> Why are you assuming that a human would be more efficient and better for the environment than an electrically powered robot?

Because bicycles use 5x less energy per mile than electric scooters, which would be a reasonable analogue for slow electric delivery robots [0].

> It is very inefficient (approx 25%) to use food as an energy source,

By comparison, fossil fuel conversions are about 30-45%, depending on the energy source [1].

> and humans are always burning energy. They can't turn off at night or when they are idle. I think it is very likely that the robot would be better for the environment than the person.

That's a really, really weird baseline to use. Turning off a robot when not performing a task is standard procedure. Turning off a human when not performing a task is not standard procedure, and is frowned upon in polite society.

[0] https://www.statista.com/chart/28710/energy-efficiency-of-mo...

[1] https://www.eia.gov/electricity/annual/html/epa_08_01.html (Smaller numbers are better. To find efficiency, divide 3412 (1 kilowatt*hour in Btu) by the value in the column [2].)

[2] https://www.eia.gov/tools/faqs/faq.php?id=107&t=3

>I think it is very likely that the robot would be better for the environment than the person.

This is your mind on HN.

Which is the cheaper power source per mile, I wonder? Electricity or bananas?

...so, are assuming that the humans stop eating/existing after you replace them with a robot?

What does "good" for the environment even mean? I always assumed it means "good" for human purposes. But if we replace humans with robots, then the goodness of the environment seems somewhat moot.

Oceans filled with plastic would be "good" for something. Just probably not us. Maybe robots?

> Why are you assuming that a human would be more efficient and better for the environment than an electrically powered robot?

Well for one thing, the robot doesn't need to exist at all. Humans are going to be eating and breathing regardless of demand for burrito delivery.