The AI and Singularity hype irks me, because I'm genuinely in agreement with Peter Thiel's argument that technological progress is actually decelerating relative to how it was moving 60 years ago once you look beyond the advancements we have experienced in information technology and finance.
True that. 1880-1950 (70 years) took us from train-and-telegraph to jet-and-atom bomb. Horse and carriage was the top-tier tech for personal transportation in 1880, and they were just starting to understand that diseases are often caused by these little blobby creatures called germs. 1950-2020 (70 years) will take us much less distance. Almost everything we're developing seems to be about information processing. If y…
The AI Misinformation Epidemic
211–220 of 286 posts
Re: The AI Misinformation Epidemic
#212From all I can see, nearly all the present AI boils down to one word -- hype. From what is really going on, nearly all of it looks like (1) in some cases a lot of new data, (2) some new, faster processor hardware, e.g., based on graphical processor units although the x86 processors are astoundingly fast anyway, able to manipulate the new data, (3) for what manipulations to do on the data, some tweaks of some of the w…
You give your work a catchy name because you want other people to help develop it, via attention, promotion, word of mouth, etc so other people can find it and fund it. To do so you speak in their terms and their terms are catchy one-liners. Don't be bitter about it, some people just don't have the level of concentration to deal with our specifics.
What is going on in such cases -- learning, neural -- is at best some applied math, but the learning, neural, and intelligence are trying to suggest, hint, hint, that the math is somehow close to what intelligent humans with biological brains do.
The effects of the hints are strong enough to give lots of newsies another of their favorite means of getting eyeballs and clicks, that is, there are big threats, the sky is falling, nearly all humans will lose their jobs, and the robots might take over as in some movies.
Yes, somewhere in a buried paragraph there maybe an admission that the learning is likely nothing like what humans, kitty cats, puppy dogs, whales, or crows do, but that disclaimter, if want to call it that, is hardly a drop of water on the hype inferno.
At one time some IBM publicity got some housewives to believe that iBM's computers were "gigantic electronic human brains".
I know; I know; such hype is not the worst problem in the world and, really, is affecting me hardly at all. But when I was in an AI group at IBM's Watson lab, still I was embarrassed to say artificial intelligence. I knew, yes, that the meaning was supposed to be just a computer doing something that would require human intelligence if a human did it, e.g., play chess well, but that the approach of the AI chess, a big tree search with alpha-beta pruning, etc., was nothing like what humans did playing chess. But with this definition some 0-1 integer linear programming for scheduling at an airline should be far above merely intelligent since it is far above what an unaided human could do -- although humans invented the applied math that is 0-1 integer linear programming.
I call the crucial core technology for my startup not AI but just some applied math. The world is awash in good applied math -- statistical hypothesis testing, the fast Fourier transform, numerically exact matrix inversion (number theory techniques), Wiener filtering, deterministic optimal control, and hundreds more -- all without a single claim of anything intelligent in any of the books, papers, courses, etc.
Re: The AI Misinformation Epidemic
#213Earlier quoted context omitted.
No car? A world where you could walk to the things you need because they're close? Cheap, safe, non-polluting, no maintenance car? Self-driving, self-cleaning car? Automated horse with roof and luggage compartment - no need for paved roads, muscles are quieter than engines?
> Cheap, safe, non-polluting, no maintenance car? We've made tremendous progress in that direction since 1960. What's your point? > A world where you could walk to the things you need because they're close? That's not a technology problem; much of the developed world is like that and more of the world is getting to be like it with progressive urbanization. The US made a deliberate social choice to focus on auto-centr…
That it's easy to imagine things being better...
Re: The AI Misinformation Epidemic
#214Re: The AI Misinformation Epidemic
#215Earlier quoted context omitted.
Sure. The first NBER paper in [1] features the following jewel (p.34): "B. Calibration To quantitatively decompose the contribution of different factors to the growth of shadow banks and fintech firms, we first have to calibrate the model to the conforming loan market data." I can tell you with a straight face that is not normal science. Economists themselves increasingly recognize so-called "calibration" is a farce.
If the paper gets traction and the model specification is indeed not robust, the first thing you'll see in a few months is something like "paper xyz: comment" driving holes in the methodology getting even more traction. Empirical microeconomics is fairly open about methodology flaws and critiques. Also, by the way, you see pretty much the same type of thing in a ton of fMRI neuroscience, medical and psychology studie…
You should start with 'calibration' in economics. No, it is not quite what you (seem to) think it is. No, it is not quite "the same type of thing" as p-hacking and low-powered studies in psychology. (Whose poor reliability, by the way, is almost common knowledge by now.)
Look it up, and see how the sausage is made.
Re: The AI Misinformation Epidemic
#216Earlier quoted context omitted.
Peter Thiel has made some other interesting observations: - that our language about the ‘developed' vs ‘developing' world is excessively bullish about globalization while implicitly pessimistic about technology. - that government has changed from thinking that progress can be achieved via planning, into thinking that it's more just there to watch random forces & statistics evolve the world. This change in mindset awa…
interesting points in this podcast http://www.talkhouse.com/comedian-tim-heidecker-talks-with-a... Adam Curtis argues that we no longer subscribe to grand visions and as a result we're floundering as our internet-induced bubbles send us into every more fragmented echo-chambers We're never going to achieve anything great again...it's the beginning of the heat-death of human progress! I think that the recent idolatry o…
I disagree with that. The majority of opinions in previous generations were discounted. A small percentage of people in power said: "We want this, therefore we will have it, and the rest will be damned".
