Look, kid, I got a master's degree in a related field. I have a vague idea of what's possible. It is true that there are ways to make the world better with genetic engineering. However, nature is a mess of stuff that was never intended to be purposefully edited and the problems are virtually endless and essentially intractable. It's nearly impossible to make a drug that won't kill some of the people you give it to, and genetic edits are even harder because they are various hijacked viruses that cause endless problems.

WIth something like machine intelligence or nanotechnology, you're talking about systems that are fully artificial and vastly more predictable. Also, nobody (or very few due to accidents) has to die to advance the field. Thus it advances thousands of times faster than biological engineering.

Now, there are uses for genetic engineering. Specifically, in editing bacteria or other very small, very rapidly replicating creatures to make something useful for humans. Also in other purposes related to food and medicine production. But this is always going to be limited by the limits of biology.

Yes, that would be the obvious eventual solution, there just does not yet exist a way to do that.

The optical interconnects are to carry large quantities of information with low latency between chips that have to be physically spread apart a number of meters because there are thousands or millions of them. There would also be on chip busses that are electrical.

Perhaps - why not integrate the memory and the processing into the same physical chip? Especially for things like artificial neural nets, where essentially the algorithm is accept inputs from connected synapses, generate output from inputs + synapse weight, transmit output.

On every tick the synapse weight variable is accessed for every node. So you essentially have a chip where you must access every memory location every tick, in parallel across millions or billions of artificial neurons. You also have insane network traffic - you probably will need direct optical interconnects.

Per wikipedia, "Energy density is the amount of energy stored in a given system or region of space per unit volume or mass". So other people aren't confused, they are just choosing the more relevant of the 2 measures. The reason it's generally more relevant is because weight more directly determines the energy required to carry the energy storage medium. For phones, obviously both a heavier weight and a bigger size are disadvantages, and the ~1000 cycles you get from lithium-cobalt is close to the ~3 year life of a phone anyway. So it's probably the optimal solution.

For hoverboards, well, making it heavier might actually improve its stability, so you have a point there.

I disagree. I think you're completely wrong and you should know it if you think about it for a few seconds.

1. Genetic engineering takes 28 years minimum to get even initial results.
2. "Peak compute" is for 2d chips that aren't tightly integrated. If we tried to make computers with similar memory and processing power to the human brain, we'd need many square kilometers of total chip area and to use optical interconnects. That's a whole additional set of improvements - making sets of chips that can be manufactured cheaply and can be tightly integrated with each other.

#2 gives you results in months or a few years at most. Eventually #2 should lead to emulated human intelligence or AI, which would be soon be smarter than human biology can even reach even in theory...

Hooray? There have been breathless articles about how diamond or CNT or whatever stomps silicon flat for 30 years now. The problem is that silicon is a moving target - it keeps being improved. If CNT or diamond is fundamentally better than the best possible silicon, which it probably is, the only way it can "catch up" is if silicon is improved to it's practical limits. That might be just a few years away - there's talk of the next few die shrinks being the last ones for silicon before physics don't allow any further improvement for 2d silicon wafers. (and 3d has the fundamental problem of trapping heat and much more difficult manufacturing)

Still, this is cool. I wonder if large scale power switch transistors can be a new future use for CNT tech? If they have better current flow and less "on" resistance, superior to silicon, that would be great.

It's not the same energy density, it's about half. This is workable for a car - although it wouldn't have Tesla like range or handling, but you'd get 600k miles before the batteries were under 80% capacity - but impractical for a competitive phone.

Ok, you've changed your argument. I concede this is a massive social problem because the truth is, if several million drivers lose their jobs rapidly over a few years as automation becomes huge (I'd expect it to take about as long as it took for iphone like smartphones to become the overwhelming majority of phones sold - 5 to 10 years, once automation makes economic sense).

But it's not one you can claim won't happen. The costs of the physical hardware in each vehicle won't stop it from taking over. How much does it cost to keep a car driving 24 hours a day now? Say 2 drivers make 40k a year in successive 12 hour shifts, plus the costs of the maintenance. If the automation is 3-5k worth of electronics and servos plus a subscription fee for the update service plus liability insurance...

Ok to be honest I see your point. For a long time it would probably make more sense for services that run commercial vehicles to automate. Individual consumers like yourself would still drive themselves because it would be cheaper, even if trucks and buses and delivery vehicles rapidly all became automated.

The only business model that would make sense for individual consumers to automate is if you got charged _per mile_. So you pay the company that is responsible for the automated system and updates the software and maps and insures it a per mile fee. So it would be proportionally cheaper...no, still cheaper to drive yourself. Huh. That fee would have to be less than the increased liability of driving yourself, and it would not be cheaper for a long time.

