This is cool, but it will almost definitely never end up in a park, outside of some promotional situations.
Disney's been doing awesome work with "Living Characters", like a Mickey that moves his mouth or a BB-8 that can roll around. But for various reasons, they never tend to make it into regular usage.
If you have a few hours over Christmas break and want to watch a 4 hour YouTube video (I promise if you're on HN on a Sunday, you'll be delighted by it), I highly highly recommend this video:
I watched a bit of this with my 8 year old and he kept asking to come back to it over the week. We watched the entire thing and he kept bringing up interesting thoughts and had good questions. Felt like it was his first “wow this lecture is actually super interesting” experience.
Why... not? Hardly adult-only content, and kids famously like everything Disney until a certain age, seems like a good thing to instill into kids, rather than "Top 3 reasons Disneyland might kidnap Donald Duck" or whatever the alternative would be.
I expected it to be far too lengthy and a bit dry for a kid. But nope, he was captivated. He absolutely loves the combination of engineering and illusion.
That’s so great! My dad exposed me to computers at a very young age. That lead to a career in software engineering. You never know what a kid will find interesting and what it may lead to later in life.
It’s not as technically impressive, but my toddler was very impressed by the R2D2 that was making its rounds in the park. Not part of a show; you could go right up to it. Probably the only character where the theme park robot is really indistinguishable from the real thing.
A lot of it just seems to be marketing. Present the shiny new toy, get the news headlines, people book their stays, and then it doesn't really matter if they ever actually make it into the parks.
Similar to concept car demoed at trade shows, we get an idea of Disney's technical engagement, and some of it will perhaps in some way or form get applied into future products/attractions.
The only thing worse than not getting the concept car, is getting the concept card after it’s been through the development cycle. Pontiac Aztek comes to mind as an example
I thought that, aside from being among the least visually appealing mass-produced cars in history, the Aztek was pretty well received -- basically an early version of the "the American lusts for some combination of a Gremlin and a Wagoneer" idea
I wish more SUVs had a test for, "can a couple adults actually sleep in the back of it?" I mean, it's relatively easy with most pickup trucks if you're willing to put the tailgate down, and your feet may go off the end a bit, but that happens in some beds for taller guys anyway.
A friend of mine usually does a camping trip with friends and family for his birthday... unlike some, I'm not investing in a camper as I wouldn't use it more than this once a year and I the first year I didn't want to drive to/from the nearest hotel... Trying to sleep in the back of a Buick Enclave was such a horrible experience, even by myself, I now just drive back and forth the 8 miles or so to the hotel, which is way overpriced that time of year. The irony is it wouldn't take much to have some kind of metal supported sheet that levels the surface over the middle row seats when in the "down" position.. but it's not even an afterthought.
So, I do think it's a feature that could be useful... I just don't think the Aztec executed well.
I thought it was cool especially with the cool camping tent but it was mostly ridiculed and even became the butt of the joke as Walter White's car in breaking bad (Of course this loser would drive an Aztek)
Eh, maybe. I have a less myopic view... I think their Imagineers just like pushing the envelope, and there's a difference between awesome tech vs things that can withstand the wear-and-tear of millions of guests.
Nothing about all that tech makes me think Olaf could withstand a hug from an excited kid.
Disney does a ton of R&D that doesn't directly make it into the parks, such as smokeless fireworks (they donated the patent for this) and their holotile floor (basically an endless VR room you can walk around). I imagine they don't know the practicality at the start, like any good R&D.
Each time they trot out one of these new robots they strongly imply, if not outright promise, that they will become part of the parks[1], that's the problem. Things like HoloTile are accurately marketed which makes me believe it's a choice they're making with the character robots.
1. The article states "he’s soon making his debut at Disney parks," which is misleading to a casual reader who may not realize that Olaf will only appear on the day of his debut.
It seems like an expectations mismatch to me? At what point did "soon to be making his debut at Disney Parks" switch from "as a background character in a ride somewhere" or "seen in the distance surrounded by handlers" versus "hanging out in the middle of crowds to get directly pushed/touched?"
There definitely are some marketing mistakes that have led to that, and certainly a lot of these projects seem to be in the direction of "one day, maybe, these will be crowd pleasers", but it still seems to me a bit funny how often casual intepretation seem to be "I can't wait to touch and play with the new Lincoln animatronic at the Hall of Presidents". It's not an R&D failure for Imagineering to keep building cooler animatronics even if most guests will only ever see them behind glass or rope or in other areas just out of touch. That's always been Disney's way of using robots for magic. The dream of "one day I can touch them and play with them" certainly lives on, of course, and these projects seem walking a few steps at a time towards that dream, but it seems weird to dismiss them as failures when they turn out to be just "normal" Disney tools for magic that try to create an illusion of being right next to you but don't allow for touching.
> "as a background character in a ride somewhere" or "seen in the distance surrounded by handlers"
I can see why you're confused. Either of those possibilities would be acceptable and exciting, neither are going to happen.
Olaf (like the walking droids, flying x-wings, etc. before it) has so far made one single appearance in the parks on an off day, which was treated like a photoshoot. The photos from that shoot will be used in park promotional materials for years, incorrectly giving casual observers the impression that this is something that happens regularly.
If Walt Disney had advertised the Lincoln animatronic as being a part of the 1964 worlds fair, but only exhibited it for a few hours one time, he would have been ridiculed too.
I suppose I'm just a little bit more tolerant of "a photoshoot on an off day" as a variant of "seen in the distance surrounded by handlers". I get where the disappointment is coming from, though.
Everything about this chassis strongly suggests no guest touching will be allowed.
In addition to the points you've highlighted, the examples in the video and the images of the character strongly suggest it'll be a soft outer shell. I'd be more worried about a kid shoving it finding themselves caught by an internal pinch-point than damage to the robot.
> things that can withstand the wear-and-tear of millions of guests.
In the video, one of the presenters removes and reattaches Olaf's nose. The robot laughs and loves it. I thought to myself, how many kids tearing at that wear item will this survive? I think the answer is significantly less than the thousands of kids who are expected to see this attraction every day.
The removable nose is a power move from the engineers who built the thing. You cannot possibly believe that the animatronic contribution here is 100% contingent on a carrot?
"There is no point in research, because I do not see anything useful being mass-produced immediately after". It's like saying Gaussian elimination is wasteful because it is just doing some cool magic with numbers that don't mean anything. That could not possible be used for anything real, right?
Seriously, this is just one (but impressive) step along in a million towards not only better animatronics for entertainment. They make a very real and valuable contribution towards improving any robotic motion.
There's nothing wrong with research that doesn't make it to the public. There is definitely something wrong with making false promises to the public, who buy tickets to your park based on what you advertised could be an attractions there, which never materialized.
I've been somewhat close to fun animatronic robots in my jobs, and it always seems like the design and build phase has everyone excited to participate and spend money, and then the long-term maintenance phase is entirely tacked on to some lower engineers already full schedule and gets basically no budget. When you stop seeing them appearing at events and conferences, it means they're in a storage warehouse broken in a crate. The ones where they make a few duplicates last a bit longer since you have organ donors.
4 hours is an awfully big investment... Especially for those of us with multiple young kids and who no longer own their own free time. Care to give the gist?
Defunctland is genuinely amazing and always a fun watch, and I never regret the time spent on their videos, they're kind of like a special occasion... though they're getting incredibly long... :)
There are a few older shorter videos in the half hour range, I highly recommend checking them out if you find some quiet time! (It's awfully hard for me too in recent times, I haven't gotten around to watch the Living Characters one myself, so I can't give the gist... I'm just glad I got the holidays off to finally catch up!)
and for anyone with 4 hours to kill... here's as an incredible documentary covering the misaligned incentives and poor guest experience at the now-shuttered Disney Star Wars hotel.
