Posts by that one in the corner 5578 publicly visible posts • joined 9 Nov 2021
> {copyright} exists to protect money. No other reason.
Stated crudely enough to be inaccurate, as I'm sure you know. Copyright exists to protect the interests of creators and allow them to make a living - purely in order that they spend time creating things - without the need for a patron. Especially as the ability to reproduce increased over time. Copyright and the controlled selling of mass reproductions allowed us to provide patronage en masse, all of (what we now refer to as mere "consumers") providing a small portion of the creators interests. And, yes, these days that means money, simply because that is a good mechanism for handling such portions: individual patronage often being in the form of goods and services provided, including lodgings.
> Almost all of the copyright legislation is created by Hollywood media magnates
The current rotten state of the copyright legislation has been caused by those magnates - more accurately, political systems that enable those magnates to affect the legislation. But that is another matter.
The concept of copyright is definitely a Good Thing. Its execution - varies.
> software is not a creative work. It's a tool.
A *completed*[1] piece of software is a tool - we freely admit that in the documentation: compilers, linkers are part of our tool chain. Other software may be "just" a utility whilst bigger lumps are called applications.
But *making* the software and completing it or adapting it, *that* is a creative act. There are lots of stages that are trivial, repetitive and can even be totally automated[2], but the same is true for other creative pursuits: I'll sand with a dremel because it is less repetitive and tedious than using a teeny tiny sanding block[3].
> Its value comes from what it does or what it can be used for, not from merely existing
That is a trivial observation, too trite to even be a truism. Name anything that literally "has value" merely by existing! "Value" is something that we attribute on the basis of our personal preference or how we react to someone else's preference or actions with respect to it. That driftwood has no value to her because it won't burn cleanly in the hearth, but he can see how to use it in an artwork. Even money - a fiver has no innate attraction to me (the plastic ones are quite irritating sometimes) but I give it value because of how it can influence others actions (like letting walking out of the shop with a new pair of socks without angry shouts).
[1] ok, we all know that no software is ever really "complete", but you know what I mean.
[2] closing brackets, expanding a template to insert an entire loop - as time passes we attempt to increase this list, in order to spend more time on the interestingly creative bits.
[3] and my results are certainly unique! Hopefully I'll get better with practice and someone may be able it is "creative" with a straight face! It is *meant* to be a dragon, btw.
Presumably from the script kiddies and wannabes who realise they'll never make it as programmers without every single ounce of spoon-feeding they can get from Copilot.
Whatever happened to the shy nerds who used code as their means of expression?
Colour me confused; as a non-US type person, I'd been under the impression that the FBI only worked within the US and the CIA dealt with anything from outside it. But the phrasing of:
> Abbate did acknowledge that sometimes people on domestic soil get caught up in a sweep. For example, if the FBI is collecting foreign intel on a hacker in China, and that cybercriminal is working with people in the US.
seems to indicate a situation where the FBI started out solely with an interest in China, only later expanded to looking within the US (and then being too broad there, hence the complaint against them).
Anyone care to point me on the correct path to understanding here? TIA.
No roasting from this direction.
> why don't we just statically link everything
Some people just seem to love complexity?
If you have 147 shared libraries then you have a Grown Up program that Needs An Installer or, More Manly yet, you get to learn how to set up a Snap wotsit or a Flat thingie or shove it into Docker no matter how pointless that is for your Users' Use Cases. But your CV grows.
Used to get this with Windows applications as well: I'll arrange everything to build statically, with as few DLLs as possible (sometimes Windows forces it on you - e.g. using multimedia timers) then someone comes in for a short time and poof! DLLs everywhere ("it saved so much time, didn't have to compile anything"), no idea if they are all still needed ("I just copied them all in, that was easier"), we don't have the debug symbols for them...
Still had to create a new release just to distribute, e.g. an updated SSL library, whether that was statically linked or in a DLL ('cos you can't rely on the User updating "just this one DLL" and nor should you).
Sigh. Time for Grumpy Bedtimes.
Talking about LLMs, including ChatGPT:
> It’s not that they sometimes “hallucinate”. They hallucinate 100% of the time, it’s just that the training results in the hallucinations having a high chance of being accidentally correct.
