Gary Marcus

advocates of #machinelearning, I am told that you all know that (current) #ML is limited. fair enough. but which limits are you willing to *publicly* acknowledge?https://twitter.com/NotSimplicio/status/1173373706674085888 …

Different ML researchers have different degrees of optimism, but the position that current methods will carry us all the way to general AI seems to be rare. And in public, AI skepticism has always been a much more popular position to take.

Most AI practitioners don't look for general AI, and don't even have public opinions on it. ML is all about experimental statistics, and if you say that out loud, you will probably not meet with much resistance.

and i still await examples from ML practitioners about the limits they are willing to *publicly* acknowledge.

This seems like it could be addressed by a Chalmers’ style “state of the discipline” survey. Construct a survey of various questions or limits that interest you and ask for either anonymous or direct responses.

I have already reviewed the limits that i see, in http://rebooting.ai  w Ernie Davis. this thread isn’t about what the limits are, it’s about what limits practitioners acknowledge publicly. common attempt at defense is to allege limits are widely acknowledged. are they?

What I’m saying is that I suspect you have the reach to just publicly ask top people about some set of limits you consider representative. Then you’ll have at least a decent empirical representation of what real contributors to the discipline acknowledge about those limits.