at scale, it's ugly and really shouldn't displace farmland and/or forests except in exceptional situations.
When you do the arithmetic, nuclear (fission or fusion) seems to be the only long term option for supporting 8 or 10 billion people at a reasonable standard of living. See UCSD astrophysicist Tom Murphy's DoTheMath.ucsd,edu archived posts for the supporting arguments ... and the math.
Other than that, most of your ideas are fine. Really. They simply don't look to be adequate to make up the gap between what is needed and what other technologies can provide.

First of all, "Net Zero" has NEVER been a realistic goal except for a few places with small populations and abundant hydroelectric/geothermal or other non-hydrocarbon power. Published national goals are consistently missed -- everywhere and always. Whereupon the IPCC laments (yet again) that humanity is doomed. That's quite possibly true. But not from climate change I think.
However, I think the fear here, may be somewhat realistic. A few AI machines are probably not much of a problem any more than a few hundred supercomputers are a problem today. A world where everyone is using AI might conceptually have many millions of AI machines. I don't think that will happen. But I guess it could. And THAT would be a problem at least unless/until improved algorithms and better hardware reduce power demands to something tractable -- which may take quite a few decades
One answer, which almost certainly will not happen, would be an international agreement to limit AI to research on a one facility per 50 million people basis. That's something less than 160 machines worldwide. Smaller countries could form consortiums to reach the 50 million threshold. Germany and France would each get one machine, Japan 2, the US 6 -- maybe 7 by joining their excess population over 300 million with Canada for a joint facility. Such an agreement should also forbid any commercial applications until they have been thoroughly vetted and everyone agrees they are harmless.
But that'll delay AI . Indeed. So what? It's not like AI is anything the world actually seems to need.
Tesla recalls 239,382 vehicles over rearview camera problems

I disagree. The mirrors are fine for keeping track of traffic in back of one. But when backing up, a rear view camera mounted lower down and including much more of the area in back of and near the car is far more useful. Of course, if you never have to back up, you probably don't need a rear view camera.
Just when you thought terminal emulators couldn't get any better, Ghostty ships

Re: Emulators?
"I tried vi once ..."
Exactly my experience actually. I tried vi back in 1992 or so and found myself looking at a blank screen with a blinking cursor. It let me type stuff and echoed it on the screen, but discovery of what to do if I wanted to do more than type was entirely lacking. Like you I ended up killing it with the power button. Of course, vi discoverability is much better now days. But the few times I've tried it because it promised some nifty capability, it looked usable, but the nifty capability turned out to do something that wasn't quite what I needed. So, I continue to use emacs. The good news. Emacs has great discoverability and configurability. The bad news. Many configuration changes require writing lisp code. Lisp is not even remotely my favorite programming language,
Trump's tariff threats could bump PC prices by almost half

Backwards
If the tariffs targeted only luxury goods -- Caviar, Champagne, and such, they would indeed be a tax on the wealthy. However Trump was talking about across the board tariffs on everything foreign including essentials. Aside from being a rather dumb idea with a long history of awful results, that would surely hit the working class harder than the wealthy who have a far higher savings rate and thus pay a tax lower percentage of their income than the average citizen. Since the tariff revenue would presumably be used to finance yet another tax cut for the wealthy, the result would be yet another transfer of wealth from lower and middle class Americans to the wealthy.
Boffins ponder paltry brain data rate of 10 bits per second

Re: Thought experiment.
Some English language speed readers can manage 1000 words per minute with reasonable comprehension. That is almost 17 words per second. And words are obviously way more than 1 bit each. I think these folks might have a very different definition of BIT than those of us in the computer and communication field use.
Starlink direct-to-cell is coming to Ukraine

