Unfortunately, the experiment came to an abrupt end when they threw "ET: The Extra Terrestrial" at the AI, whereupon after an hour of trying different tactics the AI decided that the only way to win was to send a power surge through the system, frying the only working Atari 2600 the researchers could dig up.
This still classifies the AI as coming up with the best solution to the game ever implemented.
Yaz
Or, "Don't take life too seriously... it's not like it's permanent."
At that age it's usually not a problem, you're far more likely to do something reckless and stupid that will have consequences for the rest of your life. I'd at the very least temper it with a bit of "Enjoy today, plan for tomorrow". Sure, life might throw you a curve ball but if act like every day is your last the odds are pretty good that you're wrong and have to live with yesterday.
But let's be serious. The placebo effect is one of the most effective thing in medical problems. The problem with it is that if you don't believe in it, it no longer works. Building false theories that makes sense for most people is therefore a skill that can be much more effective than finding real cures.
And the "anti-placebo" effect if you know doctors and nurses are liars and frauds so you think the actual treatment which has a pretty good track record is just more astrology/homeopathy/placebo bullshit? I mean you have to have a rather big medical community that knows this is as good as sugar pills. And you have to quite often tell that truth to limit the resources taken away from actual medical treatment to spend on placebo. Yes, the truth can be tough to deal with. No, having our real doctors pushing snake oil and superstition won't help. Now if we were talking about better psychological care to people suffering from severe physical conditions I'd be all for that, but not this.
The question is... If you are in your own home, does the robot count as a bartender, or is it an appliance? My guess is the latter, the responsibility belongs to the operator.
Liquor licenses apply just to the sale of alcohol, if I'm at a private party and mix a round of drinks I don't need to follow any regulations except those that generally apply like serving alcohol to minors. And if a minor orders it from the robot, I shouldn't be in any more trouble than if they go to my fridge and grab one. I guess they could require "alcohol lockers" the way they do "gun lockers" around here, but we're not there yet.
Do we have alcohol vending machines now?
Actually when I studied in Germany in the school cafeteria they had beer in the soda vending machine. Good times.
Like many other administration chores, the key management needs almost an expert system to deal with the daily operations for the non-caring, lazy, or just "regular" people.
And the "expert" system most choose is simply having an account - everyting from e-mail accounts to forum accounts to social media accounts. The users keep their password safe - that's securing the endpoints - and then you trust the system to deliver the email to the recipient and not anybody else. Because if you're handing over the keys to a third party, you might as well hand over the communication too.
Crypto is hard to get right. It's hard for the average person to know what ciphers or tools to use and which are just snake oil. It's hard to implement correctly so that it is secure. New ciphers are written by people who have a lot of experience in breaking the old ones. As the old guard ages out, I don't see the same depth of interest in the next generation. With crypto, there's no quick fix, and the new hotness doesn't come overnight.
Crypto is easy. Ciphers are easy. Here's a key you can use it to sign and verify messages, open and seal envelopes.
Using crypto is hard. People lose keys, forget passwords, don't transmit keys in a secure way, don't store keys in a secure way, revoking keys, checking for revocation, using third party services like webmail and so on. Strong crypto is like losing your house key and being told that sucks, but since it's an impenetrable bunker with an unpickable lock there's nothing you can do but start from scratch.
People want recovery options. If my house burns down to the ground and I escape with no passport, no driver's license, no identification of any kind the government will get me a new one. Work will find a way to get me a new access badge and key fob. That's why all those ways to recover your account exist, they're not necessary per se and you don't have to answer the security questions seriously. But when you have fucked up big and the answer is just gibberish you're pretty screwed. That's why people answer those with actual facts.
I very much approve of reading the actual papers. However...
Scientific papers are usually dry and hard to read.
I agree -- however, I find most people with some background in at least one science can at least glean something from reading the abstract, and hopefully some bits and pieces of the statistical analysis (something I admittedly wish I had better background in. If I could afford the time and money to go back to University, I'd love to take stats and philosophy of science).
