Please create an account to participate in the Slashdot moderation system


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Neural Net Processor... (Score 4, Funny) 186

"To make sense of all of this data, a new onboard computer with more than 40 times the computing power of the previous generation runs the new Tesla-developed neural net for vision, sonar and radar processing software."

So what you're saying is... the cpu is a neural net processor, a learning computer.

Comment Re:self-driving or assisted driving ? (Score 1) 186

The new hardware is for eventual full self-driving. Not autopilot. (Autopilot in airplanes is "assisted driving," too, but is still called autopilot... That's why Tesla chose "autopilot" instead of "self-driving" as the label for their first-generation capability.)

And yeah, there is a warning that goes off if the driver takes their hands of the wheel.

Posters better learn to actually do research before posting. Oh wait, this is the Internet, that'll never happen. :P

Comment Re:Radiation (Score 5, Interesting) 412

It's almost impossible to die of space radiation overdose. The galactic cosmic rays can't kill you via a radiation overdose, they're dose rate is much, MUCH too low.

The only thing with a high enough dose rate is solar particle event. And, in fact, there are very few that are strong enough to kill you (but note, there are winter or thunderstorms that can easily kill you if you're unprotected on Earth). One has occurred, however, in August of 1972, with a dose of about 1 Sievert, but it'd only be that high if your only shielding was a thin space suit ( here's a source for that). If you were inside a capsule or on the surface of Mars (shielded by the yes-still-significant Martian atmosphere), you'd be totally fine. Even 1 Sv not really enough to kill you. You need about 2 Sv to really be in danger of immediate radiation overdose and death. But you could vomit in your spacesuit and suffocate. However, these events are not instantaneous, you'd have a warning and the events occur over a period of an hour or several hours, so you have enough time to get inside or behind a rock or something.

No, it's nearly impossible to die from acute natural radiation overdose in space.

You'll survive the trip. The worry is about an increase in occurrence of cancer when you get back. However, in any case, the risk of cancer from living in space is less than being a smoker. Although, given the huge deal we make about the space radiation issue, you wouldn't know it. You'd think you'd die instantly or something, which just isn't true.

As far as how to deal with it, well Mars' surface has a much lower radiation dose from GCRs and especially solar flares. You're half shielded by the planet itself and secondarily by an average of around 40 grams per square centimeter of CO2 mass, maybe more at lower altitudes. Additionally, just massive amounts of rock or dirt work great. And water is more effective per unit mass.

On the way to Mars, your best bet is to shorten the trip to 90-100 days as Musk suggests, and perhaps use your supplies (water, food, maybe propellant) to shield you from solar particle events. That'd reduce your transit dose to a manageable amount. And you can also use drugs like Amifostine to avoid some of the radiation effects, especially the effects of acute radiation (we're unsure if Amifostine helps for chronic radiation). But once on the surface of Mars, it's possible to reduce the dosage to arbitrarily low levels.

But again, these are long-term health effects, perhaps like you'd see in any kind of hazardous environment. But you'll be able to perform the mission just fine.

Comment Re:Seems Promising, but Let's Not Get Too Excited (Score 5, Insightful) 170

I'm all for replications, HOWEVER: This isn't psychology or medicine. If a single, transparent, well-documented study shows that volcanic rock (of a common and well-characterized type) quickly locks up CO2, then it's not a fluke.

Science works differently in different fields because some things are easier to fundamentally understand, even with a sample size of n=1, than others where fundamental understanding is basically non-existent (i.e. we don't actually know how the mind works on a fundamental level) and you have effects so small (with so many confounding factors) that you need n=1000 to have any hope at statistical significance.

Comment Re:I predict.... (Score 4, Interesting) 224

Nothing in what you just said suggests why you think the mission will certainly end in death. That's a bold prediction, and not one anyone can make.