It's complicated
"Have I got this right, a standard mobile phone can transmit a strong enough signal to be picked up by a starlink satellite orbiting at say -checks Wikipedia- 1100 km??"
Just spent 20 minutes plowing through numbers, I'm mildly confused and quite likely wrong, but I think the situation might be something like. The original plan was to put the Starlink satellites at 1100km, but they later decided that lower orbits were better -- presumably to reduce latency and/or the number of signals that might be contending for service at any given time. Anyway, they currently appear to have something like 3800 satellites orbiting in multiple shells at altitudes between 520 and 570 km. They do plan to use lower orbits for at least one later group.
From a signal strength point of view, there's not that much difference between 550 and 1150 km -- 6db I think. But half the latency.
Then there's the fact that it's only 550 (or whatever) km if the satellite is directly overhead. The further it is from directly overhead, the greater the distance
Anyway, I expect that the 2 or 3 watt output from a cellphone probably can make it 500 km or so if background noise is low and there's a satellite near overhead. But I wouldn't be surprised that you really did have to hold the phone just right to make a call.
And in the case of the Ukraine, I wonder how hard it would be for Russia to jam the cell phone channels.
Technical issue briefly grounds American Airlines flights across US

"American Airlines didn't address El Reg's questions as to whether resolving the issue could result in additional delays or disruptions over the holiday season."
Perhaps they ignored your question because they thought it was a bit dimwitted. Why would any 21st century American not know damn well that a system wide outage on Christmas Eve of a major airline for more than a few minutes is going to cause delays and disruptions over the holiday season?
One third of adults can't delete device data

Re: Working Out their Frustrations
"I usually use a hammer for hard drives. Be careful not to smash your paving"
If they still work, I generally use dd to zero them out then donate them to one of the local recycling operations. If not, I drill a hole through the case and platter then tie the platter off with a zip tie to keep it from rotating. Probably not as effective as a really thorough bludgeoning, but strangely satisfying.
Humanoid robots coming soon, initially under remote control

Re: "A lot of the humanoids we see out there are a hammer in search of a nail,"
Actually, there's a rationale for six legs. When walking, a critter or device with six legs can plant 3 forming a stable tripod while stepping ahead or back with the other three. Then, once those 3 are planted, move the other three. No need to worry about balance which, one suspects, is a non-trivial problem for a 2 legged or even 4 legged machine.

Re: Neo will be paired with a human teleoperator
Allowing for weekends, vacation, illnesses, and emergencies, you'll likely need 4 operators per device to provide 24/7 coverage. If the operators need to be familiar with the specific home environment as in, for example, where stuff is stored, where the trash containers are, where the fire extinguisher is, etc, etc, etc. You may need more.

Re: "A lot of the humanoids we see out there are a hammer in search of a nail,"
Same question I had -- why humanoid? Bipedal works for us humans, but it's not a very stable platform. Four or six "legs" might work better. And why only two arms, three or four might be better for some tasks? Can't count the number of times I've wished I had at least one more arm. And a tentacle or three perhaps with with eyes on the end might be handy for many jobs around the house.
Would I want one around the house? If I were temporarily or permanently incapacitated, perhaps. Otherwise, no, I don't think so. And in any case, odds are that household robots that aren't more problem than solution are likely several decades and many billions of dollars away from realization.
Some future generation's problem one hopes.
Guide for the perplexed – Google is no longer the best search engine

Re: This was predicted - in El Reg - years ago.
I disagree somewhat. As does the article BTW. Google's problem doesn't seem to be so much lack of innovation as letting the marketing folks take over the results page with somewhat profitable, but dubiously helpful to the user, sponsored crap. (Enshitfication as Cory Doctorow would have it).
Strictly speaking, that's kind of innovative. But it's perhaps also kind of dumb.
Trump administration wants to go on cyber offensive against China
Iran-linked crew used custom 'cyberweapon' in US critical infrastructure attacks

Re: The problemis
I should possibly have explained, but one can get only so much into one post. IMO, the problem isn't technology per se but rather the fact that it seems mostly to be built with way too much "Golly, Gee Whiz, Look at what THIS can do" and way to little "What could possibly go wrong?"
The result is all too often flawed attack surfaces stretching as far as the eye can see. We have what looks to be attack surfaces and flaws to last many decades already. The last thing the Internet needs would seem to be more flawed attack surfaces. You are free to think that more technology can somehow solve that problem without introducing as many exploitable flaws as it resolves. Maybe someone somewhere is smart enough to do that. It's not me. Or anyone I've ever encountered. No offense, but it's probably not you either.