If papers come to conflicting conclusions, it's hard to figure out which is right. If you're in the field, you read the papers (or at least glance at the abstracts), and have a good sense of which studies have been confirmed and which disproved.
I think part of the problem for many peoples that the state of knowledge in science isn't a binary proposition. In much of science, the answer to 'Is X true?" boils down to five possibilities: 'yes', 'yes with caveats', 'no', 'no with caveats', and 'uncertain' ("more research in this area is required"). So if you're seeing only a few studies, and they seem to be contradictory, the conclusion you need to take is simply "this area requires more research".
And that is my problem with how most people approach science. They see one study, and say 'Science says X!", when in reality, it's really just one study that says X. Unless you have a massive body of scientific work behind a concept (such as evolution, or gravitation), one can't really make any claims as to what "science" says. Consequently, you also shouldn't be disappointed if future research on a new or lightly researched area of science later produces a paper with a contradictory view -- you can't feel that this means that "Science was wrong!". Science is seldom, if ever, "wrong" -- but how much importance people put into preliminary/early/initial results can certainly make them mistakenly feel that way.
I'm somewhat reminded of how people with multiple sclerosis reacted to Dr. Paolo Zamboni's "liberation therapy". Here was a medical doctor who produced a paper where he looked at the neck veins of a group of people with MS, found they had some narrowing of the veins and iron deposits in the brain, and came up with an angioplasty procedure to open these veins up, believing that MS was caused by "chronic cerebrospinal venous insufficiency" (basically, insufficient blood drainage from the brain). He tried it on his own wife, she subjectively said she felt a bit better, and suddenly MS sufferers around the world were flying to third-world countries to have this done to them (for a fee, of course), and in some countries (like Canada) were begging their national governments to bring the procedure on-shore and to make it part of the social healthcare system.
Unfortunately, Dr. Zamboni's research was deeply flawed. Firstly, his study wasn't "blinded". It also didn't have a comparison group -- he didn't even look for vein narrowing in non-MS populations. Thirdly, he didn't disclose that he had financial ties to a company that made equipment to treat the condition he had "discovered". These are all problematic, but IMO the worst was really the lack of a comparison group for control purposes. As two further studies have shown, the type of vein narrowing Dr. Zamboni detected are equally prevalent in both people with MS and people without MS.
Now MS is a terrible disease. People who suffer from MS live in a sort of quiet bravery, in constant struggle against their condition, and with a lot of hope for a cure. I hope one is found. Unfortunately, all too many of them jumped on this one, and got ill-advised procedures done, which in some cases has led to a worsening of their symptoms, and even death. Damage has been done, all because one paper made a lot of hopeful people jump up and say "Science says X!", when really all that "science" should be saying (and what most scientists in this area actually said) was "this is a potentially interesting result -- let's do more research in this area to see if it goes anywhere".
Truth be told, the vast bulk of science output tends to come down to "more research in this area is needed". That's what the "common man on the street" misses when they see a "result" in a single paper. Sure research needs to be taken only as a stating point for more research -- and not an end-point recommending what people need to do. As a process, science requires a lot of time to really get to the point where it can say 'X is true (with caveats)'.
(As an aside, one general scientific area that I'm interested in is negative results: science that doesn't work. The kind of science where the preliminary result says "we hypothesized this might be true, but we now know it isn't". Not enough of this sort of science seems to happen (or get good funding/publicity) anymore, unless you're disproving someone else original, positive result. The negative result is extremely scientifically honest, and sets up some useful boundaries for what we think might be true. Unfortunately, even scientists want to be the people who discover the next big thing, and many funding organizations want to fund science that leads directly to some result they can sell, making it harder to do science that discovers where the edges of knowledge are.)
Yaz
(For the record, educated as a scientist, but not currently working in primary research)
It would probably look something like this with the tagline "Encryption? I'll break right through that defence."
The rest of the world don't want products with official US backdoors though. So you'll have a very hard time selling anything US made abroad and you'd have to ban foreign imports that don't comply with your backdoor policy. Probably also all second hand private imports like eBay. And open source. If the NSA didn't cost the US enough money already, it will after that. I remember a time when you had to fight to get non-crippled crypto out of the US, only 40 bits for us schmucks. I guess now you'll have to fight to get non-crippled crypto back in...