A nice way of phrasing the problem, which I'm shamelessly nicking (sorry, I mean "am excerpting a portion of under Fair Rights[1]) from a discussion of this case over at https://reason.com/volokh/2023/06/06/first-ai-libel-lawsuit-filed/?comments=true
[1] Which is an US concept, but as it is a US website I'm ripping off - sorry, there I go again - and The Register now self-identifies as USazian I'm probably alright
> I can't see anywhere this article says that the defamatory speech was published anywhere
Confusing, isn't it?
Since posting (above) I've been doing a bit more searching and I think I'm going to have to change my opinion about the liability of "Fred Riehl"[1] - he *is* a lazy "journalist", as he likes to use ChatGPT to generate his "stories" (although he does admit that in the byline, so that is one thing in his favour[2]).
*But* as far as I can find out, Riehl never actually published any article containing defamatory statements, as we (including myself) have been assuming in the comments here. Instead, it currently looks like all that has actually happened is that Riehl asked ChatGPT and got back a response mentioning a *random* "Mark Walters" and in doing so, it was ChatGPT that was publishing this information!
Now, it appears that Riehl and *a* Mark Walters[3] are buddies in the "pro gun rights"[4] movement - Walters writes[5] for the website and his radio show has been promoted on the website - so *obviously* when ChatGPT talks to Riehl it must be referring to *that* Walters, hence ChatGPT has published defamatory statements about his colleague. All the contacting Gottlieb was to check the facts, just in case his Mark really was a wong'un - or more likely in order to be able to say to the world "look, even Riehl can manage to do this must due diligence, so ChatGPT must be really negligent".
So, ChatGPT is, apparently, "publishing" when it spits out a response and, despite having to retract my previous idea[6] about Riehl (for a new and even worst one, but hey), I still stand by my belief that this is all an attempt to get publicity for two otherwise totally pointless individuals and their ridiculous website and radio show.
[1] His own website and linkedin profile name him "Fredy Riehl" (having both "d"'s may have been too bourgeois for him)
[2] About the only thing so far; maybe he likes fluffy kittens as well?
[3] Hereinafter referred to as "the idiot doing the suing" M'lud
[4] Nope, already made that joke
[5] https://www.ammoland.com/author/markwalters/
[6] Science - we change our ideas as the evidence leads us
Bingo.
Which is why suing OpenAI is ludicrous - given, from the article:
> Riehl contacted Alan Gottlieb, one of the plaintiffs in the actual Washington lawsuit, about ChatGPT's allegations concerning Walters, and Gottlieb confirmed that they were false.
So the only case for being negligent with the truth would be against Riehl, who either didn't bother learning the basic limitation of ChatGPT *and* contacted Gottlieb to late (basic lack of fact checking) or did so prior to promulgating the incorrect statements and went ahead anyway.
However, given that there is (so far reported) nothing to connect the non-existent Mark Walters to the gun-promoting radio host bringing the case, the whole thing is just a publicity stunt anyway, as Riehl is connected with another gun-related website, ammoland.com
You appear to be correct. The only identification of "Mark Walters" given (in all the reports of this that so far found) was that he lives in Georgia and held a role at SAF.
The Mark Walters who is making the complaint is a radio talk host who doesn't claim to have held a post at SAF. If there is any actual reason to connect the two then it isn't being reported.[1]
This Walters wasn't even the first hit for a search on the name plus "Georgia" and when he does show up it is only due to this case. His pro-"gun rights"[2] radio show finally appeared after a bit of scrolling.
Given that, if there *is* a case to be made[3] it can surely only be againt the journalist, the only one who could be caught be the requirement he must "prove that the defendant was at least negligent with respect to the truth or falsity of the allegedly defamatory statements" (because, well, only a human can be negligent) then this looks to me like nothing more thsn an attempt to get publicity for his radio show.
[1] maybe Walters the radio host is is going to argue in court that he is a well-known embezzler, so it must be referring to him?
[2] guns have rights? But does he support the right for two guns of different calibre to co-habit in the same box? Won't somebody think of the ammunition!
[3] which I severely doubt
Also similar to the technique used in civilian phased-array radar systems or the phased-array antenna available for 5G phone signals.