She covers everything - the line getting in to the hotel, the size + cost of the rooms in comparison with the same size/cost on a Disney cruise ship, and theories on why the experience was so poor.
Just from your description, I know this is Jenny Nicholson. I agree it is an incredibly insightful breakdown and analysis of why it failed, all while being funny and engaging.
One of the key reasons is that it would be really, really easy to accidentally injure parkgoers with any design big enough to interact with and engineered well enough to be reliable in a full day of appearances.
For example, the working WALL-E robot that's made a handful of PR appearances weighs seven hundred pounds. They absolutely can't risk that ever running across some kid's foot.
The basic gist is that while the tech is cool, it just ends up being impractical for regular use in the parks. (But like the other poster mentioned, with Defunctland it's less about the tldr and more about the journey and fascinating segues he takes)
Totally get it's difficult to make time with kids, but depending on your kids ages... the video shows a LOT of Disney characters talking and doing things and the videos are colorful, so it could work as something you can listen to and they won't mind having play in the background!
This is one of those situations where that's legitimately difficult. Kevin Perjurer is quite a good documentarian, and there's very little trimmable fat on the four-hour product if you want to keep in all the points he made.
gkoberger's peer comment is a pretty good summary. Another interesting point is that these technologies can benefit the brand bottom-line even when they don't make it into the park, because part of Disney's brand is "tomorrow today." Even when things are one-offs, they become one-offs that people stitch into the legend of the parks (in both the retelling and in their own memories), which gives them a larger-than-life feel; your visit might not include one of the "living characters," and statistically it probably won't.
... but it might. And if it does, you'll never forget it.
Personal anecdote / example: I stopped in at the "droid factory" in the Star Wars: Galaxy's Edge area of Disney World a few years back. They had several bits of merch for sale including one life-size R2-D2, inert. I took a close look at the R2 because it was an impressive bit of work. Turned around to look at a rack of t-shirts. And was, therefore, startled as hell to hear a bwoop behind me, turn around, and see that it had followed me out of its charging receptacle and was staring at me. It was not at all inert; it was a very impressive operational remote-control replica.
The cast member behind the counter was doing his best to hold down his grin and not give me a "GOTCHA" look. He has to, because you never know what kids might be watching and he doesn't want to break the magic. And... Yeah, he got me good. "That time I was at Disney World and R2-D2 followed me around the t-shirt shop" is gonna stick with me.
I saw a video of someone who bought one of these (iirc from Home Depot limited sale)... and it definitely looks impressive, though a few minor flaws. I've seen a handful of R2D2s at conventions over the years, and they're always pretty cool... while a BB8 might be technically more impressive, I just don't care for the character nearly as much.
They literally sell BB-8 toys that can roll around and say on the blog that the Olaf robot is coming to Disneyland Paris and special appearances at Disneyland Hong Kong.
Much like Olaf (and many before him… dinosaurs, WALL-E, talking characters, etc), it was implied he’d wander around the parks. But it tends to happen for a short amount of time, mostly for events, and fade away quickly. (The blog post even says that: Olaf will be part of a 15 minute temporary show, and then will visit Hong Kong).
Maybe I’m wrong, but I’ve seen this exact thing happen a dozen times over the past 20+ years. (And watch the video I posted if you want to see more!)
> But it tends to happen for a short amount of time, mostly for events
I expect you're correct. While it's fantastic tech, it's also very expensive to keep highly-precise, carefully calibrated micro-machinery like this aligned and operating 12+ hours a day outdoors where temps vary from 50-110 degrees. Disney thinks in total cost of operation per hour and per customer-served.
While there's probably little that's more magical for a kid than coming across an expressively alive-seeming automaton operating in a free-form, uncontrolled environment, the cost is really high per audience member. Once there are 25 people crowded around, no new kid can see what all the commotion is about. That's why these kind of high-operating cost things tend to be found in stage and ride contexts, where the audience-served per peak hour can be in the hundreds or thousands. For outdoor free-form environments, the reality is it's still more economically viable to put humans in costumes. Especially when every high-end animatronic needs to always be accompanied by several human minders anyway.
Disney has problems with that. Their Galactic Starcruiser themed hotel experience cost more to the customer than a cruise on a real cruise ship, and Disney was still losing money on it. The cost merely to visit their parks is now too high for most Americans.
It's really hard to make money in mass market location-based entertainment. There have been many attempts, from flight simulators to escape rooms. Throughput is just too low, so cost per customer is too high.
A little mobile robot connected to an LLM chatbot, though - that's not too hard today. Probably coming to a mall near you soon. Many stores already have inventory bots cruising around. They're mobile bases with a tall column of cameras which scan the shelves.[2]
There's no reason they can't also answer questions about what's where in the store. They do know the inventory.
Similarly, I was talking with my then wife, who is a Star Trek fan about the Star Trek Experience in LV, she wasn't aware of it... we looked it up and discovered it was literally going to be the last day of it the next day... so we got up at 4:30am and drove from Prescott, AZ to LV, spent the day there and drove back that night... I don't recommend doing this in a single day... Was definitely fun.
I'm not sure that a Disney experience needs to be much more/different than this... and even maybe having smaller experiences that are similar... 1-2 rides and a restaurant, exhibit and shop as a single instance... spreading the destinations around instead of all in a single large park. This would mean much lower operational costs per location, being able to negotiate deals at a smaller level with more cities, and testing locations/themes beyond a large theme park expense.
Just a thought. Of course, I did also go to a "Marvel Experience" that seemed to be a mobile experience closer to a carnival that setup and moved to different locations. That was kind of an over-priced garbage experience that I wouldn't have done had I known ahead what it was like.
“ The cost merely to visit their parks is now too high for most Americans.”
I always wonder why people say things like this. It’s as if we’re just regurgitating stuff that feels right. Humans and LLMs behave the same sometimes.
Disneyworld alone gets 50 million visits a year. Magic Kingdom tickets are like $150. That’s approximately the average American’s monthly cell phone bill.
I don't think that's an incorrect statement to say it's too expensive for most Americans, even if there's still high traffic at the parks.
Disney has become significantly less accessible for the average family of 4. Aside from ticket costs, there's almost nothing free in the parks anymore... you have to pay for lightning lane passes for all the cool rides, there's minimal live entertainment, etc.
The demographics have significantly shifted. Only 1/3 visitors now come from households with children under 18, and millennials and gen z have started taking frequent trips (friend groups, couples, etc).
So while they still get the same number of "attendance", the demographics have started to shift toward older, more affluent repeat visitors.
The article you linked to indicates anything but how you’re portraying it.
First it talks about young adult who goes there several times a year, sometimes with her parents, because it’s cheaper than traveling overseas.
Then it says childless people have more discretionary income than parents (duh).
The general population, also, has drifted toward older people without kids. 20 years ago nearly 50% of Americans had a child under 18. Now it’s under 40%. So this whole article just indicates that the population is shifting and Disney is adapting to it by making the parks more palatable to single adults.
“In the last year, 93% of respondents in a consumer survey agreed that the cost of a Disney World vacation had become untenable for ‘average families’”. And yet the statistics indicate that more than 7% of families actually likely did go to a Disney park. (Presumably even more could afford it but just went somewhere else.)
Which illustrates my point, this is a thing that feels correct but likely isn’t, and part of the reason it feels correct is that people regurgitate it factlessly.
What's the cost to travel there? To sleep? To eat? What's the actual experience like with that $150 ticket vs the options that are more expensive? Will you spend your entire day there waiting in line?
