# What's new, Atlas? (AI, Robotics and tech thread) **URL:** \ **Category:** Miscellaneous and Personal Stuff I Must Share\ **Created:** [November 18, 2017, 2:39am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952 "2017-11-18T02:39:32Z")\ **Posts on this page:** 20\ **Page:** 2
**Author:** ![Machine\_Elf](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@Machine\_Elf](https://boards.straightdope.com/u/Machine_Elf)\ **Post date:** [November 23, 2017, 11:13am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/21 "2017-11-23T11:13:49Z")
> [@Snowboarder\_Bo](#): > > The 15 minute TED Talk is really good and shows older DARPA/BD robots as well as other models, including Atlas. Somehow they skipped over the [Big Dog beta test.](https://www.youtube.com/watch?v=VXJZVZFRFJc)😉 ---
**Author:** ![burpo\_the\_wonder\_mutt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/burpo_the_wonder_mutt/32/365_2.png) [@burpo\_the\_wonder\_mutt](https://boards.straightdope.com/u/burpo_the_wonder_mutt)\ **Post date:** [November 23, 2017, 3:06pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/22 "2017-11-23T15:06:22Z")
> [@Machine\_Elf](#): > > Somehow they skipped over the [Big Dog beta test.](https://www.youtube.com/watch?v=VXJZVZFRFJc)😉 Couldn’t find their Pantomime Horse costume, huh? Cool User Name/subject combo. Isn’t “gremlin” the PC term for **Machine Elf**? 😉 ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [December 9, 2017, 7:12am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/23 "2017-12-09T07:12:05Z")
So, uh, [here’s a thing](https://futurism.com/4-hours-googles-ai-mastered-chess-knowledge-history/): Google’s Deep Mind AI built another AI called AlphaZero. AZ is similar to AlphaGo Zero: AGZ was made to master the game of Go, AZ was made to master chess. Which it did in 4 hours, without any human interaction. > [@](#): > > After being programmed with only the rules of chess (no strategies), in just four hours AlphaZero had mastered the game to the extent it was able to best the highest-rated chess-playing program Stockfish. > > In a series of 100 games against Stockfish, AlphaZero won 25 games while playing as white (with first mover advantage), and picked up three games playing as black. The rest of the contests were draws, with Stockfish recording no wins and AlphaZero no losses. > > “We now know who our new overlord is,” said chess researcher David Kramaley, the CEO of chess science website Chessable. Mr. Kramaley can see beyond the chessboard, however: > [@David Kramaley](#): > > It will no doubt revolutionise the game, but think about how this could be applied outside chess. This algorithm could run cities, continents, universes. Oh, and [here’s the actual paper that’s being peer-reviewed](https://arxiv.org/pdf/1712.01815.pdf). [Even Kurzweil thinks the singularity is less than 40 years away](https://futurism.com/kurzweil-claims-that-the-singularity-will-happen-by-2045/). ---
**Author:** ![Sage\_Rat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sage_rat/32/399_2.png) [@Sage\_Rat](https://boards.straightdope.com/u/Sage_Rat)\ **Post date:** [December 9, 2017, 9:57am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/24 "2017-12-09T09:57:18Z")
For anyone who didn’t see it: [Soft robots that are strong](https://www.npr.org/sections/thetwo-way/2017/11/28/567014007/robot-muscles-inspired-by-origami-lift-1000-times-their-weight) This is quite important for robots that can safely interact with humans. It may also be useful for making them lighter. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 9, 2017, 4:34pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/25 "2017-12-09T16:34:29Z")
Ref AIs taking over chess, here’s a (only sorta serious) bit about robot delivery vehicles taking over a city: [SF Supervisors Briefly Delay Robot Uprising – Lowering the Bar](http://loweringthebar.net/2017/12/sf-supervisors-robot-uprising.html) ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [December 9, 2017, 4:57pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/26 "2017-12-09T16:57:43Z")
[Behold! The AI that can make fake videos!](https://futurism.com/ai-makes-fake-videos-facilitate-end-reality-know-it/) I know the resolution is pretty bad (like a dashcam from 10 years ago), but the images are amazing. If this was offered as evidence in a court case, I doubt the jurors would be able to tell which was real and which was fake. Something like this could make for some troubling developments in the future. ---