For some stupid reason, many people seems to conflate "some risk of a bad thing happening" to "it's a certainty that the bad thing WILL happen and will happen to everyone every single time." It's how NASA's 3% cancer risk from space radiation from a Mars mission becomes "your organs will be boiled! and it's impossible because you'll die during the mission from space radiation." This is just dumb. Space radiation isn't even as bad as smoking, and except for well-characterized and easily mitigated problems with acute doses (the biggest risk is if you have electronics which can't withstand the radiation and so fail, but that's easily engineered away), you're not going to die during the mission at all.

The first Shuttle flight, for instance, had a, I don't know, 10% chance of failure. It worked, because if you have a 10% chance of something happening, that means that you also have a 90% chance of it not happening.

I predict that getting to the surface of Mars in 9 years is much more realistic technologically today than getting to the Moon in 1969 (just 7 years after JFK's 1962 Rice University moon speech) was.

And the first flight probably won't kill anyone.

Comment Small minds, here. (Score 1) 306

I was about to come here to disagree with Bezos, partly about his arguing with the caricature of Musk's position strawmanned by the press. ...but you all have a huge misconception that must be addressed:

1) Getting to space need not be ridiculously expensive, and no, I'm not talking about a tether.
Reusable rockets running on methane (or hydrogen) and oxygen can be quite efficient. Natural gas happens to be the cheapest form of energy on the planet right now, but we can also synthesize it using electricity. And rockets are actually much more efficient than we give them credit for. It principle, with reusable rockets (and perhaps launch assist for the initial portion), we can get the price to orbit down to $10/kg. Perhaps about the same as the cost to fly around the world. We may not get there for a while, but there IS NO PHYSICAL REASON why cost can't get this low. This seems outrageous now because we throw the whole rocket away each time and so it appears unyieldingly expensive at 3 orders of magnitude higher ($10,000/kg), but Bezos' whole business in spaceflight is to pursue this reusable technology. Mastering reuse (which includes using appropriate materials for the conditions in question and developing appropriate automation) really could reduce cost that much.

2) You can get iron-nickel alloys in space that are already pre-refined.

3) Putting a Gigawatt solar array in space /someday/(when prices are lower) is not insane, even if it is today. The energy it generates will exceed the energy needed to launch it in a few days or weeks if properly designed.

I actually disagree with Bezos to some extent. But let's have the conversion start with some facts.

Comment Re:Automation is a GOOD SIDE EFFECT of minimum wag (Score 2) 1023

In all seriousness, what exactly was it that you were trying to say? Are you for minimum wage, no minimum wage, more automation? I couldn't follow.

For example, what does this mean?

Again, I see automation in response to a wage hike as a good thing. Ultimately, provided we maintain full employment, this will help everyone. Given our modern technology, human labor is worth more than $5/hour even if the workers do not have the bargaining power to get a higher wage. So employing people at below $15/hour in positions that could be automated if they were paid a livable wage is actually a misallocation of human resources.

Are you saying they need more than $15 per hour, or between $5 and $15 or something else?

At least $15/hour.

If we had adjusted minimum wage for per-capita gdp growth (a kind of measure of economic productivity) since the 1960s, it'd be up around $20/hour or so. $15/hour is comfortably below that, so I'm certain a $15/hour minimum wage would not bring devastating inflation or anything like that. We can, as a society, afford to pay people at least $15/hour. And with more automation, we could afford to pay even higher wages.

Comment Automation is a GOOD SIDE EFFECT of minimum wage (Score 5, Insightful) 1023

Automation is a good thing. That a livable minimum wage encourages some companies to automate is also a good thing. We MAY need to use other policies to maintain full employment, but at this point, I don't see why anyone should be making just $15/hour.

A big criticism of a minimum wage is that it's "not a free lunch" and just causes inflation. But if a minimum wage encourages automation, then it actually increases per-person productivity, thus partially paying for itself and keeping a minimum wage from being purely inflationary (there will, of course, be some amount of inflation due to a minimum wage increase, but nowadays a small amount of inflation is actually a good thing).

