Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Gemini Robotics

deepmind.google

Post reply on HN
Print view View original

571–580 of 588 posts

Re: Gemini Robotics

#571

Post by YeGoblynQueenne »

calmbonsai wrote: post #262

The issues with all of these robotic demo videos is "repeatability" and "noise tolerance". Can these spatial reasoning and end-effector tasks be reliably repeated or are we just looking at the robotic equivalent of "trick-shots" where the success percentile is in the single digits? I'd say Okura and Vinci are the current leaders in multi-axis multi-arm end-effectors and they have nothing like this.

No, it's the WYSIWYG model of robotics: the robot can do exactly what you see in the demo

e.g. the robot can put that particular fake banana in that particular bowl placed in that particular location. Give it another banana and another bowl and run for cover.

Re: Gemini Robotics

#572

Post by YeGoblynQueenne »

refulgentis wrote: post #25
fusionadvocate wrote: post #15

Robotics has been trying the same ideas for the last who knows how many years. They still believe it will work now, somehow. Perhaps it goes beyond the brightest minds at Google that people can grasp things with their eyes closed. That we don't need to see to grasp. But designing good robots with tactile sensors is too much for our top researchers.

I'm a bit confused. Ex-Googler so maybe I'm just spoiled by access to non-public information? But I'm fairly sure there's plenty of public material of Google robots gripping. Is it a play on words? Like, "we don't need to see to grasp", but obviously that isn't what you meant. We just don't need to if we saw it previously, and it hadn't moved. EDIT: It does look like the video demonstrates this, including why you can…

>> Like, "we don't need to see to grasp", but obviously that isn't what you meant. We just don't need to if we saw it previously, and it hadn't moved.

Blind people can grasp.

Stretch goal: small kids can walk on their hands.

Re: Gemini Robotics

#573

Post by YeGoblynQueenne »

YeGoblynQueenne wrote: post #572
refulgentis wrote: post #25

Earlier quoted context omitted.

I'm a bit confused. Ex-Googler so maybe I'm just spoiled by access to non-public information? But I'm fairly sure there's plenty of public material of Google robots gripping. Is it a play on words? Like, "we don't need to see to grasp", but obviously that isn't what you meant. We just don't need to if we saw it previously, and it hadn't moved. EDIT: It does look like the video demonstrates this, including why you can…

>> Like, "we don't need to see to grasp", but obviously that isn't what you meant. We just don't need to if we saw it previously, and it hadn't moved. Blind people can grasp. Stretch goal: small kids can walk on their hands.

Ahahahaha.

Sorry, but I just burst out laughing at my own comment, when I considered the technical difficulties in trying to teach a robot to handle the change of context needed to balance on its hands, rather than its feet, let alone walk around on them. Ahaha.

Re: Gemini Robotics

#574

Post by gertlex »

robotresearcher wrote: post #556
gertlex wrote: post #151

Earlier quoted context omitted.

It's somewhat debatable based on lack of results that have made it to market. In addition to the other comment mentioning Boston Dynamics, they are also the employers of a lot of folks that were formerly at the Open Source Robotics Foundation(?) (OSRF) (it's more complicated than that) which is behind the ROS1/ROS2 framework that are widely (not universally) used; They also have an internal division or whatever, Intr…

The OSRF people went to Intrinsic, not Google proper.

Thanks for the correction.

Discussing with a coworker yesterday, they also pointed out the distinction I'd missed, which is that there are separate robotics groups doing research (Gemini) and applied work (Intrinsic? physicalintelligence.company; these are under the Alphabet or X projects umbrella, I infer? Really haven't paid attention.)

Re: Gemini Robotics

#575

Post by nthingtohide »

fc417fc802 wrote: post #561
nthingtohide wrote: post #527

Earlier quoted context omitted.

Humans not agreeing has more to do with the fact that humans are called upon to take decisions with imperfect information under time constraints. If each human could pause the state of the world and gather all information and then decide, they would act humanely

The existence of trolley problems is a counterpoint to your comment.

