Nice catch - I was completely oblivious to it till you pointed it out, and it was an accident, I swear! Lazarus Long says punsters and practical jokers get keelhauled or staked out on anthills :-( That's no fun
Slick idea - I'd be concerned tho that, given the semi normal rough distribution of heat on the face (there's a reason head wounds bleed so much) that this attempt to game the system would be easily detected as gaming the system. The good news is any idiots building a product by bolting together parts from OpenCV that they barely understand would probably fall for it. Me, I'd be embarrassed if I didn't tag you as anomalous.
HOWEVER - if you were really really good and accurate with the application of the blocker (cause if its zinc oxide it does work in both directions) and you backed it up with makeup for the visible light side of the house, you probably could fool most by establishing a completely different facial geometry.
Anyway, certainly a concept worth ruminating on. :-) I think it would beat me at my imaginary best if done well. Have to find a counter, I guess.. Like gait :-)
Probably - its an arms race after all - at this time for the stupider systems the following steps can confuse them
1 - use makeup techniques to make the nose fade into the face and don't let it get a profile view - noses are very important - a prosthetic nose probably won't help tho, cause the nose is an orientation characteristic more than an identification one
2 - eyes and the facial geometry around them are secondary, so no nose and giant movie star sunglasses adds to the machines confusion
3 - collar worn sparkle lighting directed upwards is baffling to machine - silly things keep picking up the light sparkles on the face as feature points (heh)
4 - makeup patterns based on the WW2 ship camouflage mess up geometry detection
Now, all of that will defeat a simple recognizer but if the developers had a lick of sense, the system will still pop it out as anomalous.
Mission impossible facial masks could certainly help, but pretty much everyone is removing the IR filters at this point and the facial thermography would be all messed up
Then there's all the work going on in recognizing more than just a face - I've seen some stuff on incorporating more of the body into the recognition pattern (I've always wanted to watch one of those systems misrecognize Mitch McConnell as a turtle :-) As the available horsepower continues to expand, more of the totality of the individual can be recognized (my own accidental contribution to this is that cheap stereoscopic vision makes it a lot easier/more accurater to pick out foreground from background)
At the end of the day, I think Scott McNeally had it right in the '90s - "You have no privacy. Get over it" I am of mixed minds about this stuff in public spaces, the good and bad examples are evenly distributed
If the technology just stayed at facial recognition, maybe - I work on stereoscopic vision, so you'd have to change your head geometry too for my cameras if you're in parallax range..... and we're getting to the point of being able to do realtime skeletal pose analysis -- https://storage.googleapis.com/tfjs-models/demos/posenet/cam...
So, it's an arms race :-/
The "touch" button thing was a big deal back then - Apple and others kept reminding devs that is was not a good idea to tell users to "hit" keys as sometimes this was taken a little too literally :-)
What is interesting (thus far) to the responses is that no author of a paper has shown up going 'you're stealing my work' - nor is any money gathered by most of these organizations returned to research - myself, I've always envisioned Elsevier as the dutch dude in Austin Powers - One possible cure I heard from a librarian sobbing over the bill was that if there were companies that could hold copyright and companies that could distribute, but no one company could do both there might be some market tension
I agree on the assessment of the article, however not with the conclusion that blockchain is driving this market alone - its certainly there, but I don't know any ML practitioner who answers 'how many gpu's' with anything other than 'how many can I have/afford ?' We're starting to get more and more projects reporting their processing time in GPU/years - the singularity is bull, but ML ain't going away:-) That said, if people start dumping their mining rigs, I'm on ebay looking for them :-D
OK, that was fun - dug up an old quote from the Ithaca Intersystems linker manual I always loved (quote, not manual) -- "This program is dedicated to and named after a good friend, Link Hogthrob. "
We should do our best to get it back into port - part of the whole mess is that the name AI implies things about ML systems that simply aren't true - as a side note, we should also probably start using the word tensor more accurately, we've now enraged enough physics and math folks :-)
I ask it sometimes too - my favorite response opened with 'you want maintainability or speed' So I picked speed. They immediately set the first branch at mod 3 with the obvious justification - hired on the spot :-)
Huh - I do ML and a lot of bit twiddling with trained feature detector outputs - intake goes into a stage, feature detectors light up, and feature combos get fed to subsequent input via lots of bit banging - can't afford for it to be slow in the middle of that pipeline!
