This Twitter AI lobbyist phenomenon is so weird. When did it become normal to give any attention to these "intellectual" types who seemingly don't do AI work but love to rave about it on Twitter with non-technical arguments?
LeCun just got rhetorically beat here because he's not a fucking rhetorician/lobbyist/politician like all "AI ethics" person.
Just FYI, as I said in another comment in this case "many (senior) researchers agreed with Gebru's opposition to LeCun's original point - see tweets by Charles Isbel, yoavgo, Dirk Hovy embedded here https://thegradient.pub/pulse-lessons/ under 'On the Source of Bias in Machine Learning Systems' (warning - it takes a while to load). There was a civil back-and-forth between him and these other researchers as you can see in that post, so it was a point worth discussing. Gebru mostly did not participate in this beyond her initial tweets as far as I remember."
Also, for anyone who thinks Gebru does not have real credentials as an AI researcher, she did get her doctorate at Fei-Fei Li's lab at Stanford (which is now the lab I am part of), and did research at both Microsoft and Google and has co-authored papers with a variety of other well respected researchers. Just sayin', she's obviously not as senior as Lecun but let's not caricature her as if she has no expertise in this area.
The algorithms and data involved are too sensitive to be discussed publicly and the reasoning is acceptable enough to even the most knowledgeable people. They can't even be pressured to prove that the system is effective at it's primary purpose.
This is the perfect way to begin opening the backend doors.
I agree with the rest of your points. The problem is that we don't know if Apple implemented this algorithm correctly, or even this algorithm at all, because the source code isn't subject to review and, even it were the binary cannot be proved to have been built from such source code. We also don't have proof that the only images being searched for are child abuse images as they claim.
Security by obscurity has never been particularly effective, and there are some articles which allege that detection algorithms can be defeated fairly easily.
I think "spirit of SumatraPDF" means that he wouldn't npm install 800K lines of code into his project 5 minutes after creating it.
I look at the speed and efficiency of VSCode and it gives me hope that it's possible to write the high quality programs that we're used to from C/C++ using web technologies. It's just that right now we don't have enough good programmers creating good programs to incentivize the others to do better.
VS Code is neither fast nor efficient. I'm not hating on it as I like it more than most Electron apps, but I use it on a laptop with a Core 2 Duo P8600 CPU (among other things) and it's pretty sluggish.
I don't know that VSCode is a great example. I just started it and SublimeText and VSCode takes an order of magnitude more memory. If I watch the Task Manager as I type into each, VSCode also uses about 10x as much CPU which translates into worse battery consumption.
Aren't they doing the complete opposite by, for example, not adding touchscreen to Macs and handicapping iPadOS on purpose so that it can't be used for real work despite having an M1?
They benefit way more from selling multiple devices for different purposes than they would from selling "THE" ultimate device.
LeCun just got rhetorically beat here because he's not a fucking rhetorician/lobbyist/politician like all "AI ethics" person.