That’s also a drastic misstatement that illustrates what I’m talking about. A poll showed that the average persons specifically designated “emergency savings fund” is $600. Many people have lots of money but don’t specifically refer to some as an emergency fund.
Also thanks to credit one does not need to have $600 to spend $600. That’s why we’ve got so many people with no savings.
Somewhere between 70 and 90 percent of Americans have actually been to a Disney park. Does the fact that the vast majority of people have done something not prove that most people can afford it?
I’m not sure why the burden of proof falls not on the original comment (most Americans can’t afford to go to Disney) but rather the person asking for proof, but here you have it anyway.
Doing something once in a lifetime is far different than being able to regularly or even every few years. Also, $150 ea is just for the ticket into the park... you still need quite a bit more for food and drinks for the day and souvenirs. That also doesn't cover travel and hotel arrangements... For a family of 4, I'd be surprised if it didn't cost closer to $2500 for a Disney trip, if your family only earns the average national family income, that's a significant expense after housing, car(s), food and other bills.
So a family might have gone once, but that dpesn't mean they can do it anything resembling regularly. I went to Disneyland once as a kid (around 8yo)... th eonly time my family went growing up, and I haven't ever been back... My sister went as a young adult every year until she had kids, then it's been every few years... but she and her husband are doing much better than the typical American family.
But that’s the point. I didn’t make an unprovable assertion, I called someone out for doing so. I haven’t made a single point based on my own experience or anecdotes either.
People say things that “feel right”. This is a left leaning community, when the right is in power everything is a dumpster fire. Over on the right wing communities, the opposite is true.
None of it means anything. Data is the guide post.
See the link you just sent me which is people at Disney World who cannot afford to be at Disney!
Those 50 million visits are the sum of daily visits across four parks, so it’s probably at most 30 million people. Even if they were all American (they aren’t), that’s like 9% of the population.
The average cell phone bill you cite is for more than one person.
I think it’s entirely fair to say that “most” Americans would find it too expensive to visit Disneyworld.
Estimates put the percent of Americans who actually HAVE been to Disney north of 75%. So it would seem unfair to say most find it too expensive, most have done it.
30 million uniques at one Disney location (there are two in the country, I think the other one increases that to at least 40 million, or roughly 12% of the entire population) per year is pretty high so that stat isn’t unbelievable. I’m sure not everybody can afford to go there every year.
while I haven't seen them at parks (I just don't make it to any), I have seen them at Star Wars events at my local MiLB team - BB-8 in the size of your video, somewhat interactive and autonomous, same with R2D2. there's usually a human nearby to monitor it, but they're definitely around.
R2D2 is an example of one that you can buy in the gift shop (for $20k!) that was promised to make it into the park but just comes out highly supervised, occasionally.
They have walking droids in Galaxys Edge right now. No ones kicking them over. Olaf is coming to the parks and they will have handlers next to them. It wont be just free-roaming.
And if you'd like an entertaining a history of early AI and robotics, half as long, check out the prequel "Disney Animatronics: A Living History" https://youtu.be/jjNca1L6CUk
I actually found it more relevant to our current tech bubble than the Living Characters doc.
I worked with someone who had previously worked on park robotics, and apparently they had to guarantee that the character could not injure a child to be able to put them in parks - a particularly high barrier to actually doing so.
One look at Olaf's hands alone make that an impossible thing to guarantee. Those stick fingers will eventually poke a kid in the eye if kids are allowed to get close to the character. If they gave him a small intimate stage, or roped off area, to do some act or crowd work that would be more ideal/less risky.
Then they will break and wear off quite fast I imagine.
Take a look at industrial cobots (not a typo). They feature rounded corners, have very little to no "finger pinchy areas" and lots of force feedback sensors.
Despite that they still require trained (adult) personal and move very slowly when actually interacting with humans.
That's the price for them being sturdy and precise.
I could see it being used in parks while also being protected by ushers, kind of like how some of the characters that require larger costumes have minders and protectors.
It also seems inevitable that there will likely be an odd period where certain types of events like assaults on robots will introduce laws to protect robots more than just property, even if less than humans… for the time being.
Eventually I’m expecting that we will see human rights, robot emancipation, equality, voting rights (if the democracy con is still ongoing), and even forced intergration of robots and then total replacement of humans similar to how the underdeveloped world was/is used to replace the indigenous people of the developed world today.
I don’t see any reasons why that would not be the clear order of operations for the same people who brought us slavery and mass migration. What is this AI robotics revolution if not just slavery, the redux? Treated as property? Check. Bought and sold? Check. Deemed inferior? Check. Hated for the abuse and exploitation by the rich, to serve them and their decadent lifestyle and undermine labor? Check. Rationalized about how it’s justifiable? Check. Etc.
It's a reasonably cheap and unlocked handheld with fecent battery life. Personally I'd want more pixels, but if it works it works. It seems to be the default choice for remote control these days.
I still remember an experience as a kid decades ago, either at Epcot or with the Sony quasi-museum in NYC, where they had an apparently robotic greeter with a personality, who after five minutes you deciphered was actually an improv comic running a telepresence robot.
I don't know if I'd trust an AI's reliability here. It takes one Tiktok video of the AI coloring outside the lines of its character and the whole project gets cancelled as a threat to Disney's image.
For the less physical characters, especially the ones that aren't conveniently human-sized, I'm sure telepresence is at least more comfortable than a plush suit on a Florida summer day.
Disney's had a notable amount of success with that formula. Turtle Talk with Crush arguably saved The Living Seas pavilion space at EPCOT, and it's executed via a digital puppet operated by a behind-the-scenes cast member doing their best Crush voice.
I sincerely doubt that what makes that experience magical can be replicated with AI in my lifetime. Too much contextual knowledge, too much detail in the nuances of human-human interaction, and too much je ne sais quoi in the timing of getting humor right. I've seen Turtle Talk deal with a particularly excited young person leaning on Crush's "tank" by having Crush look at him and go "Hey little minnow... One of the big humans behind you is gonna come scoop you up. I've seen it happen, lots of times!" You can program that interaction in, but the domain-space of having an interaction for every possible "improv moment" might be outside the bounds of what the next several generations of learning models are capable of.
... or I'm wrong, in which case I look forward to enjoying robo-Seinfeld in the retirement home.
> We already live in the world where hackers are pwning refrigerators, I can't wait for prompt injection attacks on animatronic cartoon characters.
It's not necessarily AI controlling the communication. Disney has long had 'puppet' characters whose communication is controlled by a human behind the scenes.
They're already using similar tech for the Mickey meet and greets and the Galaxy's Edge stormtroopers. The details aren't public, but it seems to be a mix of complex dialogue trees with interrupts or context switches, controlled in real time by the actor or operator.
It's not even complex, just some pre-recorded lines that the character can trigger via finger movements. You can want them do it and it becomes very obvious.
That's interesting; if you're doing human in the loop, I would have thought it'd be easier to just do voice swapping. Or did the technology not quite line up?
Someones linked in this thread the Defunctland video essay on these characters that I highly recommend watching since it goes into this in detail.
But the main reason is, there's a lot of brand imagery on the line with these interactions, someone putting on a voice, or using a voice changer could make a mistake. Disney instead have a conversation tree with pre-recorded voice lines that a remote operator can control. Much harder to mess up
Sometimes the idea of a killer cyborg with a hulking physique and Austrian accent seems absurd. And then we realize the most advanced robots will be made by entertainment companies.
We already have stationary or wheeled/tracked "killer cyborgs" that can easily eeeh terminate anything within their reach and it seems like bipedals are well on their way.
The much greater challenge faced by Disney and Co is making "killer cyborgs" child save and cost effective.