You want to proceed carefully with a Pi Zero - the chip is an ARMv6, so you'll end up building more things and you loose some optimization - you might not like the performance :-(
As another practitioner, I wholeheartedly agree - most of what is called AI today is ML, and thats basically pattern recognition via boundary relaxation solutions - it is necessary for "AI" but not sufficient - BUT - my problem with the authors position is that he keeps not noticing A stands for artificial - we've had other solutions in other techs, Mycin and Dendral come to mind - no, they weren't AI either, however I think when future work begins to incorporate more than the PDP groups seed stock, we may see AI's begin to appear - and then the arguments will start all over with 'too narrow to be intelligent' - first we'll have to figure out 'it looks like cognition' before we move on to anything remotely resembling actual cognition - at present, not only can we not do it, we have never ending arguments about its very definition.
I'm wondering if the work on adversarial systems, this one being quite interesting, can help us with our giant bugaboo of "OMG, its overfitted :-(" Right now we model, train, test, fail, and start all over again, and usually fiddle with the hyperparameters to boot - what would happen if we turned training into a two phased approach, with a BEAM/GAN whatnot used on each cycle to measure how 'brittle' the backprop is? The idea being to round down the spikes in the learned model by penalizing the backprop when it is too narrow - training would take longer, but we'd throw away fewer sets, I'd think
To me as well - first it starts with "I don't want to lose any of my soldiers but I must be able to kill theirs" and then more and more questions about machine autonomy keep popping up and finally we've got machines capable of making the decision on their own, and then we're pretty much doomed - not the snazzy terminator robots, just a lot of dumb plastic machines lower on the intelligence level than a flatworm, however that won't make us any less dead as a species. If you aren't willing to risk lives you absolutely shouldn't be killing people, and we're getting to the point where we really need to give it up as a strategy before we do ourselves in.
Check out stuff that Betrand Meyer, Eiffels dad, has written - he pretty much beat hell out of software mishaps to show how he'd tried to compensate for those issues in Eiffel. -- here's one of the best known -- https://archive.eiffel.com/doc/manuals/technology/contract/a...
Hmmm - I've always thought you could see the effect of the vertical stack every time you were out for a walk and found yourself four feet in the air with your brain going "My bad dude - stick, not a snake "- I only read the linked article, but on the surface, it seems like imputing magical properties to the stack it doesn't need - if you just look at it as a mechanism for allowing certain signals to move across the hierarchy because they are extremely beneficial in bundling the wider range of information that's rising up from the broader lower hierarchical layers. My 2c admittedly.
We're completely invaded in NH - best trick I've learned is leave a shallow pan of water in kitchen with desk light shining on it all nite - rinse and repeat every day :-(
Yeah, if what Quinn says comes as any great surprise to you, well, that's just sad. Pity too, she can be funny as hell - and I'd have probably signed up for NYT instead of just going in anon
wowzers - cool - I learned it in the early '80s from a guy who taught me - remembering some weird computers that usta be lying about, not at all surprised to see the barb directed at univac then - I believe it was aimed at Sperry in my day :-)
The NYT shows an ever more feeble understanding of brand and reputation in this age - Quinn's writing and public commentary really shouldn't come as a surprise to them, any more than going to a talk by Brenda Laurel and having her call out questions with 'Yo' and shoot down bad questions with 'That's a stupid * question' - I don't always agree with her, but I have always seen her as authentic (both of those fine women, actually)
The classic joke in the '80s was that a systran project meant to translate Russian into English wasn't too good outside of purely technical work - The proverb 'The spirit is willing but the flesh is weak' became 'The vodka is good but the meat has gone bad.'
Agree 100% if you're implying Common LISP/CLOS as the most underrated.... it's fun to show people examples from the Art of the Metaobject Protocol and watch their heads explode :-)
Having a tech background does allow you to better understand what is going on and how it might provide value - conversely, a finance background is probably a good thing to have if you're looking at the IPO for a B2B financial services company
There's a reason we put bindings on books - to protect them - I know what some of my older math books from the '80s look like, and if they'd been spine in, they'd be dead - to me this translates to 'books are things and I don't give a damn about their actual contents' I think I might seriously walk out of someone's house if I saw this.
Paraphrasing Mr Adams - “The story so far:
In the beginning LISP was created, .... and then Guy wrote the Common LISP Bible. This made a lot of people very angry and been widely regarded as a bad move.” - or, Common Lisp - the LISP communities attempt at a non-proliferation treaty with itself. :-)