The problemis
If you ask me, the problem is that too many people think that not exposing critical infrastructure is simple. Y'know what? It's not simple at all. It's hideously difficult. Quite likely impossible with all the features of the modern Internet in place.
I doubt the answer is more/better technology.
I suspect the answer if there is one is likely to be a hell of a lot LESS technology.
And I'm pretty certain that the answer is not AI.
Keep in mind. Not only is the internet becoming a very bad neighborhood. It's getting worse. And, guess what, On the internet, every scumbag, sleaze artist, and malicious national agent on the planet is your next door neighbor.
Cruise shutdown blastzone increases – Microsoft takes $800M charge

Re: Oh dear
FWIW a $0.09 EPS hit looks to be about 60 hours worth of profits from the Microsoft empire. Significant, but genuinely it does look to be but a scratch.
The numbers I used are $0.09 write off from the article, quarterly EPS = $3,30 for the quarter ending Sept 2024 source -- the all-knowing Internet. I certainly might have that wrong. It's late. It's been a long day. I'm 85 years old. I never was all that great at math. ... And I really don't much care. Feel free to correct me if you have nothing better to do
Blocking Chinese spies from intercepting calls? There ought to be a law

CAMS
Advertising spies? Heavens no. They are simply Computer Assisted Marketing Services (CAMS) -- Tiny cogs in the wheel of marvelous technology that drives the wonderful (Donald Trump's word, not mine, I'd say instead "wildly overextended") American Economy.
You Donna mess with the CAMS or we breaka you hands.
Microsoft dangles $10K for hackers to hijack LLM email service

Why?
"This simulated service uses a large language model to process an email user's requests and generate responses, and it can also generate an API call to send an email on behalf of the user."
Sounds like a mechanism for machines to babble endlessly with each other with no human oversight or input whatsoever.
What would it be good for? As far as I can see, absolutely nothing.
British boffins build diamond battery capable of working for a millennium or five
Microsoft teases Copilot Vision, the AI sidekick that judges your tabs

It's Incredible
Microsoft management? Am I the only one reminded of the mythical Bob Parr, cartoon hero of The Incredibles. "They keep creating new ways to celebrate mediocrity..."?
This sure sounds to me like the dumbest idea since Clippy. At least with Clippy, the creators had the excuse of not knowing how annoying an intrusive AI assistant can be. Now, two decades later, we know. But they persist anyway.
Microsoft: Another Chinese cyberspy crew targeting US critical orgs 'as of yesterday'

Maybe it's time to rethink this Connected World thing.
It seems to me that the problem is simply that, for uses beyond entertainment and everyday commerce, the internet simply is not very secure. Further, contrary to the belief (more accurately -- forlorn hope) of way too many, it seems likely that there is no simple fix. If there were, it would have been implemented 25 years ago. Encryption and two factor authentication have their place. But they really are not universal solutions -- especially for the physically handicapped, those who lack reliable cell phone coverage (lots of us in rural America) and not loaded on top of the truly awful user interface on smartphones where they are likely to be a substantial usability issue for many of us.
Truth of the matter is that if you are going to behave as the worlds' Protector General (liberals and many moderates) or the world's biggest and toughest bully (many conservatives) you simply can't put command and control of critical infrastructure on today's internet. Sooner or later you're going to annoy someone into retaliating, And retaliation via today's internet is quite inexpensive. Not to mention the ransomware problem.
The only solution I'm aware of is that used by the military and some parts of government three decades ago when last I worked in that world. Physically completely separate installations for secure and non-secure work. Even that isn't secure if an employee leaks data or engages in sabotage -- but that's a different issue. The trouble with dual networks/critical data partitioning/etc is that it's quite expensive. My guess is that in many/most cases, the costs of operating that way would exceed any savings from operating online.
I'm sure I'll garner a bunch of downvotes from those who don't want to hear what I'm saying. But kindly do me a favor, If you disagree, by all means downvote. But also compose a brief reply and explain why you disagree.
Google DeepMind touts AI model for 'better' global weather forecasting

