Another thought: It could be the beginning of a self driving food truck.
Much the same reactions pop up after every batch [1], including the "sign of YC decline" comment elsewhere in this thread. Maybe they're right, but we won't know for a few years.
There's one big factor here for sure though: nascent things seem trivial, so it's easy to dismiss them. That's the null hypothesis about dismissals—most are fully explained by it. It's also the #1 problem of early-stage investing. pg has written about this a ton, most memorably (to me) at [2].
Little things are easy to belittle. Their value is hard to see, their triviality is obvious, and the vast majority turn out worthless anyhow. If your goal is to be right as often as possible, it's a sure bet. But startup investing is all about detecting that hard-to-see value, and for that it would be a bad sign if these reactions didn't pop up after every batch.
I do wish the HN discussions could be a little more varied, though; at our worst (or maybe our median) we mostly emit versions of "No wireless, less space than a nomad, lame." But this may not be possible. Dismissals have statistics if not economics on their side, "wait and see" is not a compelling message board argument, and the discussions invariably have incomplete information—launch stories don't come with the complete vision and plans of the startup nor the thought process behind the investment.
The habit of looking for ways that things could work instead of reasons they probably won't is too rare to hold its own in a public discussion. That's sad, though, because it's so intimately related to the creative process, and if there's one thing I could change about HN's culture it would certainly be to enrich it with more of that.
There are obvious tradeoffs, but I could see it happen
Assuming the machines are too expensive at first for an 'everyone buys their own' model to scale at all, seems a better way would be to do away with the (massive) cost of keeping premises in prime locations, and still hiring at least some staff to take care of things like security and cleaning, and switch to a 'machine makes food in small cheap warehouse, food is delivered to people's homes, or very small outlet points in cities' model.
In fact it makes me wonder why this model isn't more widespread already (there are takeout restaurants, of course, but they're all still in relatively expensive locations that you can always walk into). This would work fine with people cooking the food, obviously, but somehow it seems a bit odd. Maybe the cooking being automated is enough of a shift from tradition to make another big change like this seem more natural?
http://www.paulgraham.com/swan.html
Personally, my bet is that Nebia is that startup this batch, and the vast majority of the other ones fail.
Came to a similar conclusion in the past but looking through the current batch I was impressed with the sheer number of projects and their applications. Makes software only companies look somewhat lame. Remember a lot of these hardware startups start off looking "toy-like" at the start, and will evolve.
Particularly interested in the sandwich robot. Lots of problems to solve there involving supply and safe storage of edible materials.
Bank tellers, petrol attendants, supermarket cashiers and packers are all examples of where machines have replaced people. Where time, cost and convenience are concerned it appears people are redundant with the exception of services with very high level social interactions.
"Can it match the higher level of quality provided by a local restaurant?"
Nutritional match: Let's face it, no machine will ever match the quality of of favourite pâtisserie [0] in terms of taste, style and presentation. However a machine with a suitable API and focus would be able to outdo out-scale the best trained chefs with individualised nutrition.
Given the western world has a glut of cheap, fast and fatty foods, a machine able to build a menu selection given the individual nutritional and taste of thousands of individual customers can excel in terms of choice, cost and diet.
[0] Although you cannot see sandwiches, baguettes, they are available: https://www.flickr.com/photos/bootload/albums/72157622498559...
You get the idea, right?
As soon as you're paying over $10 (or really over $6-7) it's hard to imagine a massive benefit from today's robots. But I could definitely see cooking robots dramatically change quality of food available in the $1-9 range, especially in locations without a full kitchen available, and especially breakfast / brunch stuff.
Edit: I upvoted you after people started to downvote you. Not sure why they are doing it. Your point is valid.
Plus, going out for lunch is 25% about food and 75% about getting out of the office and walking around in the world.
I'd imagine the primary market for such a vending machine would be the small population of workaholics who are too far gone to take a 30 minute break in the middle of the day, or McDonalds competitors in high-minimum-wage areas. But honestly, it seems like McDonalds would be able to beat any comer (maybe even without having to license any IP).
Tough market :-)
oh shit, startups that people thought were bad ideas are unicorns.
Some time later .....
I don't get it. Hmm ... UNICORN FUND IT!
Jokes aside: I think some of these are interesting, rather than gimmicky. But it all lies in the execution.