There is so much literature on Trolley problem that they will come out enlightened on so much stuff.

Re: Gemini Robotics

#576

Post by nthingtohide »

echoangle wrote: post #529
nthingtohide wrote: post #527

Earlier quoted context omitted.

Humans not agreeing has more to do with the fact that humans are called upon to take decisions with imperfect information under time constraints. If each human could pause the state of the world and gather all information and then decide, they would act humanely

Not at all. Even if you have a completely hypothetical situation with well defined circumstances, the responses will be all over the place. Just think of abortion, wars or legalizing drugs. People disagree completely over those because nobody agrees which choice would be the moral one.

You are suggesting today's state of affairs to remain the same after people do the pause the world literature survey. This is exactly opposite to my point of view. At least they will be well informed as to why their views are different.

Re: Gemini Robotics

#577

Post by tsunamifury »

echelon wrote: post #355
tsunamifury wrote: post #304

Earlier quoted context omitted.

Google has been looking for post-ad post-search revenue for almost a decade now. They certainly won't dominate forever and have several signals flashing red for a few years now.

YouTube has bigger revenues than Netflix. While the majority of that revenue is from ads, they get it by providing immense value in the form of near-unlimited entertainment. That's just one of their many business units.

Ok. Do you always make these non value comments?

Re: Gemini Robotics

#578

Post by nthingtohide »

root_axis wrote: post #528
nthingtohide wrote: post #526

Earlier quoted context omitted.

All of fiction is a distortion of sorts. Consider Wall-E movie fat people. The AI advancements shown in the movie should transitively imply that biotech, biomedical progress would be so high that we would have solved perfect health by then.

Not really. As history shows, progress in one field of science/engineering/philosophy doesn't necessarily imply progress in others.

Which is because we don't have intelligence on tap. If we did, we would actually put intelligence to use in all the subjects. I would rather put intelligence to use on biology rather than philosphy.

Re: Gemini Robotics

#579

Post by Beijinger »

seanmcdirmid wrote: post #514
Beijinger wrote: post #475

Earlier quoted context omitted.

Well, they are of performing well, this means they might be bad or they might be undervalued. It was asked, what is undervalued? Not what did perform well (and might be overvalued). Bust is unlikely with an ETF. They rebalance without you having to do anything. Most tech might come from China in the future.

They aren't really performing well. Take this one: https://finance.yahoo.com/quote/CQQQ/ It is at the same price it was at in 2019. You can't rebalance them because the techs all boom and bust together, there isn't any point. Still, if you think Chinese techs are going to take off big soon, now is a great time to get in (or at least as an emerging market hedge).

You don't rebalance. The ETF does this for you.

But yes, CQQQ is exactly the one I am going to buy.

Re: Gemini Robotics

#580

Post by namaria »

ninetyninenine wrote: post #389
n_ary wrote: post #291

Earlier quoted context omitted.

When I see pro-AI promoters giving weird analogies that do not fit, I get a feeling that they are either management class(aka MBA types) or are not serious professionals. One key point we must first understand, coding is NOT software engineering or even programming! Writing code is the last bit, a minimal fraction of the job description(unless you are actually indie dev or working for consulting firms). The core task…

Your dismissal of AI as merely a glorified autocomplete tool correctly acknowledges its current limitations—but it reveals an alarming blindness to the aggressively upward trendline of technological progress. Yes, today’s AI primarily simplifies mechanical coding tasks; your assessment of its present role is accurate. However, your argument dangerously ignores the relentless momentum and historical pattern of breakth…

The funny thing about the current AI hype and people echoing it is, if these technologies were as remarkable and promising as you say than they a) don't need the hype; b) you'd realize that hyping them would harm your own self interest because the rational thing to do would be to leverage them quietly while you have an edge on non-believers.

Thus the rabid propaganda is great evidence that these technologies are not revolutionary.

Post reply on HN