Well, (1 and 2) are specific to your situation, not most people's situations. (3) is more a "by definition" thing - if you can't trust the car to find a spot to park in an ordinary lot somewhere and not hit any people or other cars in doing so, it's not really a smart enough vehicle to be considered fully automated. (4) is possible to deal with, it's not the same as public transit because people using your car as a taxi are not anonymous. Also, it could make a stop at a cleaning place before returning to you once it's near time to pick you up for the day.

As for the cost of $3k worth of electronics, no, you're the one misunderstanding auto manufacturer costs. TLDR, they only make 10-15% profit, not 300%. So why does your van's option cost 3k when it's $500 worth of electronics? The reason is both the cost of labor and integration and the bigger factor is you ARE getting shafted on that option. That doesn't mean that if the base cost of the bare van were 25k, that it actually costs only 7500 worth of metal and machining to build the van. It's probably $21000 for the machining and labor and everything else.

Capitalism doesn't work that way in a competitive market, in a competitive market the prices of things tends to drop to the actual total cost + the cost of capital. The "profit" that Toyota earns is mainly a fee for the billions of dollars of capital that Toyota has to have in order to operate.

So no, I'd expect that if the electronics were 3k, it would add 4k to what's call the marginal cost for the manufacturer. Now, the manufacturer also has to somehow divvie up a multi-billion dollar tab for developing autonomy among the vehicles it sells. Maybe the early ones will add $10k or $30k to the cost. The manufacturer also has to set aside money for every autonomous vehicle to fund the insurance policies.

In fact, being realistic here, the manufacturers might be forced to charge a subscription fee for autonomy. The fee would pay for the manufacturer to continue fixing bugs and fine tuning the software for autonomy (and updating the maps it uses), and also would pay for an insurance policy the manufacturer would have to have to pay for the damages caused when it screws up. *

* Obviously the only way the economics of this would work out is if the fee + regular car insurance was less than you pay now for car insurance, which is possible.

Not comparable. The seat belt doesn't increase the value of the car to the buyer, at least in the early days when the general public didn't like to wear them. If your car's autonomous, it's got concrete financial value.

1. You can live farther away and save money on rent and mortgage, doing your first hour or 2 of work as the car drives itself
2. You can save money on gasoline because while you're at work it can go to a charging station and spend a couple hours charging at a robotic charging station
3. You can save money on parking because it can go drive and park somewhere cheaper than right next to where you want to go in an expensive place downtown
4. And the obvious - you can register the car with an Uber like service and have it making money for you as a taxi when you're not driving it.

As I said, actual financial value. More than the cost of the hardware for autonomy. Now, obviously the car itself is a liability as well. It might crash and kill you, it might crash and injure you or need major repairs and the manufacturer might balk on paying.

This is in fact how they already do it to an extent. Nearly all the autonomous vehicles in testing have lidar sensors which do give that kind of shape for modeling. They've got some really cool, lower cost LIDAR sensors like these. http://spectrum.ieee.org/cars-...

As a side note, the actual hardware is doable. 6 or so LIDAR sensors, a dozen or so fixed position camera, a couple of radars, and a computational architecture that probably uses FPGAs and custom ASICS for the machine vision and machine learning and a bunch of ARM processors. Mere thousands of dollars worth of physical hardware at the mass production stage. The high cost is obviously the development.

There probably is a way to determine the difference in road surface. Even if it's not by seeing it ahead of time but just correctly responding when the vehicle hits the black ice patch. Good modeling of the vehicle and modeling of different alternative actions could easily result in better control on such surfaces than human drivers can achieve.

I've got a question for ya. I'm just going to take your statements as valid.

If programmers over 30 are unemployable, where is this limitless pool of 22-30 year old programmers coming from? There's a finite number of even qualified Indians...

If the median age of an employed programmer is 43 according to the BLS, where are they all working?

There aren't enough management jobs for every over 30 programmer to get one. Simple and obvious math because a manager needs at least 8 or so subordinates or they aren't really a manager...

I suspect the actual truth is more that there are plenty of programmers who are much older - I have met some personally - but that in the fishbowl of the LA startup scene and at the elite campuses of facebook and google, they can afford to cherry pick. They can only afford to cherry pick because there must be a never ending stream of newcomers to LA seeking their fortune. Sort of how nearly all women alive are outside the parameters to be a movie star (since you gotta be young, skinny, beautiful, AND be able to act) but the constant flood of them into Hollywood means they don't have to employ female actresses over 30.

It depends on what the goal is. If you're talking about educational efficiency, it makes a ton more sense. It makes a ton more sense to train people for 6 months intensively, with a repeat in 5-10 years, if it makes them 80 or 90% as good as someone you trained for 4-5 years. I have that 5 years worth of training, maybe that makes me a better engineer, maybe it doesn't. But if I could be 80% as good as I am now, for 80% of the pay, and I didn't have to waste 5 years in school, I'd have gone that route.