Even something as simple (and very, very low cost) as making up a room in a hotel, is probably 25-30 years off at least.
You don't just need a robot that can make up a room as good as a human does it, it also needs to be cheaper than that human to "hire".
This stuff is not small. The cost reductions achieved through shrinking transistors are irrelevant here. You can only go so cheap with moving parts that need to be strong and durable.
For widespread adoption, cost is probably the biggest factor. Actually getting close to the same performance of a full time worker who grosses under $20k a year for a lower price is going to take a lot longer than many people seem to think. Throughout human history, we've repeatedly solved labor problems by throwing heaps of desperately poor human slaves at them, and we continue to do so-- competing with that model will be difficult.
Fully automated "Cleaning a hotel room" is actually an insanely complex task across tons of disciplines - solving that problem will also result in you solving a lot of the Artificial Intelligence challenges.
I think what we'll see much earlier, is a hybrid solution, where automated bots that can do 99% of the tasks send an interrupt to a controller when they encounter something they aren't certain of, such as, "Is that a lifelike doll laying on the floor, or a child lying very quietly."
I think the trickier aspects of hotel room cleaning could be mitigated by creative design of the accommodations.
If it was a stark room with a bed a couple tables and a flat screen there would be little to knock off tables or break.
Make the whole bathroom operate like a giant dishwasher and come up with a more robot friendly fitted bedsheet then use some kind of industrial roomba.
It wouldn't be cheap but it can be done.
Depending on how far you are willing to stretch your definition of hotel room I bet we could automate the cleaning of those Japanese pod/tube hotel rooms right now.
Maybe those are one and the same.
I would say it's far fetched and unlikely except that's kind of how the world is today.
Some things are indeed not replaceable by robots and I think "most" professions* may not be able to be replaced by robots. These could be things like professional chefs (not fast food), programmers, and masseuse.
*By professions I mean the raw number of unique jobs not the number of people who work in those positions.
That being said in the U.S. with the "College Driven" economy I doubt it will ever affect 95% of the population at once. Generally professions will disappear one by one, but those people can retrain and if they don't it won't matter as the next generation will obviously not train to fill positions that don't exist (that's not logical). Eventually the new generations train to fill different job positions.
http://www.humanesociety.org/assets/pdfs/hsp/soaiv_07_ch10.p...
On balance the development of cars was good, obviously, but let's not kid ourselves, the downsides were significant.
Rather most of them are indebted for decades to pay for their complex equipment.
I believe the introduction of robot will follow a similar pattern, some industry will see their number of employee reduced drastically but they won't necessarily increase their margin. The customer will likely win.
The thing is that people who were employed in farms moved to factory. Nowadays I don't know where people who lost their job to robots will move.
Even when ordinary people will reap benefits from robotization, the big bucks will go to the Google or Apple of the day. We might have our daily needs met by the future economy, but big companies will dominate the finances. We'll get to play with the new toys, though.
We already replaced humans for many tasks such as assembly etc. The next step is cognitively challenging but physically easy tasks like driving cars.
A lot of "simple" human tasks such as making sandwiches in a restaurant require fine motor skills, creativity and other very difficult robotic traits. Those jobs will be done by robots long long after we as humans stopped driving cars or building them. Given the budgetary difference between making a sandwich and going into battle, we can be pretty sure sandwiches will come last. If anyone is still around to eat them by then that is.
Barring "replicators" (a la Star Trek), or labor costs rocketing through the roof (perhaps from something like a "minimum income") I don't expect the general composition of restaurant staff to change in the next hundred or even two hundred years.
A portion? Yes. 95%?
Now what would be more impressive would be for it to get down on it's knees and autonomously change a tire. 4/10 if it does this in a controlled environment, 8/10 if it adapts to tools/bolts being dropped, 10/10 if it does all the above outside on the side of a road.
One thing to keep in mind is that many 'menial' tasks are more complicated than they seem. Because much of the control is done subconsciously they don't seem that hard and they are harder to reverse engineer.
Also you say "no costs" but do you honestly think these would be maintenance free? There would probably be a fairly significant cost to keep a fleet of these robots running as they are incredibly complex as compared to say an amazon warehouse robot.
The hardware is almost there, at least, for general tasks. I'm not sure about software. Is there anything recent showing machines reading human-targeted instructions to accomplish some task (change a tire, set up a desk) and then fulfilling the task, without having been programmed or trained on that task already? That would be impressive.
Btw, why do people always worry about the jobs technology will take away? The good old days when everyone was a farmer.
Because if technology destroys your industry you lose. Automating textile work lead to greater wealth for the world as a whole, but it didn't make textile workers better off. The fact that a new job was created somewhere doesn't matter to them if they can't do it.