PewDiePie is #1 in subscribers on YouTube with 35 million and played a major role as himself in two episodes on South Park, that you've never heard of him probably means your interests don't overlap much. Just like I'm sure there's some big Bollywood celebrities that I have no idea who is, no idea who the other guy is though or if he's got any claim to fame outside his game but PewDiePie certainly qualifies, he's not a superstar but I know many celebrities with less claim to fame.
Enforcement doesn't have to be perfect to be worthwhile. If you look at things like land mines and cluster bombs they have become politically very difficult for developed nations to use, and seen as a sign that the user is evil. I'm sure in the future there will be more killer robots, but you probably won't see most countries creating squads of them or using them too openly.
Doubtful, the reason they've become politically very difficult is not their use during war it's that they last well beyond the war. That there's collateral damage might not be popular, but between enemies using civilians as human shields, bad intel, accidents, misunderstandings and simply being caught in the line of fire it happens but nobody quits, just like you don't end the justice system after putting an innocent man in jail. It's a bit harder to excuse why a 5yo playing out in a field got blown to bits 10 years after the war ended. As long as the killer robots stand down when the war is over, I don't think they'll suffer the same backlash.
My understanding of those robot turrets is that they can identify human shaped targets and lock on, but they can't tell friend or foe so their default operating mode is to wait for an operator to give a fire order by feeding the video stream back to a console. They can be left in full auto mode in case of all out attack but in that mode they a just an area denial weapon, more technology than a land mine but no less indiscriminate.
So although they are a robotic weapon system with the ability to decide whether or not to fire by itself, it's not what most people think about when they talk about a fully autonomous weapon system in which a system can make strategic decisions about how to complete an arbitrary objective.
Until you implement some kind of IFF, for example it sends a directed encrypted radio ping that you'd better send a pong back. Or you implement sensors so it will sound warnings, don't shoot you if you put your hands in the air - it works on consoles. Dynamic kill zones by remote control is also a lot more than a land mine, you can for example put it in ambush mode where you let them get close before you open fire and they'll have a helluva problem getting out of range.
Of course it still wouldn't discriminate between civilians and foes, but if you're not with us... Then you miniturize it and put it on top of one of those robot dogs. Then you put in a pack and teach it the basics of covering fire, flanking and such. Sounds to me like you soon have the basics of a robotic assult troop. You just designate a position on the map as enemy territory, accept the risks of collateral and they'll cease it and subdue any resistance for you.
The second part is the important one. Neurones in the human brain have an average of 7,000 connections to other neurones. That's basically impossible to do on a silicon die, where you only have two dimensions to play with and paths can't cross - you end up needing to build very complex networks-on-chip to get anywhere close.
We can implement that with a fairly simple grid with pass-through, say you have a grid (x,y) and (1,3) wants to pass it to (4,7), we can just pass it right to (2,3). It can do a simple compare(x=2, y=3) not for us, if x > 2 pass right else if if y > 3 pass down until we hit the right grid node. What's hairy is understanding how to program it into doing anything useful.
No, but you're doing real-time 3D vision and context-sensitive pattern recognition with an amazing degree of parallelism any time you got your eyes open. Cue the "I'm blind, you insensitive clod" jokes. Do you know what the processing speed of a neuron is? Roughly 0.2 kHz, give or take a little depending on type. The Apple I from 1976 runs circles around a neuron with a 1 MHz processing speed. The difference? We have a *lot* of neurons with a *lot* of connections.
The brain proves we can do a lot more of extremely low power massively parallel processing, we've only gotten started with GPUs. They have thousands of shaders, the brain got 100 billion neurons. And yes, a tiiiiiny processing unit is a better analogy than a transistor, they're much more than that. Individually they're not much, but we make up for them in volume.
All seems condemned in the long run to approximate a state akin to Gaussian noise. -- James Martin