But I guess all of us imperialist warmonger commentards are going to be more familiar with the military application and it does tie nicely into the use within war-torn areas (technology giveth and taketh away).
> It's a real shame that every vehicle doesn't have the same cameras and sensors and that it doesn't log all those inputs and control inputs too. It would be incredibly useful in both determining if any laws had been broken and who was liable for insurance purposes.
Giving that much data to the insurers (and you know they'll get it) could lead to some interesting results:
* BMW drivers see their insurance go up because the insurers can no longer pretend that they know how to use indicators.
* Audi drivers tailgating picked up by their own forwards parking monitors, insurance goes up.
* Everyone who loudly proclaims that they are "better than the average driver" gets the annual report that shows reality; insurance goes up
Of course, it won't all be such good outcomes.
> Meta's LLaMA, a seven-billion-parameter LLM
Aaargh - it is a seven-billion-near-as-dammit-arbitrary-number LLM!
A seven-billion-nadan LLM! There you are, "nadan" - doesn't that even sound more exotic and intriguing than "parameter"?
A "parameter" is an input that has understood meaning - if you "know the parameters of the problem" it means that you can identify and *describe* each parameter, *explaining* how they affect the outcome. That is why we say that a function has parameters - which we give meaningful names to - and we pass over values that we know to be sensible (well, when thing are going to right).
The LLM - and any of its relations, the varieties of Neural Nets - tweaks weights between connections as it is trained. Looking at the result, this huge pile of simple numbers is totally incomprehensible: you won't gain any useful insight by asking "Why is this 7.01" - it just is. You can't say "I want to make this change to the outputs, to do so I will change that 7.01 to 7.02", there is (as yet) no way to determine whether any given number in model even *has* an effect on the outcome (it may be blocked at any level in the layers preceding or succeeding it). Unlike a Markov Chain, the layering makes these large stochastic models totally opaque. There is ongoing research to see if this situation can be improved upon, without "damaging" the "usefulness" of these models but we are not there yet.
Rather obviously, the LLM pushers call them "parameters" - and expect everyone else to follow suit - because it makes them (both LLM and peddlar) sound cleverer than they are (old advertising trick - even when people say "oh just call them parameter, what does it matter? Don't you know language changes all the time!" the mind still attaches some of the gravitas of the word to its lesser usage).
> The researchers noted that they cannot explain why a search result is trustworthy or not.
Because they are just flinging nadans not parameters! These nets have *no* explanatory power, unlike other approaches to "AI".
> They hope to come up with another strategy to increase accuracy and reliability in the future
But, note, not to increase (well, start) providing any ability to *explain* why the result is trustworthy (or not).
> low powered ARM boxes (and certainly with old TLS boxes too) the only recourse is a hardware upgrade.
Except in the situation that the Arm boxes are still running a level of security that is sufficient for their use-case (e.g. on a LAN that is itself sufficiently disconnected from anything outside) so upgrading that hardware is a needless expense. But if you can't (easily, nay trivially) connect your Windows laptop to all the devices on that LAN then - angry faces all around.
I have SMB1 re-enabled on a Windows 10 box connected to a cabled LAN; if anyone is on that LAN then I've got worse issues than worrying about that particular Samba traffic. If -when - Windows plain stops allowing SMB1 then will I chuck all the perfectly working devices on that LAN? Or is it cheaper & simpler to stop using Windows with that LAN? (Hint: the answers are "No" and "Yes" respectively).
> The issue is that people 'want' security
People want security that matches their requirements. For Joe Bloggs or anyone who doesn't "want" to think about it (or, more likely, simply isn't knowledgeable about it and isn't in the job of becoming so) they want the *defaults* to be as good as possible. For many readers here, we want to use *appropriate* settings (and know it is on our heads if we choose badly) - and do so in a *supported* fashion, not from random web trawls.
> and they also don't want to pay anything
We don't want to pay for what we don't actually need
> or expend any effort to do
Joe Bloggs ought not need to expend any energy, he gets the defaults. For us here, following supported methods is fine, that is an appropriate amount of effort.
