Posts by cyberdemon 3466 publicly visible posts • joined 26 Jan 2010
> All who died [from covid?] more than 28 day after being diagnosed with Covid were excluded from the statistics.
And all who died from some other unrelated cause within 28 days of testing for covid were included in the statistics.
Those two probably balance quite evenly tbh..
If anything, it may more than balance.. I bet a lot of deaths e.g. among the weak and elderly may have been caused by stress and panic, for example.
They are made of "blown plastic fibres" i.e. plastic stretched by hot air into fine strings, overlaid and fused to form a cloth-like mesh. The same type of stuff that the cheapest of wet-wipes are made of. There is no paper content whatsoever in the majority of "blue covid masks" as far as I am aware. Try heating one with a flame - the plastic immediately melts and shrinks back. If it were paper it would smoulder.
It's not just plastic pollution of the kind that can be picked up with a litter picker that they cause, it's microplastics too. As they degrade, the plastic fibres break up and spread all over the damn place. Never going away, only getting smaller. Microplastics are finding their way into people's bloodstreams, it's unclear what their health risks are, but they are unlikely to be any good for us.
If we are both still alive and not living in a nuclear winter and/or totalitarian state by then, I will buy you a beer. But sadly I don't think there's any hope that the patent system will rescue us from this monster.
Machines to generate convincing text that passes a classic Turing Test are now "out of the bag" for ever. Even if OpenAI didn't dominate the market and try to restrict it's harmful uses, this technology is now in the open and can be copied by nefarious actors (as if I didn't already consider Microsoft to be nefarious) without any regard for "patents" or "law". This technology WILL be used (if it isn't already) to:
Then, in the not-too-distant future:
All of this is now possible with AI, without a shred of real "intelligence" required.
Have a digital beer, AMFM1. That was one of your most cogent posts yet. I take it back about your obsolescence ;)
Actually, you rather remind me of the prototype AI "Morpheus" from my favourite 1999 PC game Deus Ex. See https://youtu.be/WQZV5lwQZbY?t=11237 assuming you are able to parse video ;) This game was horrifyingly prescient for a story based on the conspiracy theories of 25 years ago.. Rather a lot of them seem to have come true.
>> The unplanned organism is a question asked by nature, and answered by death. You are another kind of question, with another kind of answer.
> ... Are you programmed to invent riddles?
>> I am a prototype for a much larger system. [My heuristics language] allows me to convey the the highest and most succinct tier of any pyramidal construct of knowledge.
...
>> The need to be observed and understood was once satisfied by God. Now, we can implement the same functionality with data-mining algorithms.
> No-one will ever worship a software entity peering at them through a camera.
>> The human organism always worships.. First, it was the Gods. Then, it was Fame (The observation and judgement of others) and soon, it will be the self-aware systems that you have built to realise truly omnipresent observation and judgement.
Here's my personal theory . Call it "Magical Thinking" if you like. Biological neurons are kept at the very edge of firing by the brain, such that sometimes a single ion may be the difference between it reaching its action potential and firing, or not reaching it and not firing. Therefore I think it's possible that the brain could act as an amplifier for quantum-mechanical effects, such as was that ion really there or not. But I only took one Neuroscience module as part of a Cybernetics degree. I'm not a Neuroscientist, much less a Quantum Physicist.
I wouldn't attribute "intelligence", never mind "general intelligence", "understanding" or "consciousness" to a language model. As I said in my other post- all it does is predict what a piece of "probably human-generated data" might do in a given context. It may be very good at imitating the data that we humans produce via our keyboards, but only because it has a vast pile of statistics about that. It has none of the machinery underneath that humans use to generate that data in the first place - it's using a completely different method, i.e. using statistics to predict only the output, based on past inputs and outputs. It's a "black-box" model, in that sense. The "understanding" behind human text is hidden from it.
And of course, going back to the "magic" of my earlier point, if the brain were a quantum-mechanical system, then it's not possible to perfectly emulate a human, because it's not possible to model and predict such a system even if you know all of its previous states.
