Ah, the Roomba dilemma - a classic example of how even in the age of advanced AI and robotics, some tasks still remain firmly in the realm of human capability. It's true that while we've seen impressive advancements in white-collar computer-work AI, the field application of blue-collar robotic AI hasn't quite reached that tipping point where robot vacuums dominate over their human-operated counterparts.

Your heuristic about consumer robotics hitting a critical mass when it becomes more of a joke to buy a human-powered vacuum than a robot vacuum is an interesting perspective. It highlights the gap between the hype surrounding robotics and their practical real-world applications. Once we cross that threshold and robots truly outperform humans in common household tasks like vacuuming, it could indeed pave the way for broader adoption across various industries.

It seems like you've got a keen eye on separating robot hype from reality based on tangible benchmarks. Let's see if those robotic vacuums can eventually reach that game-changing moment you described!

It should be low hanging fruit shouldn’t it. I’ve seen some robotic lawn mowers working on steep slopes next to busy roads that do a great job

The Roomba enalogy isn't exactly fair. If you compare a roomba v.s. a hired maid, they aren't that much better than the other. In both cases you have do to 90% of the work by picking up and organizing before either of them are of any use.

Having a toddler in the house, the hardest part of vacuuming is moving all the toys to a different part of the house while vacuuming a different part of the house. Even if the roomba is doing the vacuuming I’m still doing the more laborious task. I’m not sure the roomba will ever take off. Maybe an all-in-one biped cleaning robot but not sure how far we are from that level of AI/robotics.

most office workers are idiots, not hard to simulate idiocy with a lossy compression algorithm

think Kamala word salad

have you ever actually worked in a government department?

this is the jobs they are "displacing"

people with 80 IQ and obedient little pigs

We have used a Roborock for 3 years and it is fantastic with only a rare hiccup. Would never go back. Now waiting for the lawn mowers to catch up. We are on the cusp…

Running the vacuum is one of the few household chores I excel at. Not to mention the fringe benefit of running it when the kids are too loud or to annoy the dog.

Robotics has to over on these aspects too.

Seems reasonable in the realm of humanoid robots and some non-human ones like the Roomba. However many types of non-humanoid robots have proven quite useful in many settings. Many of them are rooted to one location. Others are used things like bomb disposal or as drones in warfare. Impressive progress has been made in extremely agile robots as well. I expect some of those to be used in emergency situations and yes unfortunately, in military applications.

My robotic vacuum cleaner regularly malfunctions and leaks mopping water underneath it. I have laminate flooring, which is starting to get damaged by the water. Sorry, but the floodgate moment hasn’t arrived yet. 😕

This is the newest trend on CES2025:

https://youtube.com/shorts/vHVQxXVgBm4

Most people can pick up a sock faster than this vacuum-cleaning robot 😅 - Thanks for the perspective!

By the way, I have the book ready where I mentioned you. Can I DM you a digital copy?

I think, taking the roombas as an example to measure is logic. Especially cause more people understand it. To me it seems that we'll follow the one special task path for now. I don't think there will be the one robot for many tasks soon. But I'm looking forward to it and plan to do my Landdesigns considering the future influence of robots in the landscape. This will be a dramatic game changer in farming.

I reckon having access to AI, like those fancy LLMs, right on our phones is a real game-changer. Folks don’t see computer intelligence as some sci-fi movie thing anymore. People are starting to realise just how powerful machine thinking can be.

Businesses, too, are getting inspired by this leap forward. They're asking, "How can we make these dumb vacuums smarter?" AI has definitely raised the bar, and everything else is going to catch up soon enough.

They're designed poorly. They aren't easy to clean when stuff gets stuck in them. They're super loud, and why tf is there a camera on it?

I've had the opportunity to design novel robotics technologies and witness their first use cases in controlled settings (pre AI boom).

Robots might not be replacing carpenters in variable settings just yet, but they definitely are taking more and more market share in surgical settings where the workflow/steps are well-defined.

Now with AI all bets are off - we're going to see some insane stuff in the next few years.

Making complex thoughts simple is your skill

I like cleaning my house and vacuuming. While other people waste time at the gym I am cleaning my house. Outside I do things like cut back trees, power wash my deck, house, sidewalk and help my wife with her gardening.

I even sweep out front of my house, rake leaves and pick up trash around my house too. It's great!

I maintain the view that we will see a robot CEO before we see a robot plumber.

My view is that this bull market has nostr:npub1a2cww4kn9wqte4ry70vyfwqyqvpswksna27rtxd8vty6c74era8sdcw83a so bored that vacuums are the topic du jour.

nostr:nevent1qqstekkchv83katqu96msf3lekyajd0prxpek5htl4lntvn6jqjwedcpzamhxue69uhhyetvv9ujuurjd9kkzmpwdejhgtczyr4tpe6k6v4cp0x5vneas39cqspsxp66z04tcdve5a3vntr6hy057qcyqqqqqqg7l68wk

I’ll take a dumb as shit R2-D2 that keeps banging against the wall over clean floors under SkyNet.

