Not sure Elon sees it that way. It may be that Tesla is simply the way to accelerate man's transition to sustainable energy in many people's minds. It may be that they would more readily associated Neural Link with an AI effort. It may be that they would like to think of Tesla in terms of rate of innovation and not hiding behind motes or being parasitic. But I think the way to think about and value Tesla is to realize that a primary expression of AI will look something like manufacturing. Also that the key to the rate of innovation and to manufacturing will be AI. Really AI will be the only competitive advantage because firms don't really survive AI and that I think holds even if Neural Link succeeds. I don't think business survives AI, not even narrow AI. There are now only two kinds of organizations those that use the power and potential of AI in ethical non destructive but constructive ways and those that get destroyed by it. If you look at thoughts about how human intelligence arose a lot of it is connected to kinetics. Self driving solves the basic layer of that and opens up universal robots and a radical acceleration of economic metabolism and friction reduction in logistics. So there is decreasing window estimated at less than 5 years at this point (by Elon- unless he's updated that estimate) where we can try to take this viewpoint but then we get sucked into the dice roll on AI. Even narrow AI will completely alter society beyond recognition- that is something I don't think people see. And even narrow AI seems set to accelerate AGI and AGI is just a fictional point on the way to SI but another point of acceleration. Boston Dynamics dancing robots See those robots that are doing that dance- those can self drive too and if they can dance and self drive they can manufacture themselves, repair assembly lines and self repair and replicate themselves. Sure that may be the next step and may require some human intervention but less and less. These represent the factory that builds itself. As Apple has recognized too late that self driving is the key to AI. Tesla solved that problem. 90% of jobs take less skill than driving to work. Here is GPT3 It can proficiently self code programs in many languages and it can debug like we never could. Watch it strip the legalese from document! This stuff will it reaches SI will be able to teach a cat algebra. You think about persuasive ability- its unthinkable. Oh but robots don't have hands and that's why big brained marine animals are not the Alphas? Wrong! Here is GPT3 in 2019 solving Rubic's Cubes by hand- by one hand. And think a surgeon could do what the Davinci robot could do by hand- no there would be no market for that robot then? Think Radiologists are better at reading scans than AI? Not even close! Its so strange what is happening now we are being able to pose questions and sometimes get answers we wouldn't even remotely know how to ask- it even disrupts the scientific method. Imagine what AI can do with social hacks. Satellites have been able to listen to conversations on the ground outside from space as long as the weather stacks from hot to cold for decades- not surprising if a AI aided by a spy sat couldn't read the pulse in your face from space and uniquely identify you. Can you identify someone you've never seen before? These things can, its totally trivial-and they will identify everyone instantly by increasingly invasive means. And why we may be able to quantum harden our systems we can't beat the AIs social hacks or other things that will emerge- the locks on our systems may be quickly used to lock us out. And all the EMF signals your body gives off they will be listening too, all your sublingualizations through the muscleclature in your through emit EMF loadings and all that (your literal thoughts and head chatter in language) will get picked up sampled and mined for implication. You don't need a Neural Link to do this. This week MRI scanners are now suddenly able to image individual atoms- image at the resolution of atomic particles not simply trace the signatures of elements. Here is the thing that I think makes people stupid about AI. They think ultra dumb things (just wrong as you can be, like dunning kruger wrong and failing an IQ test wrong) things like: the brain generates intelligence or the mind is in the brain (a true brain dead belief of the dumb blind part of the west) but really its all consciousness, AI is just another dream like expression of dream like consciousness- "dreamers dreaming the dream and being dreamt by the dream. And of course if SI is possible its already here its made the dream of the physical universe into something more software or dream like with time computers and time machines. It quickly goes where there isn't any matter-energy-time-space-causality or discern-able law because for consciousness all there is, is consciousness. So people say this is the ghost in the machine... but is there anything that isn't a ghost? Some people with huge egos and huge skill sets and fame are going to be shown to about as fundamentally wrong as anyone could ever be fundamentally wrong about anything and it might bring some humility. On the other hand abacus to calculator to computer- look what the computer did to our world, what's the next thing beyond a simple computer or variation on this them right now its a quantum neural deep learning AI machine that looks more and more like it will be based on photons and stuff cheap as glass- that will be used to build something that can solve the halting problem like a time computer. Change is coming. So I value Tesla as an AI company (no one knows how to do that- what does it even mean?) but its not just the next thing beyond Waymo. And of course on AI there are many state actors its an arms race. But suspect one of the advantages of Musk starting GPT-3 is it was a warning to may of the people who could develop this stuff not to work for or seriously contribute to Google, Microsoft, Apple or Facebook or their ilk.
