I don't see the difference. Is a humanoid robot not a machine? What surprises is that it took so long to achieve this. It took Honda 20 years to get to ASIMO in 2000. I'm not surprised by the fact that 26(!) years later, a humanoid robot finally exceeds human capabilities in basic tasks like running.

Such robot would easily smash the current marathon record as well and the only thing I'd be impressed by, would be the battery capacity tbh.

▲

Cookingboy 17 hours ago | parent | next [-]

But not all machines are the same?

If you were ever trapped on the top floor of a burning building, should we just send a car up there?

Nobody is surprised by advancement like this, but there is still a lot of engineering that went into it, not talking about how it’s now almost consumer grade tech that you can buy.

▲

qayxc 16 hours ago | parent | next [-]

> If you were ever trapped on the top floor of a burning building, should we just send a car up there?

Riddle me this then: what role does a robot that has been built to run real fast play in that scenario?

Sure, a good deal of engineering went into this, but the hard part - solving bipedal locomotion - had been done long before that. Apart from showing off it's not really advancing the field and so there's not too much left to be impressed by (honestly, the battery and actuator efficiency that allowed the thing to do a half marathon was MUCH more impressive, at least to me). It seems to me to be more analogous to car tuning.

▲

throw1234567891 14 hours ago | parent [-]

> what role does a robot that has been built to run real fast play in that scenario?

It will run real fast up the stairs, I guess? Is it a good enough riddle?

	▲	qayxc 4 hours ago \| parent [-]
		The thing has no real arms, hands or carry capacity, so what would be the point? Running up the stairs is the easy part after all. The difficult part in such scenario is lack of orientation due to smoke, obstacles like locked doors or fallen debris, unstable walls and ceilings, locating people and getting them out safely. All of those things require a good amount of autonomy, on-the-spot physical problem solving and judgement. To this effect, the recent DARPA Triage Challenge, which has its final taking place in November this year, seems like a much more thorough and practical demonstration of current capabilities. Something to look forward to if you're interested in the current state-of-the-art w.r.t. real-world applications of robotics.

▲

throw1234567891 14 hours ago | parent | prev [-]

If I'm trapped in a burning building and am at the top floor, send a helicopter, dude. The robot will be fine on the way back (maybe) bu I won't!

▲

paulpauper 17 hours ago | parent | prev [-]

What people find impressive or interesting is subjective, I guess.

▲

qayxc 16 hours ago | parent [-]

Absolutely. I was impressed by ASIMO and Atlas. Running fast, though, doesn't seem all that impressive by comparison, but yeah YMMV.

	▲	hakfoo 7 hours ago \| parent [-]
		It's a very pretty benchmark. Everyone has the same basic objective and measurement norm, so we can see how performance improves year over year, and how different manufacturers handle a specific known goal, even if it's not super-representative of real world use cases. Billion-dollar supercomputers don't spend most of their time running LINPACK, and some probably perform better on specific tasks, but it doesn't make the Top 500 list any less of a thing.