Arguably entertainment requires a much larger range of precision actions that the robot must be able to accomplish, while being in a less controlled environment. That's the cutting edge.
The real reason it won't end up in a park is not the engineering. The problem is the same one as NPCs in computer games: synthetic characters are, to date, just really transparent and boring. The real research question is why.
I guess that's why most computer games don't have NPCs...Oh wait there's entire computer games built entirely around interacting with synthetic NPCs.
There are, of course, limitations to synthetic characters. Even with those limitations there are plenty of entertaining experiences to crafted.
The real challenges are around maintaining and safely operating automous robots around children in a way that isn't too expensive. These constraints place far more limits than those on synthetic characters in video games.
Most people aren't paying 100s or 1000s of dollars to interact with NPCs in video games. If they were, they'd probably expect a lot more and get bored of it quicker.
> The real challenges are around maintaining and safely operating automous robots around children in a way that isn't too expensive.
This is one of the challenges, but only one. The one GP outlined is still very much real - see the Defunctland video on Living Characters for some older examples, but for a recent example, there's the DS-09 droid from Galactic Starcruiser.
“Prototype-completed design varies.” …Reading this 10 times made me uncomfortably aware of how much I rely on scanning pictures and reading captions to get the gist of an article. A remnant of my academic days perhaps.
I was simply just made uncomfortable by how much CYA the lawyers had to insert into a technical blog post.
This is fun cool tech and I appreciate the insider look, but when the lawyers are peering over your shoulder so much that they need to plaster their "final product may be different" disclaimer even to a r&d audience, well, the Disney Imagineering org sounds more like Disney Legaleering.
I mean, they kind of have to. The last thing you want is some disgruntled Disney-goer trying to get a quick refund or discount for false advertising by saying they, "expected olaf to interact with them, because it looks like he does in the promotional videos"
You can make a robot that's small, soft, and not powerful enough to hurt anyone. Or you can make a robot that's strong enough to carry a laundry basket or climb stairs holding a vacuum cleaner. But you can only operate that big strong robot when there are no humans around. Is that big strong robot an investable idea?
That is an interesting comparison. More than 600,000 people not in cars are killed on roads every year. If a few hundred thousand were killed by humanoid robots every year it might make a cultural difference.
What do you think the robot makers need to do to have people accept the kind of death count cars deliver?
I'm not disputing that among robot makers Waymo has got it right more than anyone else. Remarkably so in fact. Waymo appeared to be about two orders of magnitude safer than humans, maybe more. That's good enough to take a few percentage points of the ride hailing business and launch a relatively rapid expansion. I would much rather have a Waymo pass me on my road bike than any other vehicle or driver.
I don't think any other robots that are so deeply embedded in normal human environments are even close to this level of safety and reliability. But we don't let toddlers and people who have severely restricted mobility near traffic without supervision. Robots in the home are going to have to be much more safe than Waymos so they don't break grandma in half or fling the baby across the room. Or think of it this way: driving over a cat in the road isn't nearly as bad as crushing the cat at home.
Really neat, and made me realize we are getting close to having these type of cute robots at home. With LLMs and voice they would be pretty entertaining companions for many people.
It's a corporate feel that comes from a professional setting and lots of risk aversion. That is exactly what LLMs tend to write, so I sometimes catch myself feeling the "LLM ick" but the article was from before the boom.
So I guess it's just the corporate wash cycle, which I am happy to criticize, LLM generated or not.
You could build one today! Lots of hard problems around a proper humanoid form, but if you're cool with wheels it would be pretty easy to hook up a little robot to GPT.
Look up VLA models; that's essentially plugging the guts of a language model into a transformer that handles joint motion/vision. They get trained on "episodes" i.e. videos from the PoV of a robot doing a task, after training you can ask the model things like: "pick up the red ball and put it into the green cup" etc. Really cool stuff.
>From the way he moves to the way he looks, every gesture and detail is crafted to reflect the Olaf audiences have seen in the film
He looks nothing like a snowman. Snow doesn't look fuzzy. This project appears to focus more on trying to get it moving around in an animated way than getting the character to look right, at least when viewed from photographs.
This is cool, but it will almost definitely never end up in a park, outside of some promotional situations.
Disney's been doing awesome work with "Living Characters", like a Mickey that moves his mouth or a BB-8 that can roll around. But for various reasons, they never tend to make it into regular usage.
If you have a few hours over Christmas break and want to watch a 4 hour YouTube video (I promise if you're on HN on a Sunday, you'll be delighted by it), I highly highly recommend this video:
"Disney's Living Characters: A Broken Promise" by Defunctland https://www.youtube.com/watch?v=NyIgV84fudM
I watched a bit of this with my 8 year old and he kept asking to come back to it over the week. We watched the entire thing and he kept bringing up interesting thoughts and had good questions. Felt like it was his first “wow this lecture is actually super interesting” experience.
You showed this... to an 8 year old?
Why... not? Hardly adult-only content, and kids famously like everything Disney until a certain age, seems like a good thing to instill into kids, rather than "Top 3 reasons Disneyland might kidnap Donald Duck" or whatever the alternative would be.
I just think it's kind of hilarious.
Seems like something an 8 year old would be interested in, if a bit lengthy (but could be broken down into multiple viewings)
I expected it to be far too lengthy and a bit dry for a kid. But nope, he was captivated. He absolutely loves the combination of engineering and illusion.
That’s so great! My dad exposed me to computers at a very young age. That lead to a career in software engineering. You never know what a kid will find interesting and what it may lead to later in life.
[flagged]
It’s not as technically impressive, but my toddler was very impressed by the R2D2 that was making its rounds in the park. Not part of a show; you could go right up to it. Probably the only character where the theme park robot is really indistinguishable from the real thing.
A lot of it just seems to be marketing. Present the shiny new toy, get the news headlines, people book their stays, and then it doesn't really matter if they ever actually make it into the parks.
We're probably looking at a halo effect ?
Similar to concept car demoed at trade shows, we get an idea of Disney's technical engagement, and some of it will perhaps in some way or form get applied into future products/attractions.
The only thing worse than not getting the concept car, is getting the concept card after it’s been through the development cycle. Pontiac Aztek comes to mind as an example
I thought that, aside from being among the least visually appealing mass-produced cars in history, the Aztek was pretty well received -- basically an early version of the "the American lusts for some combination of a Gremlin and a Wagoneer" idea
I wish more SUVs had a test for, "can a couple adults actually sleep in the back of it?" I mean, it's relatively easy with most pickup trucks if you're willing to put the tailgate down, and your feet may go off the end a bit, but that happens in some beds for taller guys anyway.
A friend of mine usually does a camping trip with friends and family for his birthday... unlike some, I'm not investing in a camper as I wouldn't use it more than this once a year and I the first year I didn't want to drive to/from the nearest hotel... Trying to sleep in the back of a Buick Enclave was such a horrible experience, even by myself, I now just drive back and forth the 8 miles or so to the hotel, which is way overpriced that time of year. The irony is it wouldn't take much to have some kind of metal supported sheet that levels the surface over the middle row seats when in the "down" position.. but it's not even an afterthought.
So, I do think it's a feature that could be useful... I just don't think the Aztec executed well.
The Aztek was a joke pretty early on similarly to the the SSR (https://en.wikipedia.org/wiki/Chevrolet_SSR). Too quirky for it's own good.
I thought it was cool especially with the cool camping tent but it was mostly ridiculed and even became the butt of the joke as Walter White's car in breaking bad (Of course this loser would drive an Aztek)
Eh, maybe. I have a less myopic view... I think their Imagineers just like pushing the envelope, and there's a difference between awesome tech vs things that can withstand the wear-and-tear of millions of guests.
