But would a robot chef use an electric mixer or a hand egg beater?
So, not only can we not outthink them anymore, we can’t outrun them, either?!!
Not looking good, peeps, not looking good. 
Don’t worry. Nobody is thinking of arming these machines. Nobody
I tend to disagree. Machines have long surpassed humans in physical tasks. As noted, a self-driving car would easily win a race against a human. So the fact that this robot mimics human legs instead of using wheels doesn’t seem like that big a deal to me. More of a stunt.
I’m a lot more impressed when a machine demonstrates the ability to outperform a human in some mental activity, like playing chess.
I assume you’re joking.
But to set the record straight, Stockfish is currently the strongest chess-playing engine, featuring an estimated ELO rating over 3800, operating far above any human capability.
Computers started outperforming humans in chess decades ago, going back to Deep Blue in the 1980s when it defeated the reigning human chess champion grandmaster.
Yeah, we’re screwed.
I think half marathon is a perfectly fine and sensible long distance run for a human. Full marathons seem a little crazy to me (Farthest I’ve run was about 15 or 16 miles.) And ultramarathoners I truly do not understand.
I think you misunderstood my post. I said I am impressed by machines that can outplay humans in chess. Not I will be impressed be them someday. I’m aware that machines outplaying humans is a thing that’s already happened.
Heheh, as noted above, you’ve got to be joking. But since someone might think you’re not:
The reason it ended before it began was that the AI car didn’t detect it had a tire going down on the warm up lap, and it slid into the barrier.
Deep Blue was mid-1990s. Defeated reigning champion Kasparov in 1997. The first computer to defeat a GM, though, was Deep Thought in 1988, against Bent Larsen, rated 2560 and #39 in the world at the time. (Kasparov beat it when he played it in 1989.)
Must be my day to be misunderstood.
I was talking about a self-driving car winning a race against a human. Not a self-driving car winning a race against a car being driven by a human. I figured that would be obvious in a thread about marathons.
Ummm, yeah. Totally not clear in your original statement. A human on foot would also lose against a human driving a car. I can easily outrun my desktop computer, unless you put it in a vehicle and transport it off.
Gotcha, apologies for the misunderstanding.
Thanks for the correction.
Which makes a great intro to one of my favorite New Yorker cartoons from about the PC/XT era. On my phone so not gonna try to search it up. Easy enough to describe.
Two men are standing and chatting with a desktop PC w CRT monitor in the scene. One says “My computer beat me at chess. But it was no match in kick-boxing.”
Reminds me of a story called Q.U.R. but that’s probably not it
Futurama once featured “Wireless Joe Jackson”, a robot baseball (actually “Blernsball”) player who was nothing but a box on wheels with an arm to hold the bat.
Oh, and I suppose Pitch-o-mat 5000 was just a modified howitzer?
Since Pitch-o-mat 5000 didn’t impart any spin on the ball, it consistently threw knuckleballs.
Walking is far more complicated than chess; the fact that we do it instinctively doesn’t mean that it uses less brainpower. But then, we’ve only been playing chess for 2,000 years; it took us 2 million years to get this good at walking. Give it time.
I still feel it’s essentially a stunt.
I don’t think we’re going to take the technology that allows a robot to walk or run and develop many real world applications for it. In the real world if we need a machine to move around, we put it on wheels.
The only real world application I can see coming out of this is developing robots that can imitate humans, which I feel is a relatively small niche of ai development.
Technology usually tries to develop ways of doing things that are more efficient than the existing way rather than just imitating the existing way. That’s why airplanes don’t flap their wings and boats don’t swim. If we have a task that needs the physical capabilities of a human being, we’ll probably just use a human being. Robots will be designed to have physical capabilities humans lack.
Except for the sexbots obviously.
An autonomous walking robot has to do three separate things: it has to carry out the physical mechanics of walking, it has to perceive its environment, and it has to integrate the two into the supertask of navigation. Those are hard enough for our current level of technology that the three are being worked on separately. So mastering the mechanics of walking is a legitimate first step, no pun intended.