Update

Siikr's joke-text has been indefinitely replaced by serious-text.

It will remain this way until enough people are fighting about it.

Please do your part.

wondering how to use A* search with gradient descent given that it requires a heuristic of how far a partial solution is from the goal 🤔

an admissible heuristic, which is to say an underestimate of the distance remaining to the goal, where the more accurate the heuristic the better the search performs, but the search breaks if the heuristic ever overestimates because that can cause it to neglect better solutions.

the cool thing about optimisation by gradient descent is that it tells you exactly which direction to go in order to reach a goal whose exact location is unknown, while A* search tells you which direction to go to reach a goal whose exact location is known but cannot necessarily be reached by heading straight towards it.

then genetic optimisation can be applied when you have an unknown goal and a nondifferentiable cost function, so you don't know what you're looking for nor which direction it must be in, which is obviously not a great situation to be in.

gradient descent has always kind of bugged me for all but the simplest problems, because not everything is differentiable and back propagation is annoying and numerical futzing around is unsatisfying but the most frustrating part is that it works!

but most neural network advances seem to come from running this insanely powerful optimisation process on a network architecture that was dreamed up by a shaman behind a waterfall operating on mushrooms and intuition; I want the computers to be doing this part of the job too, but you can't differentiate over all possible networks, can you... can you?

You cannot. You may convince yourself of this by simply getting too creative in your choice of activation function.

then again I guess why limit yourself to networks, they're just one slice of the space of possible functions after all.

You tryna start a fight rn or something?

wondering how to use A* search with gradient descent given that it requires a heuristic of how far a partial solution is from the goal 🤔

an admissible heuristic, which is to say an underestimate of the distance remaining to the goal, where the more accurate the heuristic the better the search performs, but the search breaks if the heuristic ever overestimates because that can cause it to neglect better solutions.

the cool thing about optimisation by gradient descent is that it tells you exactly which direction to go in order to reach a goal whose exact location is unknown, while A* search tells you which direction to go to reach a goal whose exact location is known but cannot necessarily be reached by heading straight towards it.

then genetic optimisation can be applied when you have an unknown goal and a nondifferentiable cost function, so you don't know what you're looking for nor which direction it must be in, which is obviously not a great situation to be in.

gradient descent has always kind of bugged me for all but the simplest problems, because not everything is differentiable and back propagation is annoying and numerical futzing around is unsatisfying but the most frustrating part is that it works!

but most neural network advances seem to come from running this insanely powerful optimisation process on a network architecture that was dreamed up by a shaman behind a waterfall operating on mushrooms and intuition; I want the computers to be doing this part of the job too, but you can't differentiate over all possible networks, can you... can you?

You cannot. You may convince yourself of this by simply getting too creative in your choice of activation function.

The guy who did the first designer babies is on Twitter and he constantly posts stuff like this.

Sad to hear the guy who did the first designer babies also has bad opinions

eh,

'Bioethicist' views:

• 18% — paying organ donors is OK

• 63% — all extensions of life equally good regardless of length

• 66% — someone's life being worth living gives no reason to create them

• 40% — being blind is only a disadvantage because society is unjustly designed

https://www.apollosurveys.org/bioethics/

I think the history of biological research means you gotta be a little more specific when you're talking about what restrictions are or aren't justified; "You should be allowed to genetically edit embryos to pick out traits" is very different from "we shouldn't need informed consent for medical testing". Sucks, but it's not like it's hard to say what things in particular you think should change!

Wait hang on what?

"Being blind is only a disadvantage because society is unjustly designed" has 40% popularity among Bioethicists?

Like, I understand the sentiment, and think it would be totally excusable for anyone without "bio" in the name of their discipline, and even with "bio" in the name I could still see 40% agreeing if the statement held something like "Being blind would be less of a disadvantage if society were more thoughtfully designed", but like

Eyesight specifically is the quintessential example of an adaptation so advantageous that it has separately come about in multiple forms in multiple species. Are bioethicists just regular ethicists that added bio to their name or is there some next level take I'm missing here?