Re: Hang on a minute
The reason they do not give you a confidence interval is probably that they are not doing statistical modelling and therefore generally don't have much quantitative idea how good their forecast is. However I'm 70% confident that AI will be happy to make up a number and report it to you. e,g, "There is an 83% chance that a tornado will carry you off to Oz between 1400 an 1600 this afternoon" (Which is why it might be a good idea to keep the works of L. Frank Baum out of your AI's training materials).
Judge again cans Musk's record-setting $56B Tesla package

Re: Musk should quit fighting this
As I understand it, the agreement between Musk and Tesla depended on Musk meeting some performance and valuation goals that were, at the time of the agreement, thought to be somewhere between wildly aggressive and impossible. Somehow Musk did it. Seems to me that he certainly deserves a generous reward. Heck, Judge McCormick, says that in her opinion. This case isn't about whether Musk should be paid a lot by Tesla. It's looks to be about whether the agreement between Musk and Tesla and the subsequent shareholder vote met the minimum legal requirements for a corporation in corporate friendly Delaware. McCormick says they didn't. I wouldn't be at all surprised that her opinions stand up to the inevitable appeals.
Tesla still would seem to have a moral obligation to pay Musk a lot of money for his efforts. Since that will presumably be negotiated in Texas under that state's legal code (assuming they have such), I imagine Musk will make out OK in the very long run.
Google India probed after driver fatally followed Maps route over unfinished bridge
Chinese boffins find way to use diamonds as super-dense and durable storage medium

I think straight ASCII text and ALGOL math notation might be good for more than a few decades. Probably Unicode as well. We can, after all, read many 2000 year old and older documents. The problem for the most part isn't the content, but deterioration of the media. Heck, in High School I was forced to spend an unconscionable amount of time reading Ceasar's commentaries on the Gallic War in the original Latin. In retrospect, Spanish would have been more useful.
But you're right, Microsoft er. al. will undoubtedly do their damndest to deprecate anything simple that actually works. They seem to be rather better at that sort of thing than at quality control.

How many parity bits?
"How many parity bits are you going to need ..."
I think Hamming discussed that in "Numerical Methods for Scientists and Engineers" back in the early 1960s. My copy went missing about five moves ago in the early 1990s. I don't recall the details very well, but I think the answer might be surprisingly few check bits are needed for error correction even after allowing for the check bits sometimes being wrong.
Someone around here probably knows the answer off the top of their head. Maybe we'll hear from them.
Microsoft preps big guns to shift Copilot software and PCs

Re: Great!
"The morbidly obese orange clown's Tariff fiasco will make the Covid disaster look like a hiccup."
Most likely. Given that Economics is even less comprehensible than quantum mechanics, it's hard to be certain, but tariffs like that proved to be a terrible idea 90 years ago during the Great Depression -- triggering retaliatory tariffs ,cutting international trade dramatically, and worsening an already already awful world economic situation. One doubts they will work any better today. But the Collection of Crackpots Trump has assembled for his Great Leap Backwards will probably encourage him to try the tariffs anyway. Fasten those seatbelts kiddies it's probably going to be a bumpy four years.
Cloudy with a chance of GPU bills: AI's energy appetite has CIOs sweating

