Well, there's a few reasons - but I think the biggest help we'll have is that I expect the change (based on current progress) to be gradual. That is to say, we'll have time to adapt and build measures as the technology improves, rather than have to deal with it all at once. In a lot of ways, the change is well in progress already. Someone who wants to do something bad has, through the Internet, many more knowledge and contact resources than they would have had in 1985 (or 1885). The information age has already created security problems, but we've adapted, and we leverage the same technologies to keep us safe.

Another reason, that is a bit more "out there", is that I believe this sort of technology will likely be able to solve a lot of humanity's problems before it gets to "supervillain's assistant" or "self-interested omniscient being" sort of level. And people who generally having their needs met (or perhaps "overmet"... and may also be watched 24/7... yeah...) are less likely to cause problems with their supercomputer access. People may find that they want to play just one more level before they blow something up. And maybe finish their hyper-Doritos.

While they sound kind of hokey, I think there are some credible threats in this "self aware computer gets squirrely" vein - depending on how the AI is developed. If we build an AI based on, say, scanning a brain and recreating it - however that might work - then we end up with a very predictably unpredictable agent. This emergence could be a very "singularity" type event where we go from fairly dumb AI to very smart, world changing AI in a short time. From there on out, it would get hard to predict very fast; a weird mix of scary/great/over that humans may not keep up with.

But if strong AI grows out of, say, a Watson type "oracle" program that just gets smarter over years and years (and this style of development seems more likely), then the kinds of problems I'd expect would be much more comprehensible. Still potentially scary, but likely more manageable.

...and some of those people would want to do bad things. A bad person would be more capable of doing harm when aided by an AI doing planning, co-ordination, or execution. There's no guarantee that AIs on the "other side" would be able to mitigate the new threats (the two things aren't the same difficulty).

I think there's lots of risks associated with the rise of AI (though it doesn't seem that tech is coming all that fast at the moment). That said, there's risks involved with all sorts of new tech. That doesn't mean this is alarmist nonsense; it's worth discussing potential ways to mitigate those risks - but there's also good reason to believe we'll be able to manage those risks as we've managed changes in the past.

Yes, sorry - it depends on what you count as consumption. I think that taxing consumption is the best start for building a practically effective progressive tax, but I didn't really get that out in the comment above, my bad.

Yes a flat sales tax will often be regressive, but that's not what I (or Bill) was imagining when I was talking about a consumption tax. My comment didn't really bring that out though - so, my bad on that.

In any case, I'm imagining a consumption tax that wouldn't hit much of what a lower income person is spending (for food, clothing, housing, child care, etc - while still taxing more "luxury" consumption in those same categories). And I think this is more practical than coming at the same "wealthy people's money" in other ways - it's easier to shelter income and capital than it is to hide spending.

We charge the people doing the labor (income tax) and then *also* charge them on consumption?

From the summary:

I agree that taxation should shift away from taxing labor. It doesn’t make any sense that labor in the United States is taxed so heavily relative to capital.

So he wants move away from taxing labor. Less tax on labor. That will mean less tax for people doing labor. Less labor tax. And to replace that tax, he wants a progressive tax on consumption. Progressive means rich people pay more than poor people (not just absolutely, but as a percentage of their means). Taxes on consumption tend towards being progressive naturally, and you can exempt basic needs to make them more progressive.

Consumption taxes also tend to be more difficult to dodge, because you're buying things from a variety of sellers. Again, the effect of this is normally progressive.

You're concerned that Gates is suggesting something that would hurt poor people. You are exactly backwards - his suggestion is much more "poor people friendly" than competing proposals. The objection to Gates's proposal would normally be that consumption taxes reduce consumption and slow down the economy.

If you could differentiate between lots of breath strengths, I suppose you could have up/down/left/right. But it might be easier to just alternate (ie. you do a right breath, then a down, then a right). You could do stronger breaths to start/do big movements, then proportionally little breaths to fine tune your position. Again, I think with some training you could hit targets pretty well with few breaths (assuming you have good proportional control, which might require a different sensor than a microphone).. but I'm also just wildly speculating, and it would probably vary a lot based on the patient.

Or maybe do "blow=right, suck=down"? I know many patients would be able to suck effectively, but others wouldn't have the right facial control I imagine.