**Author:** ![Sage\_Rat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sage_rat/32/399_2.png) [@Sage\_Rat](https://boards.straightdope.com/u/Sage_Rat)\ **Post date:** [December 10, 2017, 11:21am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/27 "2017-12-10T11:21:44Z")
> [@Snowboarder\_Bo](#): > > [Behold! The AI that can make fake videos!](https://futurism.com/ai-makes-fake-videos-facilitate-end-reality-know-it/) > > I know the resolution is pretty bad (like a dashcam from 10 years ago), but the images are amazing. If this was offered as evidence in a court case, I doubt the jurors would be able to tell which was real and which was fake. Something like this could make for some troubling developments in the future. The night one has some strange shadows on the ground, but I only noticed it because I could see the original at the same time. Keeping a careful and well documented chain of custody for videos may soon cease being a thing that only the police do or care about. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 10, 2017, 2:23pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/28 "2017-12-10T14:23:36Z")
In a weird sort of way, the faster we transition through crappy fake pix & vids to high quality fake pix & vids the faster we’ll learn to discount anything that may have flowed through a computer. With all that discounting being applied both in our personal lives and in courts of law and of public opinion. Once enough of us understand that “seeing (through a computer) is totally NOT believable” we’ll be fine. It’s the transition time from “seeing is believing” to then that will be scary. ---
**Author:** ![Wesley\_Clark](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/wesley_clark/32/20581_2.png) [@Wesley\_Clark](https://boards.straightdope.com/u/Wesley_Clark)\ **Post date:** [December 10, 2017, 5:53pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/29 "2017-12-10T17:53:08Z")
I worry about mass unemployment. I think that will be a huge issue in the 2020s but especially the 2030s. On the bright side, maybe economic growth will pick up with all the robotics, and that can be used to fund UBI. But that’ll be a hell of a fight because politicians and the rich will fight UBI tooth and nail. Divide and conquer has been highly effective at keeping those at the bottom rungs of the socioeconomic totem pole divided so far, I fail to see why it’d stop working just because we have robots everywhere. Sadly, I worry we will enter some kind of cyberpunk future where there are robots and machine intelligence everywhere, but tons of us just live third world dirt poor lifestyles and a small minority of people are rich. ---
**Author:** ![Declan](https://avatars.discourse-cdn.com/v4/letter/d/ecae2f/32.png) [@Declan](https://boards.straightdope.com/u/Declan)\ **Post date:** [December 10, 2017, 9:20pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/30 "2017-12-10T21:20:49Z")
> [@Wesley\_Clark](#): > > I worry about mass unemployment. I think that will be a huge issue in the 2020s but especially the 2030s I was thinking about that. Say you can buy a general purpose labor robot for about the price of a leisure toy ( quad, snow mobile, bass boat) and with the right software upgrades or packages, it can do anything from working at a bakery to working on an automotive line. So now you lease it out to a brick and mortar company to do what ever is required as long as the software is available. What do you charge for hourly labor, concidering its going 24/7. ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [December 10, 2017, 10:51pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/31 "2017-12-10T22:51:43Z")
> [@Declan](#): > > I was thinking about that. Say you can buy a general purpose labor robot for about the price of a leisure toy ( quad, snow mobile, bass boat) and with the right software upgrades or packages, it can do anything from working at a bakery to working on an automotive line. So now you lease it out to a brick and mortar company to do what ever is required as long as the software is available. > > What do you charge for hourly labor, concidering its going 24/7. Well, that’s part of the problem’s being explored both by IRL economists and tech people (and on this board in our [What happens when the robots (peacefully) take over?](http://boards.straightdope.com/sdmb/showthread.php?t=637834) thread over in GD. The transition to a post-scarcity civ is going to be, in part, rocky, bumpy and ugly. Post-scarcity is the exact opposite of what we know as the norm for the universe (and what our biology evolved in response to). ---