Reverse engineering Registry settings to disable excessive protocols, or even just trawling the web praying someone else has succeeded in doing so, is not something we want to do (or ought to have to be forced to consider doing).
I can see lots of LANs with Samba traffic happily flying between Arm devices, NAS boxes, DVRs, media players etc - and the poor old Windows 11 laptop suddenly being shut out of the game.
Maybe Windows can have the requirement disabled, but by the poor sod who just updated and wants his photos "back"?
> I just hope the doctors that use this are required to 'sign off' ...
From the article:
"Generative AI models aren't perfect, and often produce errors. Physicians therefore need to verify the AI-generated text."
So - yes, they are.
For the time being, at least. Once everyone becomes complacent..
The use of GPT4 given here is *not* for any diagnostic purpose.
> If the 'AI' is only given the audio recording, what about understanding accents
As already said, GPT4 is getting the transcript, *not* the audio - it does not have to deal with accents
> and any images made of injuries, wounds, possible cancerous growths, rashes, etc?
GPT4 is *not* being asked to provide any diagnosis, it is not examining any images!
It is being used to convert the transcription - the conversation, with all the repetitions, hesitations and irrelevant "no need to be embarassed, I've heard it all before" comments - into a standardised format with just the useful info retained. And they are claiming that 12% of its output needs to be edited (corrected? deleted? made to sound sane!).
The article points out that the text is still being verified by the doctor (who is then correcting the expected 12% of garbage).
So the patients are still relying on the doctor matching up the output to the relevant patient (in case GPT4 hallucinates an entire case history) and spotting when something that. sounds good (remembering that LLMs have a tendency to sound convincing, with good grammar etc) is actually inaccurate.
I don't get to blame the lathe manufacturer when I drop it on my foot.
> Almost every panel The Reg attended involved experts having a chuckle at the lawyer's folly.... Almost every panel The Reg attended involved experts having a chuckle at the lawyer's folly.
So almost every panel agrees with my sentiment and yet
> "It's even more important when non-experts are using these tools that there is accountability, non-liability. If you aren't an expert, where do you go to check the tool? The onus should be on those creating it."
These - things - are toys and, as a professional, you have a duty to know that (your professional body sure as hell has a duty to you to make that clear, preferably in the rules of conduct) and be prepared to accept liability.
What those creating the LLMs should be made to do is state clearly, evey time,"this is for entertainment purposes only" - preferably within the generated text, multiple times! At least make the pillocks misusing it have to do some deliberate, knowing, editing so there can be no claims of "I didn't know".
FWIW I do understand the enumeration stuff, the underlying *mechanism* of the "why", but I've never (in my uses cases) seen it work *usefully", hence questioning "why did they decide do it like that (and not let you disable it)".
Having an old fashioned, fixed, COM port[1] means I can program multiple MCU boards[2] with a simple script[3] and just plug them in one after the other. Ditto talking walking around a lab at work, talking to embedded devices at work. Not many (any?) utilities that know COM ports exist will continually re-enumerate the COM ports and keep their UIs up to date, so even if they do the "poll every port and look for a known response" trick[4] you have to restart them every time.
[1] or an FTDI cable, glued into one port so it damn well never moves[5]
[2] after bypassing the "helpful" onboard USB-serial, assuming that is feasible
[3] like "make install"
[4] and hopefully the chars used in the poll don't send other bit of kit, still plugged into another port, whacko: did that robot arm just move?
[5] I'll need to be convinced that Win10 reuses the same number *IF* I can plug the device into the same USB *every* time: k USB ports, greater than 2 * k devices? If Win10 is repeatable, I *could* write the effective COM number onto each device - just need a tag big enough for every combo of every USB port for every computer that might be used to talk to this device...
> People then bring up Disney - but really, I'm ok with the Mouse House keeping hold of their rodent
No-one cares about that mouse, the real damage Disney does is opening the door so that every other copyright extends to stupidly long terms.
> a response to a system that is clearly broken on many levels.
You will never fix *anything* if you put off trying because "doing that one thing won't fix it all in one fell swoop". Something something Rome, one day...
Huh?
Not sure what you are trying to say, but just in case I've interpreted your comment correctly and you think that the latest Windows *can* actually
> cover their use case
You do know that using the latest Windows isn't viable if you want (need, in more than a few industrial cases) to keep older hardware running?
