When Tesla says that you must be fully attentive, with hands on the wheel at all times and not rely on autopilot to drive for you, those are not terms and conditions. They're operating instructions. If I put a metal dish in my microwave, and it is damaged as a result, I'm out of luck. The microwave may have still operated (for a time) when I put something in it that the instructions say it isn't meant to handle, but that does not mean the use was appropriate or reasonable. Similar concept here. Intentionally using a device/machine contrary to explicit operating instructions, then suing when something bad happens typically does not go very well for the plaintiff.

Keep in mind that this is probably on the order of a 0.1 version. Will it still be as limited in 6-12 months? Like I said, it seems like it's more of a proof of concept, and very likely a data collection project at this point than an actual attempt at cost control.

From my reading, the robot takes the patty off the grill when it is cooked to order, so while the system is pretty limited, it is not quite as trivial as you indicated. Though if it can't put it anywhere and get the next one when it is ready, it almost amounts to the same thing.

My expectation is that after some field testing, they add some upgrades so it can do more things. But I guess we'll see.

You don't need to replace a human worker 100%, though. If this machine handles 60-80% of the grill worker's tasks/time, you can combine that worker's job with another one that is at less than 100% utilization, and you've potentially replaced one worker for multiple shifts per day. And I'm not sure what the GP was thinking, but I'm pretty sure you're not going to give the thing Sundays off when you're implementing it as a labor-reducing tool.

As it is, though, this seems to be mainly a combination proof-of-concept and publicity stunt. Perhaps also a bit of demonstration (to workers, shareholders, both?) that management is willing and able to replace tasks/labor hours with automation. And of course data gathering, both to see how such a unit performs in a working kitchen vs a lab and to bulk up the amount of data the machine learning piece is using, which may lead to additional functionality.

"Sorry Dave. CIMON didn't say."

It might not. Or it might. This thesis depends on the notion that killing people is the goal and the method is irrelevant. I don't think we can assume that for all or even most mass shootings. If the individual in question believes a gun is the instrument they want to use to go on a rampage, limiting availability of some types of guns may well lead to the individual trying to acquire one by illegal means, which makes it more likely that the person will fail to get a gun and/or get caught trying to illegally obtain one.

On the other hand, in a case where your assumption holds, and the perpetrator simply wants to kill a group of people, renting a truck and plowing it into a crowd might work. But there are challenges to this type of approach. You need a crowd of people in an open place where you can drive a truck. A school parking lot after school might be such a place, but to be effective, the truck probably needs to gain a fair amount of speed. A school parking lot after school also tends to be crowded with cars, so it's not a particularly good place to get up to any notable speed. And to an extent, a truck requires the element of surprise since it is much easier to evade than a bullet that travels 25-50 times as fast. In other words, there are many variables and circumstances that tend to make a truck a less effective (and perhaps entirely ineffective) weapon than an AR-15. It could still make for a horrible attack, but the circumstances where that could be the case are much narrower.

I don't think so. At least, I've never seen a bus schedule where the times where "whenever you need to be picked up" and the routes are "anywhere you happen to be" to "wherever it is you want to go" with no stops in between.

I believe in this case, the "ride sharing" refers to two or more separate Uber riders/customers sharing the trip. Hence, actual ride sharing by users from Uber, the "ride hailing" service provider. I think they've had the "pool" option for a while now, where multiple riders pay 60-70% of a single rider fare to share the Uber with a stranger when there are getting picked up and dropped off in fairly close proximity. Presumably, Uber found that a lot of the additional cost/lag of taking multiple customers at a time from one area to another area could be cut out by having a single pickup and single drop-off location rather than multiple ones somewhat close to each other for each customer.

Yeah, it's merely staying on task. If they put a more-difficult-to-open handle on the door, and the robot continued working until it successfully opened it, I suppose that would be "teaching it to fight back against the door."

Actually, the Airbus effort is a small, car-like VTOL vehicle intended for short flights of perhaps 50 miles or less. A single passenger prototype has already flown a test flight. It might be 10 years or a bit more before it is commercially available (I've seen 2022 quoted, but that seems more hopeful that practical), but based on the comments here, that's much sooner than many people believe. Others are also in the testing phases with similar style vehicles.

Probably too soon to tell with any real accuracy what the operating costs will be for something like the Airbus Vahana, but I believe they are projecting an operating cost of a little over $1 per km. So, perhaps 25% more than an Uber ride of a similar distance.

Definitely something that must be addressed, though it does not seem to me to be an insurmountable problem. Obviously, the maintenance schedules and safety checks will be more like the current ones for planes than what is typical for cars, even though we may call them "flying cars". There are already safety regulations for aircraft of various sizes for operating over populated areas. This may not be exactly the same, but it's not entirely dissimilar either.

No doubt another valid concern, but all of the small, autonomous vehicles (from Boeing, Airbus, AirSpaceX, others) being touted as the future "air taxis" are electric, and some are VTOL planes rather than helicopters, both of which should help alleviate the noise issues.

Again, probably a lot closer than most people think.

It may not be as far off, or as outlandish an idea, as some seem to think. Both Boeing and Airbus are working on fully autonomous flight systems. And as I understand it, autopilot (the real kind, in planes, not the Tesla variety) can already handle all but the most challenging conditions, for which autonomous flights would presumably be grounded. Some carriers even have rules in place now limiting when crew can take manual control.

All they have to do is up their prices by 50% or so, and they're in the black. That kind of price increase wouldn't drive their sales down or anything, right?

Obviously it does not. We've discussed things with lower intelligence already. You may be getting confused due to the similarity of the noun intelligence -- which particularly when we are speaking of "mouse intelligence" or other animal intelligence, but also the wide range of human intelligence simply means mental or intellectual capacity as an attribute, which can obviously be low or high or average -- with the adjective intelligent, typically meaning: having or indicating a high or satisfactory degree of intelligence and mental capacity. Intelligence, intelligent. Different words with different meanings.

I'm sure I don't need to tell you that simulating mouse intelligence is not the same as simulating the activity in a mouse's brain. But that's neither here nor there. What I asked about was why you believe that AI has to be super human intelligence to be considered AI.

You don't have to completely eliminate low skill labor. If you can replace 30-40% of what a worker does and you have a staff of 10, that's three or four jobs down, approximately. Some things of that type can be (and are being) done today. Think order takers at fast food or quick serve restaurants -- and to see that 100% implemented, look at Wawa's to-order food service. It's all done at self-serve kiosks. And it is probably only a matter of time before someone decides to cut wait staff in a table-service restaurant in favor of the type of table-top kiosks that have been showing up for several years. That's been low priority for those places because the direct cost of wait staff is low, but eventually someone will decide that the savings in the many non-wage costs associated with employees are good enough to go all in. Not to say that those things should happen, necessarily. But they can and will.

Not even close. You just have to replace enough low-level labor that supply exceeds demand by a decent margin. Then the least skilled, least valuable laborers will be effectively blocked out of the labor market by virtue of the fact that fewer jobs are available and an ample supply of better candidates are in line ahead of them for any jobs that might open up.

Why? We recognize and can measure intelligence in animals, so there is a wide range of non-human, natural intelligence that has been identified. Why would artificial intelligence have to start above all that?