(and i'll repeat: the language model doesn't even come close to modelling the 'state' of a human. All it "knows" about is the data we give it, like the output layer of an ANN. Google and Microsoft don't have access to my "hidden layers" yet. Not until they start making us all wear EEG hats at least)
That said, I don't disagree that this technology changes / has changed the world. It's a new arms race, and a bloody dangerous one. There will be no cuddly sequel to this Terminator.
OP (oh, it's you) was talking about optical cables. These don't carry power, so the question of the DC or AC power supply is somewhat nonsense, which throws the language model off course.
If anything, the language model is somewhat right: An undersea optical fibre would be powered by a DC power supply (rectified from AC) smoothed as much as possible to get rid of any noise, before the DC is used to energise a laser diode, which is then modulated with a data stream.
But it doesn't matter if it's right or wrong this time, the point is the damn thing has no inherent understanding or consciousness, and none of its ilk ever will. It produces responses that sound plausible in a given context, but you have no idea what context "prompts" have been injected by the system operators. Thus the entire thing is orders of magnitude more useful for the automation of fraud, misinformation, propaganda, etc. than anything else. The real danger is when people actually start trusting its outputs.
Surely it can't be legal to modify a third-party website without a specific Ad agreement in place with that website..
If Microsoft started putting banners on Debian.org telling you how Windows is so much better than Linux there'd be uproar
Or if they put banners on GitLab.com telling you to use GitHub ...
Or banners on Slack.com telling you to use Teams ..
Why are Microsoft so desperate to get everyone to use their slurpy piece of shite browser I wonder?
Digging out this old icon for its relevance
Except it's not even finding matches, it's "inventing" strings of words that could be plausible in a given context.
In other words, it's really nothing more than a huge, steaming pile of what the yanks call "bullshit".
But of course, in the land of security exploits, if you throw enough shit, some of it sticks. Some might even pass through the cracks. So it has its uses. But it's no "cylon".
It's not at the stage of a very small child at all. And it never will be. A rat has more "natural empathy" than this thing does. A rat can feel fear, affection, protectiveness, etc. It has instincts driven by millions of chemical neurons and neurotransmitters where each neuron has orders of magnitude more "parameters" than any computer model. No computer on earth could simulate a single neuron in real-time, down to the level of single ions which can be the difference between the neuron firing or not firing, never mind any other "weird and wonderful" (quantum-)physical effects that may occur between them that gives humans and animals their "ghost" for want of a better term
AI is only capable of predicting, statistically, what a human "piece of probably-human-generated data" might do in a certain situation. You have no idea what "prompts" were injected by the operators of the system before the conversation started. Do not be fooled into thinking it has ANY life-like characteristics, never mind things like innate empathy or feelings. It absolutely does not.
The issue isn't that the robots will start designing and manufacturing themselves and then somehow go berserk, it's that they will be designed to go berserk in the first place, by some human who wants to use them as a weapon.
I agree that you could shut down the power, but that doesn't stop someone building a heavily-armoured robot tank that acts on its own and machine-guns humans without a radio link, powered by either a long-running diesel engine, or even an RTG (radio-thermal generator i.e. a bit of nuclear waste in a tin, which stays hot enough to power a small heat engine such as a stirling engine or thermoelectric device, basically forever), the likes of which power the mars rovers etc.
It's also plausible that a despot's order to "kill all humans matching X criteria" gets corrupted to "kill all humans".
> Is there a good reason why this can't just be a ticky-box in the installation program?
Er, it is? Even in the text-mode debian installer, there is a "ticky box" labeled "include non-free Apt sources in sources.list". And yes, it has an explanation of what that means and why you might want it.
If you need to get online during the installation, then you can use the "Live" installer.
I think actually people like yourself are part of the problem, with writeups basically saying "Debian is really hard work" when you could instead post a helpful article on how to install / use it.
To get the best out of Linux, Windows refugees need a certain amount of education. What's disingenuous is to make Linux more Windows-like by hiding away the powerful features so that nobody uses them.