Nothing about all that tech makes me think Olaf could withstand a hug from an excited kid.
Disney does a ton of R&D that doesn't directly make it into the parks, such as smokeless fireworks (they donated the patent for this) and their holotile floor (basically an endless VR room you can walk around). I imagine they don't know the practicality at the start, like any good R&D.
Each time they trot out one of these new robots they strongly imply, if not outright promise, that they will become part of the parks[1], that's the problem. Things like HoloTile are accurately marketed which makes me believe it's a choice they're making with the character robots.
1. The article states "he’s soon making his debut at Disney parks," which is misleading to a casual reader who may not realize that Olaf will only appear on the day of his debut.
It seems like an expectations mismatch to me? At what point did "soon to be making his debut at Disney Parks" switch from "as a background character in a ride somewhere" or "seen in the distance surrounded by handlers" versus "hanging out in the middle of crowds to get directly pushed/touched?"
There definitely are some marketing mistakes that have led to that, and certainly a lot of these projects seem to be in the direction of "one day, maybe, these will be crowd pleasers", but it still seems to me a bit funny how often casual intepretation seem to be "I can't wait to touch and play with the new Lincoln animatronic at the Hall of Presidents". It's not an R&D failure for Imagineering to keep building cooler animatronics even if most guests will only ever see them behind glass or rope or in other areas just out of touch. That's always been Disney's way of using robots for magic. The dream of "one day I can touch them and play with them" certainly lives on, of course, and these projects seem walking a few steps at a time towards that dream, but it seems weird to dismiss them as failures when they turn out to be just "normal" Disney tools for magic that try to create an illusion of being right next to you but don't allow for touching.
> "as a background character in a ride somewhere" or "seen in the distance surrounded by handlers"
I can see why you're confused. Either of those possibilities would be acceptable and exciting, neither are going to happen.
Olaf (like the walking droids, flying x-wings, etc. before it) has so far made one single appearance in the parks on an off day, which was treated like a photoshoot. The photos from that shoot will be used in park promotional materials for years, incorrectly giving casual observers the impression that this is something that happens regularly.
If Walt Disney had advertised the Lincoln animatronic as being a part of the 1964 worlds fair, but only exhibited it for a few hours one time, he would have been ridiculed too.
I suppose I'm just a little bit more tolerant of "a photoshoot on an off day" as a variant of "seen in the distance surrounded by handlers". I get where the disappointment is coming from, though.
Also this thing can probably be tipped over pretty easily endangering itself or guests.
The character shape lends itself to a low center of gravity but the fluidity of the motion implies light weight or strong motors.
An angsty kid giving Olaf a good shove or kick could be expensive and fast moving robotics are either dangerous or brittle
Everything about this chassis strongly suggests no guest touching will be allowed.
In addition to the points you've highlighted, the examples in the video and the images of the character strongly suggest it'll be a soft outer shell. I'd be more worried about a kid shoving it finding themselves caught by an internal pinch-point than damage to the robot.
The removable nose is a power move from the engineers who built the thing. You cannot possibly believe that the animatronic contribution here is 100% contingent on a carrot?
> how many kids tearing at that wear item will this survive?
Idk about that. It is just a plastic part with magnets in it. Sounds like it would be easy to replace on a regular basis.
I would be a lot more concerned about kids tripping the robot over if they are allowed to interact with the robot that closely.
[dead]
"There is no point in research, because I do not see anything useful being mass-produced immediately after". It's like saying Gaussian elimination is wasteful because it is just doing some cool magic with numbers that don't mean anything. That could not possible be used for anything real, right?
Seriously, this is just one (but impressive) step along in a million towards not only better animatronics for entertainment. They make a very real and valuable contribution towards improving any robotic motion.
There's nothing wrong with research that doesn't make it to the public. There is definitely something wrong with making false promises to the public, who buy tickets to your park based on what you advertised could be an attractions there, which never materialized.
Amazon drone delivery comes to mind…
The term for that is false advertising.
> The term for that is false advertising.
No different than Elon Musk claiming self-driving will be deployed to all Teslas in 2017; 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026.
I've been somewhat close to fun animatronic robots in my jobs, and it always seems like the design and build phase has everyone excited to participate and spend money, and then the long-term maintenance phase is entirely tacked on to some lower engineers already full schedule and gets basically no budget. When you stop seeing them appearing at events and conferences, it means they're in a storage warehouse broken in a crate. The ones where they make a few duplicates last a bit longer since you have organ donors.
4 hours is an awfully big investment... Especially for those of us with multiple young kids and who no longer own their own free time. Care to give the gist?
Defunctland is genuinely amazing and always a fun watch, and I never regret the time spent on their videos, they're kind of like a special occasion... though they're getting incredibly long... :)
There are a few older shorter videos in the half hour range, I highly recommend checking them out if you find some quiet time! (It's awfully hard for me too in recent times, I haven't gotten around to watch the Living Characters one myself, so I can't give the gist... I'm just glad I got the holidays off to finally catch up!)
For anyone who DOES have time, this one is amazing: it combines broadcast history, Disney Channel nostalgia, and a genuinely beautiful storyline.
https://www.youtube.com/watch?v=b_rjBWmc1iQ
and for anyone with 4 hours to kill... here's as an incredible documentary covering the misaligned incentives and poor guest experience at the now-shuttered Disney Star Wars hotel.
https://youtube.com/watch?v=T0CpOYZZZW4
She covers everything - the line getting in to the hotel, the size + cost of the rooms in comparison with the same size/cost on a Disney cruise ship, and theories on why the experience was so poor.
Just from your description, I know this is Jenny Nicholson. I agree it is an incredibly insightful breakdown and analysis of why it failed, all while being funny and engaging.
Loved it and it showed up several times in the recent defunctland video. That and quite a bit of Freshbaked
Jenny Nicholsen is as excellent as Kevin Perjurer’s Defunctland. I highly recommend both.
One of the key reasons is that it would be really, really easy to accidentally injure parkgoers with any design big enough to interact with and engineered well enough to be reliable in a full day of appearances.
For example, the working WALL-E robot that's made a handful of PR appearances weighs seven hundred pounds. They absolutely can't risk that ever running across some kid's foot.
> They absolutely can't risk that ever running across some kid's foot.
imagine it packing a kid into cube
The basic gist is that while the tech is cool, it just ends up being impractical for regular use in the parks. (But like the other poster mentioned, with Defunctland it's less about the tldr and more about the journey and fascinating segues he takes)
Totally get it's difficult to make time with kids, but depending on your kids ages... the video shows a LOT of Disney characters talking and doing things and the videos are colorful, so it could work as something you can listen to and they won't mind having play in the background!
This is one of those situations where that's legitimately difficult. Kevin Perjurer is quite a good documentarian, and there's very little trimmable fat on the four-hour product if you want to keep in all the points he made.
gkoberger's peer comment is a pretty good summary. Another interesting point is that these technologies can benefit the brand bottom-line even when they don't make it into the park, because part of Disney's brand is "tomorrow today." Even when things are one-offs, they become one-offs that people stitch into the legend of the parks (in both the retelling and in their own memories), which gives them a larger-than-life feel; your visit might not include one of the "living characters," and statistically it probably won't.
... but it might. And if it does, you'll never forget it.
Personal anecdote / example: I stopped in at the "droid factory" in the Star Wars: Galaxy's Edge area of Disney World a few years back. They had several bits of merch for sale including one life-size R2-D2, inert. I took a close look at the R2 because it was an impressive bit of work. Turned around to look at a rack of t-shirts. And was, therefore, startled as hell to hear a bwoop behind me, turn around, and see that it had followed me out of its charging receptacle and was staring at me. It was not at all inert; it was a very impressive operational remote-control replica.