Windows 10 won't fit on my (perfectly functional) Dell laptop-with-real-serial-ports-that-don't-change-COM-number-every-bloody-time-I-plug-in-a-widget [1] and Win 11 won't even try on my "newer" Dell-small-enough-to-really-carry-around laptop (which I only got last year so bog off, not buying another one).
Fingers crossed the drive backups won't have evaporated when the unthinkable happens and I need to reinstall from scratch, but at least now I can put the legit original XP copy back on and get it activated.
[1] seriously, why the bleep does Windows do that? It isn't even as if I've got that many widgets, it bumps up depending upon the software! Some day soon my box will have COM100 and I'll find out that an otherwise really useful utility only copes upto 99...
Conversely, has Enigma ever *proven* that its software actually does anything and, if it does, does it do that thing completely and with total accuracy?
If not, then does it have an argument for saying its software *isn't* "potentially unwanted"?
Even if there is a totally benign bit of software (i.e. doesn't do anything bad) sitting on my system, if it doesn't do anything good either then I probably do not want it!
Please take the following as aimed at Crockford et al not your good self:
<rant>
JavaScript can just eval JSON, any other language is going to *need* a separate parser (and, of course, in real life, JavaScript *ought* to be using a separate parser! Eval'ing text from an unknown source...).
If anyone is writing a parser for a text format, using *any* programming language (even VB6 or, gasp, JavaScript) and can not figure out how to allow for comments then - well, let us just say they should not be doing so professionally or for anything that is expected to be released for more than one other person[1] in the world to use.
Even if the claim was that JSON was meant to be viable on a low-resource MCU in an IoT device (bleeugh), JSON is complex enough that the extra states to handle comments are trivial.
</rant>
[1] their lab partner in the exercise after the second lecture on basic compiler techniques.
> It was only later that it got used for configuration files, where the lack of comments is a nightmare
Dunno about anyone else[1], but when I've chosen the format for files that are used "only" to talk between an application and a server, as often as possible[2] they are text-based *and* damn well allow for comments!
There is this little thing we like to call "testing" where we keep lots and lots of files in the repo with subtle differences between them, all commented up to the nines to point out those differences and the expected results[3]. Which are fired off/caught back using curl.
For that matter, last paid gig where I created such data (used to send a few values between a set of peer servers, "no human intervention required") the *generated* messages contain comments (telling the reader what server subsystem created the file and guiding them to the docs in the repo - 'cos I'd wasted so much time trying to figure out similar info from other traffic in that project!); the "waste" in doing so was well within the available bandwidth for the task and the payback in live tests made *my* life easier[4], so win.
[1] in particular, it seems, the people who came up with JSON and *all* of the early users who should have been screaming at them!
[2] i.e. if the time and processing resources allow - e.g. between two 8 bit MCUs with 2K RAM I'll probably forego the comments (and shorten the keywords).
[3] what, put those comments into another file? It is hard enough to get people to change comments when they change the primary file, trusting a secondary will be up to date is lunacy!
[4] my bit is working, so it *is* your code going mad; toodles.
Oh for pity's sake - you mean I've been fooled into thinking JSON is actually *better* than it officially is because of sheer luck in which parser library was chosen by the project?
Thanks for putting me straight on that.
What a totally moronic decision - taking a pure text format derived from JavaScript and deliberately stripping out comments!
Bash scripts can be commented - can these key/value pair files?
I've seen far too many data files being "parsed" by some uber-trivial sscanf()-like calls instead of at least using a simple lexer that can allow comments and whitespace handling so the files can be laid out in a readable fashion.
And for anyone who leaps in and says they are all using JSON files now, so there is a proper parser, comments and all: that means they're probably using something that is vast overkill! So are they correctly limiting the complexity of the data structure allowed, 'cos now the parser will happily let you define a list but your code only eats a single value: bet it uses the first item and doesn't raise an error that it is ignoring the rest!
> what else are you going to do if you're going to have to wait around for at least half an hour to recharge
Recharge yourself.