As to literacy: if users could read, you could have a screen of words. If they couldn't, you could have pictures for many of the things people would be trying to say often (I want to sit up, I need a tissue, I'm hungry, turn on the computer, etc..). I don't have good ideas on how you'd approach more advanced communication without some degree of literacy (you're going to run out of space for pictures..), but if you had, say, 30 common items that's at least a start.

Anyway, I think you could make a basic device like this for $25 (or so, with volume). But I have no idea how many people would want it.

I'm not sure you're wrong - I've never used one of these. The only similar devices I've seen used were based on gaze (which is obviously a different animal).

But to be clear, I would do this as an x-y grid - so there'd be an x-blow and they a y-blow, then either a click or more blows to correct (which I don't think would be necessary often with some practice). And, also to be clear, you'd often be picking words/phrases/actions rather than individual letters. I don't know the best potential scheme - but what I can say for sure is that having some feedback (via a screen) would expand the kinds of inputs you'd be able to reliably accept. Instead of 2 "lengths" of breath, you could differentiate between many (and between more lengths of pauses), if you can see where you're at. Whether a grid is the best answer I don't know - but for expert users, I'm 100% sure you could come up with a more efficient scheme than Morse code. And for more limited users (including users who can't read/spell) Morse code is going to be a non-starter.

And also, to be clear, 4 breaths won't spell many words in Morse code. Have you used Morse code? I mean, it gets a lot better if you allow abbreviated words or something... but again you pretty much need a screen for that so you can see what you're doing.

Lol yeah - I get that. What I'm saying is blowing out letters in Morse code isn't the most efficient way to communicate, even when we restrict ourselves to using breathing.

..because a lot of design work has gone into them, and they have more stuff that makes them more usable.

Off the top of my head, I think this device would be slow to use, and would require a fair amount of skill/education (you have to be able to spell words in morse code - fine for many, but a problem if you've got broad developmental issues). Physically, for many users, taking 15 breaths to spell a word is going to be slow and (for some) very tiring. As you get tired and aren't in perfect control of your breathing, even small errors in input would cascade down and the resulting words would be unintelligible. Because Morse code isn't actually well suited to the task.

Again, off the top of my head, it seems like it'd be much faster to choose common words (or pictures) off a screen (a screen wouldn't add much to the cost - probably only a few dollars). Then you could use continuous breaths - perhaps with the device measuring breath velocity when the user is able to moderate that - to alternatively move a cursor x+y (and then a short breath to "click") to choose options off a grid. With some training, I bet you could get most common words in 3 breaths this way. Or, when more choices are required, you could use the same mechanism to press keys on a keyboard (though, again, you'd definitely want word completion). For heavy users without developmental issues and with good breath control, you could build out a shorthand type system that might be fast enough for reasonably paced conversation (using breath length/intensity, and lengths of pauses between).

I'm not saying this isn't a cool thing, and it could certainly be made cheaply (though, again, a very cheap screen would make this a lot easier to use). But it's pretty much the "hello world" of assistive devices. I like when people make new things, and I like effort/attention going to problems like this, but I'm tired of how these articles tend to belittle the work done by others who've approached the same problem. It's not that nobody ever thought of making a simple breath control system before. Most likely everyone who approached the problem started by making a device much like this to test the initial breath control... and then they made fancier ones that worked better, based on the feedback they got from real users. Pretty much any suitable engineer, when presented with this problem and a cost constraint, would be able to make a similar or more usable device.

That's no disrespect to him (he's doing a good thing) - it's disrespect to this kind of breathless reporting. No, some teenager in Britain didn't come up with a way to triple the speed of your internet connection. No, some science fair kid didn't come up with a way to make solar panels 100% more efficient. People seem to love the base story here (teenager shows us all the way), but the stories almost always well overstate their case.

Your definitions are simultaneously pedantic in tone and broadly wrong.

Not all sports are games. Many people would describe (for example) fishing as a sport, but few would call it a game. There's a broad intersection of these categories, but sports are not a subset of games in either a prescriptive or descriptive sense.

Not all games are hobbies. A gladiator may compete in games, but... uh... gladiating was certainly not a hobby. "Hobby" has all sorts of connotations that are not satisfied by many instances of games or gaming. You could call them subsets of "activities" maybe, but certainly not hobby.