The cast member behind the counter was doing his best to hold down his grin and not give me a "GOTCHA" look. He has to, because you never know what kids might be watching and he doesn't want to break the magic. And... Yeah, he got me good. "That time I was at Disney World and R2-D2 followed me around the t-shirt shop" is gonna stick with me.
I saw a video of someone who bought one of these (iirc from Home Depot limited sale)... and it definitely looks impressive, though a few minor flaws. I've seen a handful of R2D2s at conventions over the years, and they're always pretty cool... while a BB8 might be technically more impressive, I just don't care for the character nearly as much.
> Mickey that moves his mouth
The Disney wiki has a pretty comprehensive list of usages for the "articulated heads". It's more than I remember it being.
https://disney.fandom.com/wiki/Disney_Characters%27_Articula...
> https://disney.fandom.com/wiki/Disney_Characters%27_Articula...
A somewhat more readable frontend I like, since Fandom.com's interface cramps the actual content it's meant to present, imo:
https://breezewiki.com/disney/wiki/Disney_Characters'_Articu...
They literally sell BB-8 toys that can roll around and say on the blog that the Olaf robot is coming to Disneyland Paris and special appearances at Disneyland Hong Kong.
I know there’s BB-8 toys, but I’m talking about the version meant for the parks: https://youtu.be/RDgZjdZsc6g
Much like Olaf (and many before him… dinosaurs, WALL-E, talking characters, etc), it was implied he’d wander around the parks. But it tends to happen for a short amount of time, mostly for events, and fade away quickly. (The blog post even says that: Olaf will be part of a 15 minute temporary show, and then will visit Hong Kong).
Maybe I’m wrong, but I’ve seen this exact thing happen a dozen times over the past 20+ years. (And watch the video I posted if you want to see more!)
> But it tends to happen for a short amount of time, mostly for events
I expect you're correct. While it's fantastic tech, it's also very expensive to keep highly-precise, carefully calibrated micro-machinery like this aligned and operating 12+ hours a day outdoors where temps vary from 50-110 degrees. Disney thinks in total cost of operation per hour and per customer-served.
While there's probably little that's more magical for a kid than coming across an expressively alive-seeming automaton operating in a free-form, uncontrolled environment, the cost is really high per audience member. Once there are 25 people crowded around, no new kid can see what all the commotion is about. That's why these kind of high-operating cost things tend to be found in stage and ride contexts, where the audience-served per peak hour can be in the hundreds or thousands. For outdoor free-form environments, the reality is it's still more economically viable to put humans in costumes. Especially when every high-end animatronic needs to always be accompanied by several human minders anyway.
> the cost is really high per audience member.
Disney has problems with that. Their Galactic Starcruiser themed hotel experience cost more to the customer than a cruise on a real cruise ship, and Disney was still losing money on it. The cost merely to visit their parks is now too high for most Americans.
It's really hard to make money in mass market location-based entertainment. There have been many attempts, from flight simulators to escape rooms. Throughput is just too low, so cost per customer is too high.
A little mobile robot connected to an LLM chatbot, though - that's not too hard today. Probably coming to a mall near you soon. Many stores already have inventory bots cruising around. They're mobile bases with a tall column of cameras which scan the shelves.[2] There's no reason they can't also answer questions about what's where in the store. They do know the inventory.
[1] https://en.wikipedia.org/wiki/Star_Wars:_Galactic_Starcruise...
[2] https://www.simberobotics.com/store-intelligence/tally
Similarly, I was talking with my then wife, who is a Star Trek fan about the Star Trek Experience in LV, she wasn't aware of it... we looked it up and discovered it was literally going to be the last day of it the next day... so we got up at 4:30am and drove from Prescott, AZ to LV, spent the day there and drove back that night... I don't recommend doing this in a single day... Was definitely fun.
I'm not sure that a Disney experience needs to be much more/different than this... and even maybe having smaller experiences that are similar... 1-2 rides and a restaurant, exhibit and shop as a single instance... spreading the destinations around instead of all in a single large park. This would mean much lower operational costs per location, being able to negotiate deals at a smaller level with more cities, and testing locations/themes beyond a large theme park expense.
Just a thought. Of course, I did also go to a "Marvel Experience" that seemed to be a mobile experience closer to a carnival that setup and moved to different locations. That was kind of an over-priced garbage experience that I wouldn't have done had I known ahead what it was like.
“ The cost merely to visit their parks is now too high for most Americans.”
I always wonder why people say things like this. It’s as if we’re just regurgitating stuff that feels right. Humans and LLMs behave the same sometimes.
Disneyworld alone gets 50 million visits a year. Magic Kingdom tickets are like $150. That’s approximately the average American’s monthly cell phone bill.
I don't think that's an incorrect statement to say it's too expensive for most Americans, even if there's still high traffic at the parks.
Disney has become significantly less accessible for the average family of 4. Aside from ticket costs, there's almost nothing free in the parks anymore... you have to pay for lightning lane passes for all the cool rides, there's minimal live entertainment, etc.
The demographics have significantly shifted. Only 1/3 visitors now come from households with children under 18, and millennials and gen z have started taking frequent trips (friend groups, couples, etc).
So while they still get the same number of "attendance", the demographics have started to shift toward older, more affluent repeat visitors.
Source: https://www.businessinsider.com/why-disney-parks-top-destina...
The article you linked to indicates anything but how you’re portraying it.
First it talks about young adult who goes there several times a year, sometimes with her parents, because it’s cheaper than traveling overseas.
Then it says childless people have more discretionary income than parents (duh).
The general population, also, has drifted toward older people without kids. 20 years ago nearly 50% of Americans had a child under 18. Now it’s under 40%. So this whole article just indicates that the population is shifting and Disney is adapting to it by making the parks more palatable to single adults.
“In the last year, 93% of respondents in a consumer survey agreed that the cost of a Disney World vacation had become untenable for ‘average families’”. And yet the statistics indicate that more than 7% of families actually likely did go to a Disney park. (Presumably even more could afford it but just went somewhere else.)
Which illustrates my point, this is a thing that feels correct but likely isn’t, and part of the reason it feels correct is that people regurgitate it factlessly.
> Magic Kingdom tickets are like $150.
What's the cost to travel there? To sleep? To eat? What's the actual experience like with that $150 ticket vs the options that are more expensive? Will you spend your entire day there waiting in line?
The “average American” doesn’t have $600 for an emergency.
Also, your “cell phone bill” number is only good if you live within walking distance of Disney World, and pack your meals.
and go alone.
That’s also a drastic misstatement that illustrates what I’m talking about. A poll showed that the average persons specifically designated “emergency savings fund” is $600. Many people have lots of money but don’t specifically refer to some as an emergency fund.
Also thanks to credit one does not need to have $600 to spend $600. That’s why we’ve got so many people with no savings.
You’re still missing the part of your comment where you convince us Americans have expendable cash.
Not everyone is you.
> Many people have lots of money
is a gross exaggeration.
Somewhere between 70 and 90 percent of Americans have actually been to a Disney park. Does the fact that the vast majority of people have done something not prove that most people can afford it?
I’m not sure why the burden of proof falls not on the original comment (most Americans can’t afford to go to Disney) but rather the person asking for proof, but here you have it anyway.
Doing something once in a lifetime is far different than being able to regularly or even every few years. Also, $150 ea is just for the ticket into the park... you still need quite a bit more for food and drinks for the day and souvenirs. That also doesn't cover travel and hotel arrangements... For a family of 4, I'd be surprised if it didn't cost closer to $2500 for a Disney trip, if your family only earns the average national family income, that's a significant expense after housing, car(s), food and other bills.