Another Alternative
Let's say you run IT for modest business endeavor. 500 desks plus another 50 PCs doing various weird and wonderful things that someone once thought would be a good idea like running a camera in the main lobby during off hours and saving the images ... and some of them actually are a good idea. Maybe even mission critical. Let's say that a routine desktop PC costs you $2000 while an AI capable PC will run $6000.
Now then, you'll likely not upgrade the 50 special purpose boxes because you know that on too many of them new hardware will mean they won't work, and debugging the problems machine by machine is costly even though many of the problems will be trivial.
That leaves the 500. Hardware costs alone for upgrading the 500 to AI capable will be 500*(6000-2000) = $2,000,000. That's a tidy sum probably not in your budget which likely assumes $2000 per desktop every three years -- $333,000 annually. Then there's the electricity cost. And likely a number of other costs. Yes, possibly you need to be concerned,
But why not treat the AI things like those other expensive office devices/nuisances -- copiers, high volume printers, and such? Buy/lease one or two. Make them available on the office network. Monitor usage. See who's using them and for high volume users find out why. Maybe even give those folks an AI capable desktop if that's what they actually need. Given any luck at all, you may find that short term you only need a handful of AI capable devices. If that. Longer term? With any luck at all, you'll be retired and playing golf in some sunny spot long before everyone needs AI on their desktop.

Just say no
Letting others go off and squander time and money on a quite dubious "technology" while you stodgily continue to do whatever it is to do the way you've always done it, is a real alternative you know. True, in the best of all possible worlds, you'd invest a bunch of bucks and catch the AI wave and ride it into a glorious future. But isn't more likely that you'll misjudge this rather nebulous thing, get pounded into the bottom, and come up spitting salt water and wishing you'd never heard of AI. After all, if AI actually works out, you can probably buy AI as a service from somebody while you catch up avoiding a whole lot of pitfalls that others have found out about the hard way. More likely than not, you'll end up money ahead and save yourself a whole lot of grief.
Australia passes law to keep under-16s off social media – good luck with that, mate

Re: "The Voices of Young People"
"Guess what? Fifteen-year-olds-and-under don't get to vote, either."
True enough. Here in the US there is a rather elaborate procedure to make sure that only those eligible to vote are allowed to. Typically, it involves showing officials proof of residency, age, and citizenship. You seriously expect Social Media sites to somehow replicate that over a digital link. I'm no fan of Social Media and I think the world might be a better place without them, But many folks do seem to find some value there. Not mine to question why. Even though I dislike them, I have to think that laying a probably unworkable requirement on them is kinda dumb.
If you ask me, the world has a more than adequate supply of dumb already. It probably doesn't need more.
China starts building world's largest fully steerable radio telescope

Very Interesting
I have no idea what it's good for, but it seems like a remarkable piece of engineering. Probably stunning to watch in action. I am curious why they didn't use a phased array -- a bunch of smaller antennae that do not themselves move, whose beam is steered electronically by altering the phase of the signals from/to the individual antennae. Not questioning their judgement, just curious about the reasoning.
US senators propose law to require bare minimum security standards
China has utterly pwned 'thousands and thousands' of devices at US telcos

Re: That explains a few things
Yeah, that was me. We were having a beer break at the listening post and someone knocked a jug of brew over. And while we were mopping up I inadvertently flipped a few switches. And ... Well, I guess I can't go into details. Anyway, sorry about the inconvenience. Would it make it right if we fixed you up with a video and text tap on the largest brothel in Tijuana? No cost to you. We already did that for the dude who came on your line.
Imagine a land in which Big Tech can't send you down online rabbit holes or use algorithms to overcharge you
China sends cloud powered by homebrew Loongson CPUs into space

Different chips for different missions?
"The Chinese chippie's wares are modest and are a few years behind rivals like AMD and Intel, but this announcement nonetheless suggests Loongson silicon is ready to be used as a space-based cloud platform."
Project management in China, the US, the Grand Duchy of Fenwick, or anywhere else will probably feel some pressure to use local products in space going gear. Also, I seem to recall from reading and informal conversations with assorted folk that one doesn't particularly want to use the latest and greatest technology in space and other stressful situations except where absolutely necessary. Better to use technology with a few years track record. There's probably also the issue of cosmic rays and other ionizing radiation which I would think likely to be more of a problem with smaller feature sizes and lower operating voltages of state of the art chips.
Or maybe I'm a victim of old age and overactive imagination. Anyone who actually knows what they are talking about care to straighten me out?