Re: While Anubis is a good concept
Scruples? of course he AI dudes are familiar with them. Some sort of ground pork and maple syrup concoction popular in Philadelphia -- Right?
2323 publicly visible posts • joined 13 Sep 2017
Putting a nuclear reactor on the moon seems well within the scope of current technology. Making a 100KW reactor work on the moon once it gets there seems a bit more challenging. We're talking a heat engine here. Where is the cool side going to dump the excess heat? Space? I'll take a pretty big radiator to dump 100KW to space I should think. And the radiator would need a sun shield? It'll be interesting to see what, if anything, they come up with.
I imagine human fighter pilots will be around for quite some time, albeit fewer of them as one assumes that over time more and more missions can be trusted to unmanned craft. Depending on the future of 100% reliable control links, it's also possible that even some of the missions requiring human control might be executed using unmanned remotely controlled craft. Human pilot, but he or she is sitting at a console in an undisclosed location.
Seriously, I doubt they'll build anything in one year other than maybe a mockup and perhaps some test jigs. We're talking nuclear here. The real reactor probably needs some specialized parts made from exotic materials. That likely means long lead times. And they probably can't order the parts until the design is locked in.
Now, if plant operators have the safety culture of the Rickover NNPP we'd in good shape.
Yes, you're right. And some of them might be. But really, you have to know that if the world is paved with small nuclear reactors, most will be run by bean counters and MBAs. And those guys will no doubt blow the damn things up with dreary regularity.
Marketing speak:
"“In the long run, while tariffs will pose certain challenges, they also provide us with opportunities to accelerate the optimization of our layout for the global supply chain,” she said. “As we expand into new markets and serve a wider range of customers, I believe this presents a challenge and also creates long-term development opportunities. Overall, we are confident that we can transform this challenge into a competitive advantage.” ®"
English Translation:
If the idiot Americans want to build an iron curtain around their country, we'll just sell our chips to everyone else.
Downvoted because sea level rise in the 20th century is estimated to have been around 20cm (8 inches) and there is no reason to believe it has accelerated much if at all in the 21st. (You can check that by going to https://tidesandcurrents.noaa.gov/sltrends/sltrends_station.shtml?id=8518750 See a frightening trend there? There isn't one. By all means, check other us tidal gauges (hint, The Presidio tidal gauge in San Francisco has the second longest US record. But be aware that you want long term records because there are definitely short term (e.g. decadal) variations in sea level at any given location. And remember that older tidal gauges were primarily intended to tell cargo ships how much water was under their keel. It's likely that some of the records include effects of repairs or even moving the recording site around the harbor possibly to locations with different local tectonics
That doesn't mean that rising sea levels won't be a problem for some over a very long time span -- centuries. But ironically probably not for Pacific Islanders. Why not? Because tropical reef corals make limestone from CO2 and Calcium molecules in the tropical oceans. And they do it at impressive rates. Tropical atolls seem to have managed to keep up with 120 meters (400 feet) of sea level rise during the continental ice sheet melt 20000 years ago to 10000 years ago. That's six times the 20th Century rate. Where will the problems be? Probably places built on fill that is compacting(many of the world's port facilities), places where fluids (petroleum, fresh water) are being pumped from under the harbor, places where the Earth is sinking due to natural tectonic forces.
I'd recommend the IPCCs Assessment Reports sea level sections. They seem to be quite good for the most part. But they are information dense and are unlikely to give you the sort of quick, easy answers the media, politicians, and the public in general favor. Here's a link to AR6 (the latest) https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-9/ If you want to tackle it.
Do data centers on a bunch of volcanic rocks 2500km from Australia make any economic sense? Bear in mind that the islands,being volcanic,have little or no fossil fuel resources. Electricity for those Islanders that have access mostly comes from imported diesel. What percentage of the population has power?According to the internet, about 33%.
Nothing against Vanuatu. Might be a tropical paradise for all I know. But it seems an odd place to locate data centers.
Similar experience here. School in a small rural town. Service provider, the district offices in a larger town about 7 miles of ancient telephone system distant. It turns out that due to the fragmented nature of US telecoms, messages from the school to the service provider went to the SP's town and kept right on going. To Burlington, VT then Boston then to New York then back to Burlington, to the SP's town and finally to the SP. Replies took the same route in reverse. Lots of latency. And a router poorly configured somewhere along the way helped out by black-holing large message packets. Result: Novell worked mostly. As did some other stuff. A bit lethargic at times. Citrix,not so much.
Well, if the Gartner Guy is right, and likely he is this time, using virtual desktops will save many companies money. But the security and reliability issues ... Wow. And then there's that "The first fix is free, once you're hooked, the price goes up" thing.
By a curious coincidence, I was pricing tulip bulbs on Amazon a couple of hours ago. They're mostly going for over a dollar a bulb this year. If memory serves that's a bit more than last year. Looks to me like a better long term investment than, for example, Tesla TSLA which closed today at 339.38 giving it a price earnings ratio of around 200. By way of comparison, the long term average PE ratio of the S&P500 is around 15. Perhaps a bit of irrational exuberance there.