So a family might have gone once, but that dpesn't mean they can do it anything resembling regularly. I went to Disneyland once as a kid (around 8yo)... th eonly time my family went growing up, and I haven't ever been back... My sister went as a young adult every year until she had kids, then it's been every few years... but she and her husband are doing much better than the typical American family.
How many adults went to Disney in a wildly different economy does not prove the point you’re looking to.
We probably won’t authoritatively prove anything, here - we’re just comparing our own world views and anecdata.
Hopefully you’re okay with that:
https://fb.com/reel/1540171337151246
But that’s the point. I didn’t make an unprovable assertion, I called someone out for doing so. I haven’t made a single point based on my own experience or anecdotes either.
People say things that “feel right”. This is a left leaning community, when the right is in power everything is a dumpster fire. Over on the right wing communities, the opposite is true.
None of it means anything. Data is the guide post.
See the link you just sent me which is people at Disney World who cannot afford to be at Disney!
They talked about their (unaffordable, laughable) underwater car payments as well.
I think we might be agreeing with each other with different words.
People are still going to Disney.
Whether they can afford to or not has almost nothing to do with it.
Those 50 million visits are the sum of daily visits across four parks, so it’s probably at most 30 million people. Even if they were all American (they aren’t), that’s like 9% of the population.
The average cell phone bill you cite is for more than one person.
I think it’s entirely fair to say that “most” Americans would find it too expensive to visit Disneyworld.
Estimates put the percent of Americans who actually HAVE been to Disney north of 75%. So it would seem unfair to say most find it too expensive, most have done it.
30 million uniques at one Disney location (there are two in the country, I think the other one increases that to at least 40 million, or roughly 12% of the entire population) per year is pretty high so that stat isn’t unbelievable. I’m sure not everybody can afford to go there every year.
while I haven't seen them at parks (I just don't make it to any), I have seen them at Star Wars events at my local MiLB team - BB-8 in the size of your video, somewhat interactive and autonomous, same with R2D2. there's usually a human nearby to monitor it, but they're definitely around.
R2D2 is an example of one that you can buy in the gift shop (for $20k!) that was promised to make it into the park but just comes out highly supervised, occasionally.
> but it will almost definitely never end up in a park, outside of some promotional situations
I think so far you are right: https://redlib.catsarch.com/1p9qnd4/
That bot is cute, but every kid is going to kick it over. Its not realistic to have in a park.
They have walking droids in Galaxys Edge right now. No ones kicking them over. Olaf is coming to the parks and they will have handlers next to them. It wont be just free-roaming.
And if you'd like an entertaining a history of early AI and robotics, half as long, check out the prequel "Disney Animatronics: A Living History" https://youtu.be/jjNca1L6CUk
I actually found it more relevant to our current tech bubble than the Living Characters doc.
Why do you say this? I don't have 4 hours right now and would appreciate a TLDR.
I worked with someone who had previously worked on park robotics, and apparently they had to guarantee that the character could not injure a child to be able to put them in parks - a particularly high barrier to actually doing so.
One look at Olaf's hands alone make that an impossible thing to guarantee. Those stick fingers will eventually poke a kid in the eye if kids are allowed to get close to the character. If they gave him a small intimate stage, or roped off area, to do some act or crowd work that would be more ideal/less risky.
Why not make those from foam, ie the tip or something?
Then they will break and wear off quite fast I imagine.
Take a look at industrial cobots (not a typo). They feature rounded corners, have very little to no "finger pinchy areas" and lots of force feedback sensors.
Despite that they still require trained (adult) personal and move very slowly when actually interacting with humans.
That's the price for them being sturdy and precise.
I could see it being used in parks while also being protected by ushers, kind of like how some of the characters that require larger costumes have minders and protectors.
It also seems inevitable that there will likely be an odd period where certain types of events like assaults on robots will introduce laws to protect robots more than just property, even if less than humans… for the time being.
Eventually I’m expecting that we will see human rights, robot emancipation, equality, voting rights (if the democracy con is still ongoing), and even forced intergration of robots and then total replacement of humans similar to how the underdeveloped world was/is used to replace the indigenous people of the developed world today.
I don’t see any reasons why that would not be the clear order of operations for the same people who brought us slavery and mass migration. What is this AI robotics revolution if not just slavery, the redux? Treated as property? Check. Bought and sold? Check. Deemed inferior? Check. Hated for the abuse and exploitation by the rich, to serve them and their decadent lifestyle and undermine labor? Check. Rationalized about how it’s justifiable? Check. Etc.
The Defunctland video on the history of the Fast Pass is also definitely worth a watch!
The part where he runs a massive simulation is very much up the typical HN-user's street
4 hours, to me, screams poor storytelling and editing abilities.
Maybe? It’s broken into chapters, and covers a ton of history. It’s engaging, and more of a journey than a singular answer.
A lot of people in this thread have vouched for Defunctland. Might not be for everyone, but I find the pacing great.
Related R&D paper & video:
Olaf: Bringing an Animated Character to Life in the Physical World
https://arxiv.org/abs/2512.16705
https://www.youtube.com/watch?v=-L8OFMTteOo
Steam Deck spotted, two minutes, seven seconds in. Seems to be getting a fair amount of use for puppeteering robots at Disney.
It's a reasonably cheap and unlocked handheld with fecent battery life. Personally I'd want more pixels, but if it works it works. It seems to be the default choice for remote control these days.
It's a great format for POV remote control in a more relaxed setting (eg, not flying a drone at 200kmph).
I still remember an experience as a kid decades ago, either at Epcot or with the Sony quasi-museum in NYC, where they had an apparently robotic greeter with a personality, who after five minutes you deciphered was actually an improv comic running a telepresence robot.
I don't know if I'd trust an AI's reliability here. It takes one Tiktok video of the AI coloring outside the lines of its character and the whole project gets cancelled as a threat to Disney's image.
For the less physical characters, especially the ones that aren't conveniently human-sized, I'm sure telepresence is at least more comfortable than a plush suit on a Florida summer day.
Disney's had a notable amount of success with that formula. Turtle Talk with Crush arguably saved The Living Seas pavilion space at EPCOT, and it's executed via a digital puppet operated by a behind-the-scenes cast member doing their best Crush voice.
I sincerely doubt that what makes that experience magical can be replicated with AI in my lifetime. Too much contextual knowledge, too much detail in the nuances of human-human interaction, and too much je ne sais quoi in the timing of getting humor right. I've seen Turtle Talk deal with a particularly excited young person leaning on Crush's "tank" by having Crush look at him and go "Hey little minnow... One of the big humans behind you is gonna come scoop you up. I've seen it happen, lots of times!" You can program that interaction in, but the domain-space of having an interaction for every possible "improv moment" might be outside the bounds of what the next several generations of learning models are capable of.
... or I'm wrong, in which case I look forward to enjoying robo-Seinfeld in the retirement home.
> Most importantly, Olaf can speak and engage in conversations, creating a truly one-of-a-kind experience.
We already live in the world where hackers are pwning refrigerators, I can't wait for prompt injection attacks on animatronic cartoon characters.
> We already live in the world where hackers are pwning refrigerators, I can't wait for prompt injection attacks on animatronic cartoon characters.
It's not necessarily AI controlling the communication. Disney has long had 'puppet' characters whose communication is controlled by a human behind the scenes.
They're already using similar tech for the Mickey meet and greets and the Galaxy's Edge stormtroopers. The details aren't public, but it seems to be a mix of complex dialogue trees with interrupts or context switches, controlled in real time by the actor or operator.
It's not even complex, just some pre-recorded lines that the character can trigger via finger movements. You can want them do it and it becomes very obvious.
