Re:It takes a lot of energy for machines to learn
by pelasaco on Wednesday December 30, 2020 @03:55PM
(#25585990)
It still takes less energy than raise humans and educate them to do mediocre jobs.
[ Parent
| on HN ]
Re:It takes a lot of energy for machines to learn
by joe_the_user on Wednesday December 30, 2020 @04:28PM
(#25586335)
There are very few "typical human tasks" that machine learning actually does better than a person at; playing game all the occurs to me. But driving, image description, language-translation, and so-forth, at the scale that a human does it, the human is generally still considered to do tremendously better. Of course, some ML programs are convenient in being able to do things at a large scale but a majority of what ML programs do is like language translation - not great and what we sometimes live with 'cause it's free.
Basically, one should really not discount human ability, especially in tasks considered mundane. Such skills often are often considered "mediocre" not because they're actually easy but because all human can do them, see:
Re:It takes a lot of energy for machines to learn
by mortehu on Wednesday December 30, 2020 @04:40PM
(#25586478)
ML is usually better at speed. For example I use a fine tuned GPT-2 for code autocomplete in vim, and even though humans could do better, they can't do it as fast.
Re:It takes a lot of energy for machines to learn
by pelasaco on Wednesday December 30, 2020 @05:41PM
(#25587106)
The same was reported last year. ML could offload lawyers, doing many tasks done by them. Not better, but faster and for sure, spending less energy.
Re:It takes a lot of energy for machines to learn
by joe_the_user on Thursday December 31, 2020 @05:38PM
(#25597983)
Right now, a lot of the low-level tasks that contract lawyers do at scale for maybe $50/hour could be done by clerks for $15/hour or clerks in India for $3/hour (or something). But the lawyer paid for their time because their expertise is need to flag the "corner cases".