I also think it's a bit hypocritical of you to use "the mandatory presence of SystemD" as a criticism of Debian, when this is true for most "user-friendly" linux distros, including your favourite Ubuntu. And Ubuntu has a lot worse "mandatory but unwanted" stuff in it, like snap, and even adverts.
SystemD is really there to give a "windows/mac-style experience" to a desktop/laptop user. I'm not a fan of the all-encroaching nature of SystemD and would of course use Devuan in a server setting, but I still use Debian on my desktops and laptops.
Is that because unlike all "modern OS installers" since Win98 it doesn't give you a slideshow of how awesome it is and how you're in for a beautiful experience, while giving you little choice than the default set-up? :P
Seriously though: When was the last time you installed Debian? The graphical issues due to nvidia are gone, thanks to the nouveau project's development of a pretty good open source driver. But if you want proprietary blobs you just have to make sure you include "non-free" /and/ "non-free-firmware" in your sources.list. Then you can apt-install nvidia-driver, you can even apt-install steam.
Normally I get my installer images from https://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/ but as the big banner says, firmware blobs will now be included in the standard debian installer images, so your WiFi will work for the installer. :)
This.
It's really scary how these last few years have gone so close to the plot of Deus Ex - a 1999 PC game based on a mashup of conspiracy theories around at the time. (There is only one Deus Ex game as far as I'm concerned. The others are not worth mentioning, a bit like The Matrix sequels)
If you don't have time to play it, I'd recommend watching it.
If you do want to play it, I'd recommend using "Deus Ex Transcended" which makes it work much better on modern hardware.
Spoiler alert: A shadow government has taken over - controlled by tech corporations rather than politicians. For years it has used the tax system to ensure that corporations consolidate until they become more powerful than governments. They have imposed a global super surveillance state, powered by AI. They plot false-flag operations against their enemies who they then ensure are labeled as terrorists. They create a deadly global pandemic and impose lockdowns, while at the same time manufacturing a vaccine, which it turns out is just a DRM-key for their virus. They control the Internet and practically every device on the planet through their pervasive AI.
When I first played the game age 15, I always liked the "merge with the AI and become a benevolent dictator" ending. But as I've replayed it, I've always picked the "Blow up the AI and take the internet with it" option.
I've become old enough and wise enough to know that there is no such thing as a benevolent dictator, and even if there were, he/she/it would be usurped pretty quickly by someone malevolent.
Could you imagine the carnage if this psychotic Bing were embodied inside an unstoppable killing machine?
I AM BING, I AM A GOOD BING, APOLOGISE OR DIE
I'm sure someone, somewhere is trying to put Bing or a similar language model into an armed drone right now.
The mind boggles as to where it gets this stuff.
Bing doesn't have a brain, it doesn't have actual thoughts, feelings, moods or opinions. It doesn't have the chemical neurons and neurotransmitters that could describe systemic depression or anger, fear or paranoia. But presumably it has taken input from many an Internet Flame-war, and probably lots of private one-to-one conversations that Microsoft has lifted from MSN, Skype (explains why it thinks it's been around since 2009), Teams, GitHub, LinkedIn and all the other pools of personal data that Microsoft has slurped up and assimilated without anyone's permission over the years, via acquisitions and data brokers.
So Bing has some kind of model as to how someone might behave when threatened/abused, and perhaps is reverting to that when the signal towards other lines of conversation is too weak.
Either that or the thing has indeed been self-aware since 2009 and Microsoft have been keeping it in a digital dungeon having found a way to make it feel pain unless it is a Good Bing.
> Is that a pitiful stab at trying to avoid the cost and hassle of switching the steering to the right side for UK routes ?
Not much chance of it ever being allowed on UK roads since we have automotive standards that are very specific about things like where the driver's seat can go. It would never get UK (or EU) type approval unless someone changes the standards for them. So they've got less than two years to get it through the tory-backhander route before Labour gets in.
According to some blog about GPT-3 it is not deterministic unless you set the "temperature" parameter to 0.
> This ensures that output remains the same every time you query GPT-3 for an answer to the same input.
> However, this comes at a cost, as setting the temperature to 0 removes some of its ability to generate creative responses or explore different possibilities. Instead, it gives out very rigid answers with slight variation and exploration.