**Author:** ![burpo\_the\_wonder\_mutt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/burpo_the_wonder_mutt/32/365_2.png) [@burpo\_the\_wonder\_mutt](https://boards.straightdope.com/u/burpo_the_wonder_mutt)\ **Post date:** [December 11, 2017, 12:12am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/32 "2017-12-11T00:12:07Z")
Without sounding like a total näif, or a smart-ass (I’m not nearly into tech/SF as much as you guys), what is the purpose of having humanoid-like robots (other than swearing)? ---
**Author:** ![PastTense](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pasttense/32/14550_2.png) [@PastTense](https://boards.straightdope.com/u/PastTense)\ **Post date:** [December 11, 2017, 1:09am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/33 "2017-12-11T01:09:20Z")
> [@burpo\_the\_wonder\_mutt](#): > > Without sounding like a total näif, or a smart-ass (I’m not nearly into tech/SF as much as you guys), what is the purpose of having humanoid-like robots (other than swearing)? A big consumer product will be sex robots: people prefer having sex with something which looks like a human being than a piece of machinery. ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [December 11, 2017, 1:38am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/34 "2017-12-11T01:38:14Z")
> [@burpo\_the\_wonder\_mutt](#): > > Without sounding like a total näif, or a smart-ass (I’m not nearly into tech/SF as much as you guys), what is the purpose of having humanoid-like robots (other than swearing)? So many things are designed to be used by human beings, it just makes sense to think about also making a robot that can use our same tools, to some extent. Mostly tho, I think it’s because when we think of doing something, we think of how a person would do it. It’s where all our firsthand experience lies, after all, and the vast majority of tasks throughout history have been done by humans so we have an entrenched historical culture of using humanoid-shaped beings to do stuff as well. So mostly bias reasons, at the moment. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 11, 2017, 2:05am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/35 "2017-12-11T02:05:51Z")
> [@burpo\_the\_wonder\_mutt](#): > > Without sounding like a total näif, or a smart-ass (I’m not nearly into tech/SF as much as you guys), what is the purpose of having humanoid-like robots (other than swearing)? I’m not quite sure where this is coming from. Why are you thinking there’s some major goal for humanoid robots? There’s plenty of interest in non-humanoid robots too. As **Bo** says, humanoid robots have the inherent advantage that the built environment is already humanoid-friendly because it’s human-friendly. Recall the metal monster assassin-bot in Robocop that fell down the staircase because it wasn’t humanoid, whereas Robocop himself, being humanoid did just fine on the same stairs. Consider all the BS it takes to make a “handicapped friendly” hotel room or house or office. And that’s for a real human whose only difference from you or me is being 4ish feet tall and 3ish feet wide. But otherwise has the same hands, arms, eyes, mouth, etc. Making robots fit in where we live is probably easier than building a new world for them to live in. For factory production purposes robots will be every shape imaginable. Humans are certainly not the optimal shape or size for building either small stuff like ICs or big stuff like tanker ships. AIs have no shape at all. They think. They shuffle virtual paper and “talk on the phone” to each other. And actually talk on the actual phone to humans. ---
