Thanks for sharing this. Although, this is the position I thought he held. Most of the concerns are in the Bostrom-way of thinking, which I still think is fundamentally flawed. Also, I find comments like these very irritating: What kind of mad scientist does he think AI researchers currently are?
I've also increasingly seen the ant analogy come up: This analogy fails because human evolutionary history wasn't rooted and directed with the purpose of serving ants. (There are other reasons it fails too, but that is perhaps the most egregious) I suppose the issue is he currently sees AI researchers as mad scientists.
He's by the tree. You shouldn't have any real quests that you need to complete to get him; after you talk to a dwarf woman by the tree, you basically just talk to him and an option will open up for him to join you.
You need to pause, a lot, to issue orders and your positioning in the game is incredibly important. You want to make sure you have one person blocking and the others attacking. If you only have the wizard, then part of the problem may be that you missed Eder who is also in the same town. He is a tank character and will make your strategizing much easier since he can be your block point. If you're a melee but not a tank, then have Eder go in first to grab them and then you sweep in from behind. It might also not be a bad idea to have your character start with a ranged weapon for an opening volley and allow Eder to get them grouped and then switch to melee and coming from behind. (Also remember that might affects your damage with all weapons/spells, so if you have high damage, you'll have high damage with anything. The only exception is if you take a skill weapon specialization, then obviously that bonus won't apply to other weapons.)
I guess there is nothing that says in MoS that Batman wasn't already an established thing, but it seems a little weird all the same. MoS made it feel like Supes was the first real super hero the world knew. Granted supes operates on a global scale rather than Batman's more Gotham focus, so it would still be a big deal.
That may be so, but perhaps ironically, I think he's over optimized for the wrong goal The fear it's spreading is not a good thing. My postdoc adviser once wrote about how he in some sense looked forward to the day when people started getting upset, because of course people will, once we actually have solved the problem. But getting upset before then on over-blown fears is not good because it may lead to stifling research with bad regulations.
It's a weird situation for me. For a long time I've had an interest in ethical AI and I think it is absolutely an important topic. But I don't buy into the fear at all. The fear is effectively predicated on humans hollywood-style accidentally creating super intelligent AI than runs off on its own and because we accidentally ran into it, we didn't know how to encode the goals we wanted (or how it can learn them) and now it's a problem. (It's also predicated on computational abilities that simply are not possible; my postdoc adviser wrote a piece on it that covers some of the topics he and I discussed after reading the book, but even then it doesn't cover all of our objections to the capabilities Bostrom ascribes to it ) It's simply not going to work like that. When we have solved AI it will be because we understand what is necessary to do it and on every step of that, making the AI do what you want is a rather critical design goal that will organically develop with our technology and is something that is even currently explicitly researched by myself and others in the field. In many ways, I feel like Bostrom's book is like intelligent design (though in contrast I do think Bostrom was making an honest intellectual effort). It sounds really compelling if you're not familiar enough with the field, but ultimately doesn't work. And the fact that it sounds compelling is why it's so problematic. And I'll add that my response is not just me. Nearly every AI researcher I know has the same reaction to Bostrom's arguments.
My motivation for this thread was because Sam Harris, who I otherwise tend to like, is getting progressively more vocal at spreading the bad arguments and I wanted to see what people's perception of it was here.