If we're paying just, say, $2/hour for people to work menial jobs, which is far below a livable wage, then they are, de facto, being subsidized in some other way. For instance, government assistance through subsidized housing, food stamps, etc. Or perhaps they're living off of charitable organizations. Or perhaps they're living off the good will of their family and/or friends. But paying a sub-livable wage is being subsidized in SOME WAY, perhaps even just being taken from that person's health. It's not a society-optimal solution.

In our society, even low-skilled workers' productivity has increased due to technology. But because there are so many low-skilled workers, their bargaining power is low, and thus their wages don't increase. Thus something like a minimum wage is necessary in order for those people to make a livable wage and to not be on foodstamps, etc.

Again, I see automation in response to a wage hike as a good thing. Ultimately, provided we maintain full employment, this will help everyone. Given our modern technology, human labor is worth more than $5/hour even if the workers do not have the bargaining power to get a higher wage. So employing people at below $15/hour in positions that could be automated if they were paid a livable wage is actually a misallocation of human resources. In a sense, by NOT paying workers a livable wage and NOT automating more, companies are, in fact, having their labor subsidized by the rest of society (government, family, friends, charities).

Comment Re:The hype train (Score 4, Insightful) 161

I have a Volt. It charges at night, and you can set the charge time to be whenever you want. If utilities simply charged for time of use (which makes the most sense as it'd be the closest to the true cost), then owners would simply choose to charge whenever the price was lowest. With a Volt, I have a relatively small battery, so while I can choose when to charge at night, I can't easily shift charging to other days. With a 215 mile battery, you can now choose when you want to charge, just like you can choose when you want to fill up, particularly if workplace chargers are installed. This has the effect of actually leveling out the peaks and troughs in supply and demand BETTER than no EVs at all, not concentrating it in a "very small portion of the day."

Also, batteries are "energy accumulating tech." Tesla is building a Gigafactory (the initial portions of which are already operational) that will surpass the current global total Li-Ion production capacity. The Powerwall, and the utility-scale Powerpack (which is actually a big deal), beat basically all other grid-tied battery tech options available today. So if you want to "invest in energy accumulating techs," I suggest you buy Tesla stock.

Comment Re:This isn't a bad thing. (Score 1) 940

It's not about people working harder, I don't know how you got that from my comment. Rich people aren't working any harder, either. It's about the gains through that work being much greater due to amplification from technology.

Another thing: a higher minimum wage would likely bring more people into the workforce since it'd actually be worth it to work more instead of just rely on the social safety net you mentioned.

I did mention "rightwingers," but that does not mean this is purely a left vs right issue. I believe it's more complicated than some ideological axis.

Comment This isn't a bad thing. (Score 4, Interesting) 940

I have been saying for years that an increase in the minimum wage can partly pay for itself by spurring automation. And that's a very good thing, for everyone.

Some business owners might prefer to pay a bunch of people $1/hour to dig a ditch using a shovel, but at $15/hour, you gotta use a backhoe.

I always find it funny when rightwingers complain that a minimum wage increase is simultaneously entirely inflationary AND that it will cause you to lose your job to automation.

I've often thought that we are using far too LITTLE automation, not too much. If burger flipping can be automated, why the heck aren't we automating it? Oh, right, because it's cheaper up-front (but not long-term) to just pay someone a poverty wage.

And it's also always funny to see rightwingers pull out the Luddite critique, i.e. that automation will put us out of jobs, when in fact we've had increasing automation for centuries, now, but not any lower voluntary unemployment. So the Luddite critique is ridiculous when OTHER people use it, but totally fine otherwise...

And then, realize that we had a real minimum wage of about $11/hour in the 1960s, when productivity was FAR lower, when we had far less economic productivity per person. If you adjusted the minimum wage for productivity growth, it'd be over $20/hour right now.

I actually think that by NOT raising the minimum wage, we've stymied technological progress. Yes, there's definitely a limit to how fast you can increase the minimum wage without hitting inflation (or possibly some unemployment), but we're not near that limit with $15/hour.

Slashdot Top Deals

Computer Science is merely the post-Turing decline in formal systems theory.