Take a nap; take a short walk; read a book; write a love letter; paint an abstract; sketch a tree; needlefelt a gnome; sandpaper a birdbox; scrimshaw; knit a pair of socks; survey the flora; crochet an octopus for premature babies; practice your Latin; go to Greggs; do calisthenics; form a support group with the other drivers; laugh maniacally; ...
That would be interesting - usually I can just go straight to whatever it is I want to use (the VLC icon is there, xplorer2 on the Start Menu as is Device Manager, thanks to Open Shell).
Otherwise the discussion in my head tends to go like:
"Blast, what is it called? It has that icon with the twiddly bit on the bottom and it's yellowish - or was that the old icon? Do I have it on Steam? Nope, nothing in that list rings a bell. Oh, come on, come on, it's the one with those cards in different colours! Is it on the D drive? Nope. Oh, hang on, it *was* on Steam after all!"[1]
How well is the chat going to deal with that? Plus, I'd now be dealing with a stupid machine that won't flipping help me do what I want to do, instead of being actively occupied in the search (and, btw, being reminded of the other fun things I've got installed, one of which may be a decent substitute) - and keeping occupied reduces frustration levels and accompanying invective...
[1] certain extra words and tone of voice have been adjusted to save the blushes of the audience.
We're waiting for a suitable time (when all the episodes of an interesting series are available to binge over a week) and I'll sign up for another 30 day trial of Prime.
So far, have done this free trial three times (you have to wait a bit in between for the period to climb back up to the full 30 days). Did also deliberately extend it by paying for one month as it would be cheaper than paying the p&p to get presents delivered to the right people at the right time.
Hasn't been difficult to cancel before the payment is due (worst case, one time I forgot where to find the cancel link and - gasp - had to do a web search). I'll admit to avoid signing up for your first free trial prematurely you do need to read the web page and not just click on the biggest button presented.
Not too proud of myself for using Amazon, but considering they've been buying up sites like Abe Books it's getting harder to avoid (after checking out all the local second hand shops, of course: but we want to read Len Deighton, not Dan Brown and guess which one appears in *every* charity shop?). But if we're going to have Amazon accounts, may as well get as much of the extra bits for free as we can.
> UEFI is absolutely hideously over-engineered,
> I'm not sure what the best solution would have been.
For a start, not to have followed Design By Committee procedures quite so assiduously. Gary Kildall's BIOS was well made and performed the task sensibly, but as you point out, it had then been hacked upon for decades[1] by whoever got the job dumped on them next.
Although, as you have (inadvertently?) hinted, at the moment the single voice with the most pull who would have gleefully taken on the job would have been Poettering and gawd help us all.
I too liked Sun's use of Forth: you have reminded me, I don't know UEFI at all well (like, how does Windows update manage to make UEFI override my choice of boot OS?). For example, how many different bytecode interpreters are in there now? Is it just ACPI?
[1] starting right from IBM's "we did it deliberately, honest, no we do really know the difference between hex and decimal" in the serial i/o routine in their first PC BIOS.
To be fair though, bomb number 20 didn't decide to ignore its inputs of its own volition[1].
But a good reminder of another great documentary about AI that we should all watch (again).
PS: the uniform still doesn't fit
[1] oops, too late for a spoiler alert? But at least I haven't told you why you should be wary of beachballs.
Ah, did you mean to say "I believe the Air Force spokesperson Ann Stefanek"?
And whilst I'm here:
> B) As an earlier poster worked out, if the AI required a confirmation to fire, it would have required a confirmation to fire on the operator or his com tower. The failure mode listed wasn't 'failure of the failsafe.'
Uh, what was the "failure mode listed" and where did you see it listed? Not seeing that in either the Reg article nor in the linked-to articles. If you refer back to the article, the suggestion is that the confirmation would just be another input to the model (e.g when the model hits a point that it needs said input's current value it'll ask for it) - and any "just another input" can just be optimised out (i.e. just don't bother asking for confirmation, he might say "no").
> D) ... Why would you waste time denying a win?