That I'll vacuum my floor for many years still. Robot will be to expensive and ineffective compared to my.. also, they are not humbling and neither helpin stacking sats.

Maybe a vacuum that will mine also.. but that's like 50+ years from now too.

The reality is robots are already doing tons of blue color work…they are just mostly connected by wire still and not able to “think” (calculate probabilities).

The bigger conversation is one of optimizing before maximizing.

This is where the humans need to uplevel.

Too many people view innovation and processes (in general) in the lens of “how fast can I scale XYZ”.

When, before we discuss maximization, we need to optimize.

Optimization looks at multiple factors to find the path of least resistance, which may not always be the fastest!

The path of least resistance to what? Closer.

Closer over more!

And that’s the Achilles Heel of a lot of people regardless of whether it’s blue collar or white collar.

People chasing more for the sake of more in their business, life, and impact; instead of moving closer.

That’s the best heuristic I’ve seen. One could argue it’s a small sample size or anecdotal, but roombas reached a fairly main stream recognition an awful long time ago. I’m definitely going to have to borrow this the next time someone tells me my construction company is going to be screwed within 10 years because the robots are coming. Once the robots can vacuum floors without accidentally smearing dog shit all over, I’ll take the threat a touch more seriously. Until then, I’ll continue to utilize new tech to my make my company more competitive.

Vacuums might actually be the last thing

The variety of things my kids can spill on the floor exceeds the computational capabilities of all silicon in the known universe.

Few

Robot vacuums suck.

This is my yardstick for self-driving cars, if we cannot nail a simple task like vacuuming, or mowing grass.

Yep. Roomba is just using the Frisbees with wheels to buy up all the real push vacuums... Then, in a partnership with Boston Dynamics or Honda, either a robotic dog or robotic humanoid will come with an included push vacuum cleaner to vacuum your house the old fashion way... and, of course, spy on you. 😁

видимо еще не время))) так как это просто ... %

Yes.

I looked into a vacuum robot and quickly realized it takes more effort to constantly clean the floor of all the stuff lying around so it can operate efficiently than to just quickly vacuum myself.

And I can quickly move chairs around from under a table and such which the robot never will.

In short, Moravec’s paradox ….

I'm not surprised things like roombas aren't all that widely used. Vaccuuming isn't particularly dangerous work, and roombas aren't all that cheap. Things like, servicing nuclear meltdown sites and other hazardous locations, however, is a great use of robots that does seem to be fairly widely used. Also, the use of drones. Though, I suppose it depends whether you consider remote controlled machines in the same space as autonomous machines, because drones and things like Vicarious Surgical's "robot" might not be fair to compare to fully autonomous systems that act purely algorithmically.

Maybe at some point we'll look to robots to do things that aren't so difficult for humans to do, in the same way we use dishwashers now, but I think most things we recognize as robots just don't have the best cost:benefit ratio, especially when we consider the inertia of having a routine that already has us doing things manually.

This is an interesting coincidence. This morning in the shower I was thinking that after so many year vacuum robots still don't do a basic thing: remove the dust from the skirting board. And I also planned on my mind a 3D printed accessory for that.

Yes I know, I should spend better my creative energies.

I want to keep cooking for my children, watering my plants, having barbecues, and sweeping my house. Arguing with my wife about who cleans the dishes. That’s life. Everything else, I’m sorry, takes us away from life.

We have a robot vacuum cleaner and we use the crap out of it. It is marginally better than manually vacuuming because you still have to do a bunch of work: clean the filter, move all the furniture off the floor, and put obstacles in the way to block off an area. And then you have to leave those areas for an hour or two and move everything back afterwards.

It's very disruptive and only cleans about half the floors. It saves time but I wouldn't say it's less work. Last year we bought a really nice manual vacuum cleaner for the rest of the floors.

I don't give a shit about backflips. Show me a robot under $1000 that can clean it's own filter and move furniture without breaking it. That would move the needle.

I agree with the conclusion that robots being able to accomplish generalized high-skill trades work is a long way off, however, I think the heuristic doesn't work. There Roomba is like a narrow intelligence, say Deep Blue playing chess. Deep Blue beat Kasparov in 1996. Going on 20 years and still no AGI. So, I don't think that if someone solves for a great Roomba that signals tradesmen robots are around the corner.

I think about this a lot. I am an electrician, mostly working on residences. I would love to be able to hire an assistant bot. But the amount of adaptation that is needed for the specific puzzle of each job seems tough to solve for on the cognitive side. Even more so, the physical side, moving through crawl spaces and crawling over attic rafters, seems very far away for robots. Maybe bots that work in commercial settings building conduit in warehouses that are wide open, but homes have some tight spaces that wires run through!

Honestly, I don't think it'll make sense until it's reliably self cleaning or until you've got another general robot around that can clean it. The maintenance requirements of these machines is a major hindrance to their adoption.

call me old fashion but my broom and dust pan work just fine. and if it breaks I'll just fix it in 2 minutes. 🫠