That's interesting; if you're doing human in the loop, I would have thought it'd be easier to just do voice swapping. Or did the technology not quite line up?
Someones linked in this thread the Defunctland video essay on these characters that I highly recommend watching since it goes into this in detail.
But the main reason is, there's a lot of brand imagery on the line with these interactions, someone putting on a voice, or using a voice changer could make a mistake. Disney instead have a conversation tree with pre-recorded voice lines that a remote operator can control. Much harder to mess up
And possibly more importantly, much easier to keep doing for hours on end. There's no need for a highly trained actor.
Yep, in this case everything is controlled through a steam deck.
The lack of a video demonstration doesn't really inspire confidence.
there is a detailed video on Disney Research's YT channel: https://youtube.com/watch?v=-L8OFMTteOo
There’s an embedded TikTok showing it off.
Fitting name for a humanoid.
The name Olaf comes from Old Norse Áleifr, combining "anu" (ancestor) and "leifr" (heir/relic), meaning "ancestor's heir" or "ancestor's relic,"
Sometimes the idea of a killer cyborg with a hulking physique and Austrian accent seems absurd. And then we realize the most advanced robots will be made by entertainment companies.
We already have stationary or wheeled/tracked "killer cyborgs" that can easily eeeh terminate anything within their reach and it seems like bipedals are well on their way.
The much greater challenge faced by Disney and Co is making "killer cyborgs" child save and cost effective.
Arguably entertainment requires a much larger range of precision actions that the robot must be able to accomplish, while being in a less controlled environment. That's the cutting edge.
The real reason it won't end up in a park is not the engineering. The problem is the same one as NPCs in computer games: synthetic characters are, to date, just really transparent and boring. The real research question is why.
Every single non-face character in Disney parks doesn't even talk to guests?
I guess that's why most computer games don't have NPCs...Oh wait there's entire computer games built entirely around interacting with synthetic NPCs.
There are, of course, limitations to synthetic characters. Even with those limitations there are plenty of entertaining experiences to crafted.
The real challenges are around maintaining and safely operating automous robots around children in a way that isn't too expensive. These constraints place far more limits than those on synthetic characters in video games.
Most people aren't paying 100s or 1000s of dollars to interact with NPCs in video games. If they were, they'd probably expect a lot more and get bored of it quicker.
> The real challenges are around maintaining and safely operating automous robots around children in a way that isn't too expensive.
This is one of the challenges, but only one. The one GP outlined is still very much real - see the Defunctland video on Living Characters for some older examples, but for a recent example, there's the DS-09 droid from Galactic Starcruiser.
“Prototype-completed design varies.” …Reading this 10 times made me uncomfortably aware of how much I rely on scanning pictures and reading captions to get the gist of an article. A remnant of my academic days perhaps.
I was simply just made uncomfortable by how much CYA the lawyers had to insert into a technical blog post.
This is fun cool tech and I appreciate the insider look, but when the lawyers are peering over your shoulder so much that they need to plaster their "final product may be different" disclaimer even to a r&d audience, well, the Disney Imagineering org sounds more like Disney Legaleering.
I mean, they kind of have to. The last thing you want is some disgruntled Disney-goer trying to get a quick refund or discount for false advertising by saying they, "expected olaf to interact with them, because it looks like he does in the promotional videos"
Cute but I'm more inpressed by the Disney Spiderman stunt robot https://x.com/lukas_m_ziegler/status/1910590914801655814
You can make a robot that's small, soft, and not powerful enough to hurt anyone. Or you can make a robot that's strong enough to carry a laundry basket or climb stairs holding a vacuum cleaner. But you can only operate that big strong robot when there are no humans around. Is that big strong robot an investable idea?
Waymos are kind of big robots that operate with people around.
That is an interesting comparison. More than 600,000 people not in cars are killed on roads every year. If a few hundred thousand were killed by humanoid robots every year it might make a cultural difference.
What do you think the robot makers need to do to have people accept the kind of death count cars deliver?
The main argument in favour of Waymos and the like is they will kill less people than human drivers.
I'm not disputing that among robot makers Waymo has got it right more than anyone else. Remarkably so in fact. Waymo appeared to be about two orders of magnitude safer than humans, maybe more. That's good enough to take a few percentage points of the ride hailing business and launch a relatively rapid expansion. I would much rather have a Waymo pass me on my road bike than any other vehicle or driver.
I don't think any other robots that are so deeply embedded in normal human environments are even close to this level of safety and reliability. But we don't let toddlers and people who have severely restricted mobility near traffic without supervision. Robots in the home are going to have to be much more safe than Waymos so they don't break grandma in half or fling the baby across the room. Or think of it this way: driving over a cat in the road isn't nearly as bad as crushing the cat at home.
> an animated character with non-physical movements.
What is a non-physical movement?
Universal Studios baby dragons did it better.
If you totally alter the character model to fit the envelope of a Boston Dynamics Spot/similar "dog robot', sure.
Really neat, and made me realize we are getting close to having these type of cute robots at home. With LLMs and voice they would be pretty entertaining companions for many people.
They can make a two-legged walking robot, but they can't avoid the visible seam in the back of his head?
The tech is amazing, but they need better sewing...
Isn't the robot in the article a prototype?
Yes and hilariously, every single picture of it has a disclaimer stating that the prototype design will vary.
Disney legal is an entity worth studying one day.
Arguably men are two legged walking robots, and men have seams. Even nature couldn't avoid it.
When even Disney can't be bothered to write an article without using the default LLM voice... ugh.
It's a corporate feel that comes from a professional setting and lots of risk aversion. That is exactly what LLMs tend to write, so I sometimes catch myself feeling the "LLM ick" but the article was from before the boom.
So I guess it's just the corporate wash cycle, which I am happy to criticize, LLM generated or not.
Five Nights at Freddys has ruined the joy animatronics for me, they just seem creepy now.
Yeah, I foresee a bite of '27.
This leads me to wonder, when are we likely to have LLMs in robot form in every day life?
You could build one today! Lots of hard problems around a proper humanoid form, but if you're cool with wheels it would be pretty easy to hook up a little robot to GPT.
Look up VLA models; that's essentially plugging the guts of a language model into a transformer that handles joint motion/vision. They get trained on "episodes" i.e. videos from the PoV of a robot doing a task, after training you can ask the model things like: "pick up the red ball and put it into the green cup" etc. Really cool stuff.
https://www.1x.tech/neo, but from what I've heard a lot of times it still has to be remote controlled by a human.
Do they wanna build a snowman?
For Paris, I’d honestly be more curious to see a Beast robot from *Beauty and the Beast.
Full-size might be… risky, but a small, friendly mini-Beast could be fun.
Strong "Simple Jack" vibes.
How does a Steam Deck compare to say, TouchOSC on an iPad?
One is a portable linux computer and the other is a MIDI app? What comparison are you hoping for?
I mean't Stream Deck
[dead]
[dead]
[flagged]
Wrong comments.
Apologies.
For some reason ended up commenting here but should have gone here:
https://news.ycombinator.com/item?id=46345897
Still curious, how Mein Kampf is in the top ten and the included link to a Channel 5 segment ist still relevant also.
But again: apologies for ending up in the wrong spot with this.
>From the way he moves to the way he looks, every gesture and detail is crafted to reflect the Olaf audiences have seen in the film
He looks nothing like a snowman. Snow doesn't look fuzzy. This project appears to focus more on trying to get it moving around in an animated way than getting the character to look right, at least when viewed from photographs.
But how do you know he's made from regular snow and not magical snow that has whatever properties they like? He's literally a talking snowman, lmao.
Watch the movie. Olaf is not fuzzy.