Because it doesn't sound like a win to Joe Public? Us Reg Commentards know that it is all just part of the process to run a sim and react appropriately to whatever it throws up, including not releasing the thing into service, but they may be concerned that Joe reads it as either "this is how the AI will always behave" or prosaically just "So they've wasted all this money and they have nothing to show for it!". Trying to explain the dev process to Joe may just be deemed to be more costly than another run of the mill denial!
> Or are we now splitting hairs over what a simulation is?
Maybe we are, if the spokesperson is using the current MilSpeak (strangely similar to ManagementSpeak). After all, the update continues
> "and remains committed to ethical and responsible use of AI technology"
which would seem to indicate they think that running - and reporting on - a "misbehaving" AI under simulation is *not* ethical and/or responsible.
But running an reporting on a wayward program run with simulated inputs and, most importantly, simulated outputs, is the epitome of ethical and responsible behaviour! Yes, as noted above, the behaviour of the program was hardly unexpected if they'd read the literature beforehand so we can chuckle at their naive surprise. But if you're going to follow that path it is *highly* responsible to try it out in a harmless environment first.
Heck, even if it turns out to have all been taken out of context, if it prompts anyone who really is creating a similar program to think about how they are setting it up, that would be a really positive outcome.
The tale my professor told was of an image recognition system to find tanks in photos. All went well in training but it totally failed practical trials out on the moors: there were never any tanks to be found!
Turned out that the pictures of scenes with different sorts of tanks mostly came from sales brochures, Janes Fighting Vehicles and the like - which were all beautifully lit and sunny, with dramatic poses showing off the tank to its best. The photos without tanks were taken in more - average - conditions. Of course, the moors were overcast and soggy, totally uninteresting to this program that had been carefully trained to detect and sound the alarm whenever it was a sunny day.
This sort of response was included as part of CompSci courses back in the 1980s, for bleep's sake: a key part of any learning mechanism is that you have no idea what/how it is going to learn[1], you just keep your fingers crossed that it will actually manage to learn *something*. Especially back in the '80s when machine resources were lower than today and it was painfully obvious that the system had gone down an inefficient route and couldn't achieve any of your goals for it, scrub and try again; nowadays just fling more cycles and memory at it.
Genetic algorithms will happily recreate the Vagus nerve's ridiculous looping down and back up again. Any signal into a system may be given "unexpected" importance, either higher or lower than *you* expected (because you think you are looking at the forest but the model can't even see the trees, just the leaves): a system that is "punished" will be trying to optimise away the punishment and just ignoring that input is a perfectly good way to do it.
The reported stories about AI programs used to be specifically about these weird results, such as the analogue circuits that "ought not to work" because the program had optimised some weird arrangement of parts that took advantage of an oscillation that humans worked to get rid off.[2]
In other words, do the bleeping background reading before trying to build a system![3]
In other other words, blasted whippersnappers, get off the hole where my lawn used to be!
[1] Or you would just, you know, program it directly.
[2] And then how the humans' approach was the useful one, as it allows building blocks to be created and assembled into bigger systems but the "clever" design approach had all the parts interacting together and couldn't scale up.
[3] Wasn't there a time when scholarship in the military was a thing of pride? The design of, success and failure of, everything from strategy to tactics to ordnance? From the importance of land surveys to waterproof boots?
Any Python install that can pull down a package and expect to run it should have the tools to generate pyc files locally.[1] Unlike C extensions, where we can no longer expect every machine to a C compiler.
So why are pyc files allowed as part of the package in the first place, as everyone appears to know that the scanners won't work on them?
(Or have I missed the part where the pyc was cunningly disguised and encrypted to prevent anyone spotting it is pyc in the first place?)
[1] e.g python.exe with the appropriate arguments!
Very true, but the article notes, he
> would also have to cover the costs for "evacuating water without permission," wasting 4.1 million liters
Now, if he is charged the going price for Harrogate[1] water from Ocado (as he will need to have it delivered...), at £1.93 per litre (note spelling) that'll be nearly £8 million, which be fairer. Oh, if only they would.
[1] For comparison, they list Fiji Artesian water at £1.98 per litre (in a box of 12; Harrogate water only comes in singles, for some reason[2]) - unless that is Fiji, Lancs.?
[2] You'll only ever want to drink one bottle of Harrogate water! If it is the Proper Stuff, it has a - certain nose!