Money laundering schemes? Closer to Ponzi schemes I should think. It appears to me that most of the "money" sloshing around in big tech is in the form of IOUs of dubious quality. It seems to me entirely too likely that the motto of the 2020s will turn out to be "Think Big, Fail Big"
So humans infringe copyright constantly.
True enough. But in the US at least, most of that usage is covered by the (legal) doctrine of Fair Use. See https://www.copyright.gov/fair-use/. One does wonder if fair use doesn't apply to AI agents as well as us humans. One also wonders if those "If we screw up and you get sued, it's your problem not ours" clauses in T&Cs are legally binding. Especially if one paid the provider to provide the infringing material.
As a result of a multitude of bad experiences with modems (and, to a lesser extent, DSL) in rural locations in Vermont. I ended up writing some articles for the Compuserve Hardware forum. Two in particular seem appropriate to this discussion http://donaldkenney.x10.mx/GLOSSARY/LINEPROB.HTM and http://donaldkenney.x10.mx/GLOSSARY/BLACKHOL.HTM
My feeling at the time? I reckoned that, all things considered, it was something of a miracle that modems worked at all with the US telephone system
Let me start off by saying that I think AI probably will have a huge impact on the world. Eventually. Decades from now.
However I strongly suspect that it is one of those problems like (safe) fully autonomous driving in an arbitrary setting or scaling hydrogen fusion back from city busting to power plant dimensions that is, while probably doable, going to take MUCH longer to mature than most people think.
Case to point. IBM's Watson. Billions of dollars invested. Scored a truly impressive performance on a TV trivia game where occasional spectacular blunders are only mildly punished. But after Jeopardy it has had trouble finding gainful employment. IBM Watson has few users. Numbers? A hundred odd worldwide. 32 in the US. Hardly world shaking.
There is a persuasive case that Watson has been a very expensive liability for IBM. Billions lost. See https://slate.com/technology/2022/01/ibm-watson-health-failure-artificial-intelligence.html
What does that say about AI? Very likely something along the lines of "Best fasten you seat belts folks. It's going to be a bumpy ride." (And yes, that's a misquote of Bette Davis in All About Eve ""Fasten your seatbelts, it's going to be a bumpy night".)
English does have a few problems. For example, it does not distinguish between inclusive and exclusive or. It also lacks a convention for evaluation phrases like A and B or C. Is that A and (B or C)? Or is it (A and B) or C? And, unlike some other natural languages it is a bit hazy about negation. The average American English speaker will interpret "Isn't the door open" as equivalent to "Is the door open?" and will respond "Yes" if it's open or "No" if it's closed.
There may be some other problems. Those are just the ones I'm aware of.
Big assumption there that it eventually passes the tests.
Even bigger assumption -- almost universally overlooked -- Someone can tell the AI exactly what needs to be done. Writing code is hard, but is sometimes done well. Specifying exactly what needs to be done is harder and in my experience is almost always done poorly. Even if your AI agent can grind out working code my guess is that most of the time that code won't do what you really want done,
What would be needed for the future tech CEOs envision would be AC -- Artificial Clairvoyance. I expect it might be a while before that is claimed, Much less delivered.
Understand the comment. And there's some truth in it. But in this case, they are hiring a contractor to buy the vehicles, and apparently haven't put much in the way of restriction on how they are procured. If it's a fixed price contract, I would guess the trucks delivered will be well used and a bit banged up.
I disagree - I'm pretty sure they will not fulfil any of the pledges
Most likely, some will, some won't.
One possible strategy. Announce a zillion dollar investment in (a) US plant(s). Send a few lawyers and suits off to talk to various states about incentives to build there. That'll take about a year. Maybe 18 months. Then announce you're going to build in West Virginia or whatever and are looking for a suitable location. Another year to 18 months. Buy some land conditional on approval of permits to build. Put together an outrageous plan and submit it. Permits denied. 3.5 years have passed, By then, the Great Leap Backward has floundered. Trump is out of office. The US economy is collapsing thanks to Trump's antics. You walk away. It'll cost you a bit, but not a lot. ISTR that Foxconn did something along that line in Wisconsin a few years ago -- perhaps not as cynically. but same result.
And if somehow, Trump doesn't crash the US economy, you can always submit a rational plan and build the damn plant if it makes business sense.
I've worked in a school. It strikes me that giving Chromebooks to students will provide a steady market for replacement devices. Not only is there a steady flow of new students and graduation of those who have grown too big for the desks, but kids tend to be a bit careless at times. If kids can and do manage to forget their coats when the temperature is hovering around -10F (-23C) outside you can guess at what the MTBLD (Mean Time Before Loss or Destruction) for a Chromebook is likely to be.