I think it's debatable whether competitive gaming is properly called a "sport". From a prescriptive standpoint, some dictionaries give definitions of sport that would be met, other definitions would not be met. Looking at etymologically, you'd assume things would count as "sports" that prescriptive definitions would not consider.

Popularly - descriptively - the term "e-sport" certainly seems to be catching on.

And it's doing so because it's useful in many ways. The substance of these competitions and their supporting organization has a lot in common with "normal" professional sports. There's teams and jerseys and player positions and sponsors. The point of words is to communicate, and calling these competitions sports is communicating a lot of information efficiently, while varying only in one (possibly key) bit - they're not terribly physical competitions (even if they do require a surprising amount of stamina and physical preparation).

Further complicating the matter is that the activities simulated in the game are also often sport-like (though often involving less killing).

If you piss around with Windows 8 for a while, you can basically simulate Windows 7. But for a long time, you'll still bump into horrible garbage - like "you wanted a weird, functionless fullscreen app to view an image file, right"? Very few things are real showstoppers, but lots of stuff is just a little worse - like they abandoned all the little refinements they've made to progressive versions over the years. Little stuff, like the behavior of the "run" dialog. It used to autocomplete well, and seemed to usually know what you wanted. Now it doesn't.

My job has me doing development on a Windows 8 machine - and it's gotten down to very few times a day I say "oh God, really?", but it's taken a lot of tweaking and adapting to get there. And there's literally nothing I actually prefer about 8. Lots of it just evidences horrible testing/design. Like your default start screen has a tile for the "math input editor" or something. That's a very narrow niche app for a desktop, non-touchscreen computer, and it doesn't work the way anyone expects. Many times I've been asked "what the heck does this do?" - and it actually took me a while to figure out. Obviously that doesn't hurt anyone much to have a stupid, useless app - but the same lack of design pervades the whole product.

It's just a half-baked mess, and I think it's earned it's poor reputation very well.

If you've got a gun in a crowd, you don't really need to aim much to hit a lot of people. The tape part was facetious, but not that facetious - the system wouldn't need to be complex at all.

And no, aiming is not really that hard (unless you have a long range, or high precision requirements). Lots of non-genius hobbyists have made systems that identify short range moving targets, point at them and shoot (look on YouTube). It's a "science fair" level project that I could have done (with current tech, anway) in high school. If you actually had someone experienced in machine vision and robotics, you could identify specific people and shoot off their index fingers from 200 meters, but you wouldn't need any special expertise to just shoot some dudes.

And being on a quadcopter actually makes it easier, as you don't have to kludge together a bunch of motors/servos (like the people fooling around with paintball guns on YouTube): you can use the quadcopters movement (the hard parts of which are a solved problem from your perspective) to target primarily by translation. Simply put, you just move left until "that moving thing" is in the middle of your field of vision. And, within 10 or 20 feet, as long as your gun is pointing basically the same direction as your camera, you're done. You probably hit a good percentage of the time.

Now it's quite possible you wouldn't get many shots off with each drone (though I think you would, as long as you put any thought into your gun mount and cartridge management) - but if you had even a modest budget you could make up for that with numbers. Again, it's a high damage, "high terror" attack with modest budget requirements, modest expertise requirements, and low personal commitment.

(The reality is that there's lots of attacks like this that you could cobble together if you were smart - this isn't a new thing with drones. But drones do give you options that are particularly problematic to defend against).

I fully agree. People are going to chase ineffective security measures for a while, but eventually people are just going to be really dangerous in ways that are difficult to defend against.

Like you say, we need a different answer. Hopefully the people of tomorrow come up with one :)

I'm a programmer who has worked on machine vision systems - so, yes I know how to do it, and yes it'd be easy. Sure, sometimes you'd get one of your drones emptying ammo into a floating plastic bag or something - but odds are you'd kill a lot of people too. Lots of non-genius amateurs have done this exact thing in their backyard, to shoot squirrels with paintball guns or whatever. It's easy. And, again, it doesn't matter if some of your drones fail to kill people.

I don't know what you mean by "jam" an automated drone. What, you going to EMP it? Or blind them lasers. That requires absurd precision, and has easy counter measures (eg, a narrow field of view, optically enforced by a tube, or more than one camera, or just a drone that moves fast/eratically and spins, or that isn't really stupid).

Summary: these would be easy to make, and hard to defend against. Really.