Forgot your password?
typodupeerror

Comment: Re:Ummm.... (Score 4, Insightful) 167

I have a different theory. His comic appeal to people who merely believe themselves to be above average.

...but it can't appeal to people who really are above average, because it doesn't appeal to you! Right?

So, can you recommend any webcomics that appeal to people who bolster their own sense of superiority by accusing others of feeling superior, and then mocking them for it? Maybe something with "Projection" in the title...

Comment: Bah, character-set ignorance. (Score 1, Informative) 35

by jeffb (2.718) (#47796007) Attached to: Iceland Raises Volcano Aviation Alert Again

I feel embarrassed every time I see an English-language site render this as "Bardarbunga", when that "d" should be "th". Yes, the letter "eth" looks like a lowercase d with a crossbar and erectile dysfunction, but it's pronounced like "th".

They should render the a-with-diacritic as "au", too. (Maybe even take the "g" to a "k".) But while there's a long and stupid tradition of dropping diacritics without rewriting the vowel, there's no damn excuse for getting it this badly wrong when you've got to replace a letter that simply doesn't exist in your target alphabet.

Comment: Re:Baby steps (Score 1) 282

by jeffb (2.718) (#47795923) Attached to: Hidden Obstacles For Google's Self-Driving Cars

I did, and I chose that word carefully. If we have a fleet that accumulates experience over time, I expect its performance to improve by a compounding percentage over time. That fits the precise definition of "exponential growth".

You may disagree with my optimistic outlook, but I stand by my choice of words. If anything, perhaps I should have said "exponentially more situations over time", but I think that actually dilutes the point a bit. All the same, I accept that reasonable people may disagree with my wording.

Comment: Re:pulsing (Score 1) 72

by jeffb (2.718) (#47795413) Attached to: Magnetic Stimulation Boosts Memory In Humans

Nope. "High tension" means high voltage, which is done so they can get away with low current, which means low magnetic (b-field) coupling. And the rate-of-change is also low, because it's a 60hz sine wave, not an aggressive fast-rise-time pulse. Finally, it's a line (approximately), not a coil -- the magnetic flux is proportional to the number of turns, and for a transmission line the number of turns is 1.

Now, the electric field effects from high-tension lines are another matter entirely.

Comment: Re:Baby steps (Score 1) 282

by jeffb (2.718) (#47793191) Attached to: Hidden Obstacles For Google's Self-Driving Cars

Exactly.

As TFA illustrates, Google's autonomous cars aren't yet ready to drive on every road in the country. But the same could be said of many, many people who are driving those very roads at this very moment.

We don't need a system that always outperforms the very best human drivers. Even if it only outperforms 95% of human drivers, it will still make the roads safer for everyone -- even that 5%, because they'll be at less risk from the 95%.

And remember, there are an awful lot of people in that lower 95% -- heck, in the lower 50%, or the lower 10% -- who are absolutely convinced that they're in the elite 5%.

Comment: Re:Baby steps (Score 1) 282

by jeffb (2.718) (#47793179) Attached to: Hidden Obstacles For Google's Self-Driving Cars

I think it's the same as a lot of other AI applications. Because autonomous cars can be so much better than human drivers in so many ways -- more and better senses, faster "reflexes", less susceptibility to distraction/impairment/fatigue, inter-vehicle communication, learning from the experience of an entire fleet (including the vehicles that "died" in serious accidents) -- they may never need to be ANYWHERE CLOSE to "human-level AI".

As a very simplistic analogy, consider anti-lock brakes. It takes humans a huge amount of experience, along with lightning reflexes, to deal with rapidly-changing road conditions. Machines don't have anywhere near the human capacity to attend to weather conditions (and weather forecasts), interpret subtle changes in the appearance of the road surface, remember which local roads are prone to ice or standing water, and so forth. But because machines can observe the actual behavior of each wheel of a car and modulate the brakes independently in response, much more quickly than any human's reflexes, they can reliably outperform even experienced and attentive human drivers. And that's with 1970's mechanical technology.

Comment: Re:Baby steps (Score 2) 282

by jeffb (2.718) (#47793143) Attached to: Hidden Obstacles For Google's Self-Driving Cars

I'm sorry, but "there will always be situations where a human performs better than AI" sounds an awful lot like "I won't wear a seat belt because it might trap me in a burning car".

I really don't mean to be a jerk about this, but didn't you actually just utter pretty much those exact words?! -- from earlier in your post:

I'm sure that there will always be a few situations where a skilled human driver will make better decisions, and produce better outcomes, than standard automation.

So, given that you said that and that you were "sure" of that statement, does that mean you also don't wear a seat belt because you're afraid of dying in a car fire? Just wonderin'. :)

My point was this:

There are a few situations where you're worse off wearing a seat belt than not wearing one. There are people who have died because they were wearing a seat belt.

Those situations are immensely rarer than the situations in which a seat belt will save your life, and since "accidents" are inherently unpredictable, you can't tell in advance when you should or shouldn't wear a seat belt.

Given these facts, it's really really stupid not to wear a seat belt, even though there are some situations in which it might harm you.

Similarly, self-driving cars will eventually reach a point where they'll sometimes kill you, but far, far more often save your life. At that point, avoiding them because you're afraid of the rarest scenario will be an equally stupid decision. It's one that people will make, though, because people are demonstrably terrible at this kind of risk evaluation.

Comment: Re:Baby steps (Score 2, Insightful) 282

by jeffb (2.718) (#47791823) Attached to: Hidden Obstacles For Google's Self-Driving Cars

It will be decades before these vehicles can handle real life situations. You will need AI that can improvise as well as a human. Good luck with that.

I'm sure that there will always be a few situations where a skilled human driver will make better decisions, and produce better outcomes, than standard automation.

I'm equally sure that there will be exponentially more situations where standard automation will make better decisions, and produce better outcomes, than average (or even well above-average) human drivers.

I'm sorry, but "there will always be situations where a human performs better than AI" sounds an awful lot like "I won't wear a seat belt because it might trap me in a burning car". It's not wrong, but it is foolish, and it's a poor decision.

Comment: Re:You know what else produces magnetic fields? (Score 1) 72

by jeffb (2.718) (#47788417) Attached to: Magnetic Stimulation Boosts Memory In Humans

In related news, since staring at the sun through a large telescope can damage your retina, DON'T DARE LOOK AT YOUR PHONE SCREEN! It emits DEADLY PHOTONS of electromagnetic RADIATION!!11!

Something to keep in mind should you ever decide to timidly stick your head out of your cave.

There's no such thing as a free lunch. -- Milton Friendman

Working...