We look more at the negative effects that our overreaching plans have on groups in the minority and on the environment itself. There is no such thing as a grand vision that doesn't have an unwanted effect on some group. It's much harder to unilaterally enforce a vision when everybody gets to voice their opinion.
Re: The AI Misinformation Epidemic
#217Earlier quoted context omitted.
I think you're really underselling things like 3G data plans which are now worldwide. That didn't exist 10 years ago. Even just the widespread use of SMS has helped medical teams assess and treat diseases in remote areas. When people complain about some lack of technological advancements, it sounds to me like they're just not aware of what we have accomplished. And I am not surprised that Thiel is in this category. H…
There's a big difference between technological advancements and engineering & standard development. Deploying a cell network is an engineering and logistical task, not a feat of invention.
People that say that have no idea how hard it is to make a network large. The complexity of expanding connectivity is our modern feat.
Re: The AI Misinformation Epidemic
#218Earlier quoted context omitted.
Very much so. More and more i feel that he PC and smartphone has just been about recreating the mainframe using smaller components. The laser may well have been the last really new thing to come about.
Alan Kay spoke about it length in many of his talks. For example: https://www.youtube.com/watch?v=gTAghAJcO1o Excellent video. I highly recommend everyone working in or with IT to see it. I think he is right. PCs and smartphone are mostly recreating thing from the past, and not the best things either. However, this isn't because we invented everything there is to invent. It has more to do with how research in computi…
Re: The AI Misinformation Epidemic
#219Earlier quoted context omitted.
Yup. But it may be that some people with enough information and time to watch the stock and relevant data on it literally second by second could make a killing. And some people might be in a position to influence the time of the turning point and stimulate an avalanche of a fall. Broadly, no way do I believe that many people, not even fleets, especially not individuals, will put up with the charging time of a battery…
The charging time of the electric car is overnight in your garage, while you're asleep. So the vast majority of the time for the vast majority of people, that's not a relevant concern. On road trips, 30 minute breaks every few hours aren't the end of the world for most people. For self-driving, there's no way that's going to be held back once it's better than a human driver. The demand is going to be incredibly high,…
> Have you ever driven a Tesla?
No, but I fully accept that a powerful electric car would be a total hoot to drive. So, it would be quiet, smooth as silk, and be the fastest accelerating wheel driven vehicle so far. The key, of course, has been known for 100+ years: A series wound electric motor has, from the view of the math, infinitely large torque at stall, that is, before the armature has started to rotate. Then, before the armature is rotating very fast, the torque is still very high.
Also, get to put a separate motor at each wheel and do without a transmission, drive shaft, differentials, etc. So, get rid of a lot of mechanical parts. Also get rid of the often corroded exhaust system.
The torque and power can get from a carefully designed electric motor now is astounding.
Moreover, there may be some good opportunities to adjust dynamically, many times a second, the motor torque separately at each wheel to eliminate wheel spin, which actually hurts acceleration.
Again, my main concern is charging time: E.g., for summer driving, will want A/C, and that can take a lot of power; in the winter, want heat, and that also takes power.
Do a lot of high performance driving, and a Tesla is to be regarded as a high performance car, accelerating onto Interstate highways, passing cars, trucks, farm equipment on two lane roads, driving at 90 MPH, etc., can use a lot of battery energy. Once when I drove from my home in Maryland to my office at FedEx in Memphis, I drove my Camaro nearly the whole way at about 90 MPH -- didn't get stopped. Once when I was driving at night from Maryland to Indiana to see my fiance, I drove my Firebird nearly the whole way at 90 MPH -- didn't get stopped. I was lusting for pencil beam high speed driving lights, an air horn that could be heard for miles, and a more powerful fuel pump for pulling long hills at 100 MPH at full throttle at about 5000 RPM (Camaro 396, Turbo 400 transmission, second gear, 2.56 rear axle ratio) -- did that daily driving home from the Johns Hopkins University Applied Physics Lab!
When Dad and I went fishing in Arkansas, I did the driving, before dawn, while he slept, and I put our Buick on 100 MPH and kept it there. For people driving near the Rocky Mountains, they can want a lot of battery energy to climb hills.
Mostly just for commuting to work, I've put 200,000+ miles on each of three cars, a high end Camaro, a Buick Turbo T-Type, and a slow one, an SUV.
So, to me, a car means a lot of driving, maybe fast driving. So, that would make me concerned about Tesla charging time. That a Tesla can be considered a high performance car gives me more concern about charging time if that performance were used very much, e.g., as I have used high performance cars in the past.
My other concern about a Tesla would be the lifetime of the battery after a few thousand full rechargings as might be needed for 200,000+ miles of high performance driving.
For high performance driving, I'm lusting for a supercharged Corvette, and I'm happy that all it needs for energy is premium gasoline! :-)!
Re: The AI Misinformation Epidemic
#220Earlier quoted context omitted.
Hard to time that kind of thing. Not worth it when you risk losing your shirt. You can lose an infinite amount of money shorting, whereas going long you might only lose all of what you put in.
don't options solve this? Buy a put at 50, watch it fall, then exercise. Doesn't work out, and you're only out the premium.