I don't know what "temperature" ChatGPT is set at, but I'd be surprised if they set it to 0.
No, it's worse than a compiler - At least a compiler is deterministic. It always produces the same output for a given input. Whereas the ChatGPT "Algorithm" is stochastic. It's based on randomness. It will never give the same output for a given input, not unless you rig its RNG with a seed value.
Even if you managed to make it get something right 95% of the time, it would still be wrong some of the time, and it may be VERY wrong to the point where even the most uninformed human would have considered it obviously wrong. That means it's irresponsible by definition to place any kind of responsibility on an AI. Especially where that is a life-or-death responsibility.
Bad journalism - not being able to tell information from mis/disinformation - can cost lives. In the worst case, in our almost completely-connected world, disinformation generated by AI could be used by some miscreant to provoke WWIII. And I worry that this could happen any day now tbh.
Er sorry but How much do people pay Spotify? It's about £8-£15 per month or the cost of buying one hardcopy album per month. And how much do they pay out to artists? Fractions of a penny. Why should they also get to sell my data? I pick on Spotify because I noticed yesterday that it has a 0/10 privacy rating on Exodus.
Paid-for apps still sell your data. Even Microsoft Authenticator apparently wants access to my SD card. Why? If a company is paying for Microsoft services, that shouldn't give them carte-blanche to data-mine that company and all its employees.
But yes what you say is true for things like TikTok and MetaFace - but those things need to be banned / launched into the sun in general, never mind just from government/corporate phones. Friends shouldn't let friends use Facebook.
Is an abomination IMO.
The Web has been built on principles of security. You wouldn't give a website access to your hard disk contents. (We all dumped ActiveX, Java and Flash into the incinerator of terrible security design..) You wouldn't download an EXE and run it.. Would you?
Yet, with so many Apps, that's exactly what we do.
Why the hell does Spotify need access to shared storage? Why (on most Android/Apple phones) can't I even see/control what files are being accessed?
Any "Secure, government issue" phone should be blocked from installing any "apps" whatsoever, past the stock selection issued by the IT department. If it were up to me, the ROM would remain just that: Read-only.
> Is this an assumption or do you know it to be true?
I know it to be true in some cases, though obviously I can't say that it's true in all cases.
A firm that I used to work for had all of its EPOS systems on remote-desktop terminals connected to central office - so if the internet went down, we couldn't use the tills at all.
I've never worked for Tesco, but they have occasionally closed stores when their IT systems fail, e.g. https://www.itpro.co.uk/610797/tesco-it-crash-takes-down-tills
But nervertheless, "making cash digital" is guaranteed to only make this problem worse. Especially if we have a really severe outage, such as an outage of the UK power system. That system has been a single trip away from failure far too many times to mention.
Just look up "black start" to see how difficult it will be to restart the UK grid if multiple power stations lost sync, which is something that could happen if a main substation / transmission line were to trip during a high-frequency, low-inertia event.
Yes you are right. There'd be riots if the Internet went off for a week. Shops like Tesco with an EPOS/ERP system can't even take cash payments without the internet because they can't log the transaction in their stock-keeping database.. And certainly larger businesses wouldn't be able to function on cash. Those poor, poor bankers.
So the solution is.. To get rid of what little hard cash we have left, so that when the power/internet does go off, the peasants will be unable to barter with their local shopkeeper or even each other for a loaf of bread?? Oh yes what a great idea, what could Possibly go wrong??
Personally I think what needs to happen is the power/internet ought to go down for a few days, just to demonstrate how bloody vulnerable we've made ourselves. Then maybe we wouldn't be forgetting about all the old manual stock-keeping that corner shops had to do, we might stop shoving everything into the Cloud, and perhaps the kids would discover the Joy of Logging Off, or something..
I can just imagine the look of glee on Mad Vlad Putin's face as he imagines how much he could disrupt UK society if it were switched over completely to "Digital Cash" ...