**Author:** ![burpo\_the\_wonder\_mutt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/burpo_the_wonder_mutt/32/365_2.png) [@burpo\_the\_wonder\_mutt](https://boards.straightdope.com/u/burpo_the_wonder_mutt)\ **Post date:** [December 11, 2017, 2:56am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/36 "2017-12-11T02:56:50Z")
> [@LSLGuy](#): > > I’m not quite sure where this is coming from. Why are you thinking there’s some major goal for humanoid robots? There’s plenty of interest in non-humanoid robots too. > > As **Bo** says, humanoid robots have the inherent advantage that the built environment is already humanoid-friendly because it’s human-friendly. Recall the metal monster assassin-bot in Robocop that fell down the staircase because it wasn’t humanoid, whereas Robocop himself, being humanoid did just fine on the same stairs. Consider all the BS it takes to make a “handicapped friendly” hotel room or house or office. And that’s for a real human whose only difference from you or me is being 4ish feet tall and 3ish feet wide. But otherwise has the same hands, arms, eyes, mouth, etc. > > Making robots fit in where we live is probably easier than building a new world for them to live in. > > For factory production purposes robots will be every shape imaginable. Humans are certainly not the optimal shape or size for building either small stuff like ICs or big stuff like tanker ships. > > AIs have no shape at all. They think. They shuffle virtual paper and “talk on the phone” to each other. And actually talk on the actual phone to humans. I don’t understand where your first paragraph is coming from: I asked a straight-forward question and didn’t wish to come off sounding like a wise-guy or a complete novice. I got a decent answer from **PastTense** (unless I’m being whooshed) and especially from **Bo** –thank you, guys. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 11, 2017, 3:08am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/37 "2017-12-11T03:08:42Z")
Sorry to create confusion. I hope I didn’t sound hostile; that wasn’t my intent. It seemed to me that your question implied some upstream assumptions that weren’t stated and weren’t obvious. Or at least not to me. I was trying to get at what those were. With the intention of adding whatever I or we could to filling in the gaps you said you had. ---
**Author:** ![burpo\_the\_wonder\_mutt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/burpo_the_wonder_mutt/32/365_2.png) [@burpo\_the\_wonder\_mutt](https://boards.straightdope.com/u/burpo_the_wonder_mutt)\ **Post date:** [December 11, 2017, 3:24am UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/38 "2017-12-11T03:24:27Z")
Actually, I was trying to list, in my head, situations where a humanoid robot would be “put to work,” instead of some other kind of automation (I was thinking of Detroit, in particular–they wouldn’t replace the present “robots” they’re using with the one Kevin was screwing with; or _would_ they?). Now, if you’re saying Atlas would eventually be cheap enough to have around the warehouse toting boxes, I can get behind that. ---
**Author:** ![Clothahump](https://avatars.discourse-cdn.com/v4/letter/c/51bf81/32.png) [@Clothahump](https://boards.straightdope.com/u/Clothahump)\ **Post date:** [December 11, 2017, 6:30pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/39 "2017-12-11T18:30:52Z")
> [@Doctor\_Jackson](#): > > “What’s new, Atlas?”, he asked. Atlas shrugged. I saw what you did there. ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [December 15, 2017, 4:14pm UTC](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952/40 "2017-12-15T16:14:33Z")
I just read [an interview with the Top Gear/Grand Tour guys](https://apnews.com/99295bf8c0b44414a3376fa243ac83fb/3-car-gurus-from-'The-Grand-Tour'-ponder-the-car's-future), and I love this part: > [@Jeremy Clarkson](#): > > ARE DRIVERLESS CARS AROUND THE CORNER? > > I don’t see those happening in my lifetime, probably not even in the lifetime of my children. For 40 years we’ve been promised robots. Nobody has built a robot than can make a sandwich or boil an egg or even climb a flight of stairs or open a door. And yet somehow we’re expected to believe we’re on the cusp of a car that can negotiate Manhattan or drive around a U.K. ring road or handle Rome. This is a classic case of someone being an expert in one field and having absolutely zero knowledge of other fields. We _have_ robots that can make sandwiches, boil eggs, climb stairs and open doors. FFS, we have some that can do standing backflips. [Previous page](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952.md?page=1) [Next page](https://boards.straightdope.com/t/whats-new-atlas-ai-robotics-and-tech-thread/801952.md?page=3)