In the case of Google Search, the old non-AI search was pretty good, and I assume it still works as I don't always get an AI summary for search requests. So I reckon that shutting down AI every now and then won't affect search all that much. At least for me. And likely for most users. Of course I/we don't know how much resource the old search takes. The energy savings may not be all that great.
fraudulent stock pumping lies...
Indeed. I've long maintained that Musk's claims for his vehicles' capabilities are irresponsible and dishonest. They seem to go way beyond "puffery" -- e.g. "The Best Frozen Pizza In the Universe" which is permitted in American law. IMO they endanger both the occupants of his vehicles and innocent bystanders.
It's good to see a court hold his company responsible.
But as a test case, this really seems to suck.
Most likely the wasps were experimenting with some sort of nuclear weapon. And what better place to do so than a site where nuclear waste is readily available? Since the wasps seem to have abandoned their lab, we can assume that they have perfected whatever device they were developing and have moved on to production at a larger and more suitable facility. Doubtless all this will be made clear in the not too distant future.
"Were they on the path to going bust anyway?"
That was my thought as I read the article. Sounds like their basic problem was that their operation was chronically(?) losing money thus leading to insolvency. The cyberattack was only an added aggravation? Seems to me that if they were even close to profitability the instinct of 98% of the managers in the world would be to blame the cyberknaves for all their problems and to promise to come back stronger than ever.
"Puzzled why anyone would put an Apple product and cheese inside a microwave oven ?"
I'm told that the glue Apple uses to fasten their non-user replaceable batteries into the cases adds an unusual and quite unique flavor to some microwaved delicacies. As for cheese. A day without microwaved cheese is like ... well ... ehr ... a day without microwaved cheese.
it's probably just that I'm getting old, but more and more it seems to me that I'm living in a distopian science-fantasy novel written by a master of dark humor.
First off we have the EU. A modern industrial society which came up short in the great geophysical fossil energy resource lottery. They currently need to import fossil fuels from places no one in their right mind really wants to import energy from. Russia, the Middle East, North Africa. None of them exactly reliable suppliers. Their response is to embrace the theory of a probably imaginary climate crisis and the improbable proposition that "green" sources -- primarily wind and solar -- can provide all the energy they need. There appears to be no way that is viable without economically acceptable long term -- months, years -- storage of massive amounts of energy. At this time, that long term storage does not exist. It may not for decades, or maybe centuries, or maybe never.
Anyway, the EU genuinely does need to be concerned about energy usage even if their motivation looks to be muddled.
Then we have the tech industry which has gleefully embraced the peculiar notion that you can take huge globs of information, misinformation, disinformation, satire, and outright lunacy, throw it into a digital blender and produce a high quality, salable product. They claim to be convinced that that product will enrich mankind. One suspects that they live in somewhat egocentric universes where mankind's wealth and their own personal accounts are inextricably mixed. What's good for Sam Altman etc al is good for humanity. They think they need bizarre amounts of energy to actualize their dreams. They may well be right -- at least about the need for energy. Enrichment of mankind? I have a few doubts.
And don't get me started on the growing tendency for democracies to elect truly rotten human beings to positions of power.
Reminds me a lot of the 1920s as described by Fredrick Lewis Allen in "Only Yesterday."
The boom years of the 1920s did not end well.
Latency ...
Indeed. If you put your data center in Low Earth orbit the latency isn't too bad. But the satellite is only visible from a given location for a few minutes every now and then. If you put the data center in geostationary orbit, it's always visible from the same roughly 40% of the Earth's surface. But the round trip latency is around 600ms.
Actually Windows-95 wasn't all that bad -- After one installed about two dozen "Service Packs" (Service Pack=Massive collection of bug fixes). It was actually comprehensible to almost all users and could run for weeks without crashing. Of course it was really MSDOS 6.22 with a GUI shell. We have since devolved into the incomprehensible shambles that is Windows-11
Au contraire. Tesla looks to be an OK EV. Which is to say it's a rather expensive vehicle that takes rather a long time to fuel, works best for folks who have a place to charge it at home and live in mild climates. And it is overly computer dependent with an erratic sysop who has a rather loose attachment to reality and at times looks to be a few sandwiches short of a picnic.
Other than that, I suppose it's OK as long as you don't use its notoriously flakey collision avoidance system and drive it as you would any other car.
I'd suggest that the robots on 2125 probably will be ubiquitous. And very useful. But they likely won'i be humanoid. More likely a central chassis with sockets for appendages -- legs, arms, tentacles. And specialized tips for those appendages -- viewing, gripping, cutting, welding, etc, etc, etc... And many will swap appendages frequently as they work through a task -- observing, testing, analyzing, and finally performing the task and verifying the results. Y'know, that's WAY, WAY beyond anything we can do today. Or next year. Or the year after.
I suspect that the major near term use case for Optimus and its cousins might turn out to be modelling the latest styles in upscale clothing stores.