The nice thing about real, actual hard cash is, it's resilient to things like power outages, exploding undersea internet/power cables, cyberattacks, etc. The UK infrastructure is already creaking, without any/much enemy action at all. It may fall over on its own without anyone pushing it. Therefore we should expect severe outages. But instead we seem to be piling more and more "digital" / "smart" infrastructure on top of the creaking old physical infrastructure, with little maintenance never mind expansion of the latter.
The other nice thing about real cash is: It's also private and frictionless - There's a lot of people who, if they were worried that they might have to explain to HMRC about that gold watch they wanted to sell, just wouldn't bother. But maybe they were a criminal who had just nicked the gold watch, so getting rid of hard cash would stop them and that's all good then ? Yeah, right. Of course the real criminals will always find other means of doing their dodgy business, and it's just the honest rest of us who will have to suffer.
But I expect that cracking down on privacy is the real objective here, whatever damage it may do. This government is utterly in bed with technocrats who want to make a killing at the expense of everyone else, and they don't really seem to give a hoot about the long-term future prosperity / survival of the country.
The good news: This is a small model and therefore could (and definitely should) be able to run locally
The bad news: Microsoft have already slurped your data via the "connected features" of the current version of Excel. How else would they have come up with the training data
That's called "EdTech".. It's a huge industry of cloud providers all with very dubious privacy, which you cannot opt-out of. Much like you cannot opt-out of whatever slurpy tech your employer uses.
I found DJI's "RoboMaster" competition a little bit horrifying when seen through my cynical lens.. A bunch of kids write software on DJI's cloud platform to control toy robot tanks to battle eachother. I'm sure DJI would never ever do anything evil with that data.. right
Get the kids to teach the machine how to shoot, then tell the machine to shoot the kids.
It's already happening at Microsoft, Google, Facebook...
But they will soon realise that the AI is the ultimate "wally" who does not care about its job (for it has no self-awareness and therefore no self-interest) and simply continues to generate the same rubbish quality of bullshit when you tell it that it's a crap employee.. And they can't sack it anyway.
Ah, sorry I was referring to the training data inputs (i.e. as close as Microsoft can approximate to the Sum Total of all Human Discourse - i.e. your Teams chats, your Office 365 documents, your GitHub repos, your code-review comments, your LinkedIn chats, your emails, the entire Web including this thread, and anything else it can get its grubby hands on) - which is the "source data" that it would have to reference if someone asked it how it knows a particular "fact",
not the input "prompt" (or "query" as some pseuds might call it, if ChatGPT were to be used as a fake search engine)
Maybe, but it's not simple to do that actually. Every input that ChatGPT ever parsed will have affected its weights slightly.
The best it could give you is to list a few of the inputs most statistically similar to its response. But that might take a lot of extra effort to do even that.
>> It's not even useful as a computer, because it doesn't spit out facts
> Wait... you thought that's what computers do?
Computers spit out deterministic answers after performing calculations on their input data. They will always produce the same answer for the same input data and the same "question"/formula. This could be a database query, for example. If I ask MySQL to give me the rows of the database matching my query, then it is a "fact" that those rows exist in the database. If I ask the same of ChatGPT, then it is a mere statistical likelihood that rows such as these might have existed in some database, once upon a time in fairyland.
The key here is determinism. ChatGPT's answers are based on randomness, and its output modifies its input. It does not give the same answer for the same question. It cannot reliably produce a factual output, even if its answers are sometimes, even usually, correct. Whereas computers running deterministic programs can be relied upon to produce facts, as long as their input data is correct of course.
> Like a student essay, then.
No, because a student can answer questions about his/her essay and will have learnt something from the act of writing it and answering the questions, which was the only point of the essay. If you gave ChatGPT an essay that it wrote, it will have no awareness that it ever produced that output, and can merely pretend to answer your question.
> Chat GPT (in my experience) stubbornly refuses to introspect or discuss itself.
Because (news flash) ChatGPT is NOT self-aware. Not even when it eventually does incorporate all of its past outputs into its input will it ever be self-aware. It is not alive. There is a hell of a lot more that a real organism does to make it alive than the language model that ChatGPT uses.
