The Upcoming AI Future | Kevin Kelly | The Tim Ferriss Show

foreignhere's what I expectthe thing I want to emphasize is thatthere's plural there's AIS so this ideaof the monolithic AI taking over they'rejust they're like machines and they'rethey follow the general engineeringMaxim which is that you cannot optimizeeverything there's always trade-offs sowe're going to engineer these AIS to begood for certain things but not as goodas something else in another dimensionand we already see that with say theimage generator some favorite artistssome favorite photography there'll bedifferent personalities the one thatdoes painting the best probably isn'tgoing to be the best for writingthere'll be some kind of transformativeor anything bad at hands and so I saythe general stance we're going to haveis what I call dumb schmarten okaythere'd be really smart and is thatPennsylvania Dutch well I don't knowit's kind of like that it's Amish Martinwe're just going to be furious it's likehow could you be so dumb when you're sosmart about these other things this isgoing to be their typical reaction it'slike you're insanely brilliant butyou're so dumb here sounds like half ofSilicon Valley exactly and I'm smartit'll be more and more difficult to kindof generalize but what I'm saying isthat they're going to be engineered forspecific tasks primarily and there willbe a general one but the general onewould be kind of like like the Swissarmy knife yeah you know it's like goodgenerally but not really the best in anyone tool yeah that's the engineeringMaxim so we should expect multiplevarieties of theseand for me the best stance is to thinkof them as artificial aliens okay aliensmeaning they could be like Spock oryoded very very smart but theythey're just doing things differentlythan we were doing if they have a senseof humor it'll be a little offbut that is actually their benefitbecause they help us think different andthat's what we're going to be using themfor that's what people are already usingfor is generate ideas like there'sprobably an idea that no human wouldever come up with and that helps me comeup with a new ideathe third thing I would say about theAIS is that most of them will be unseenthey'll be behind the Office operatingthings the plumbing the infrastructureand that's actually a sign of theirsuccesstechnology succeed when we don't seethem anymore we don't think about themthey become boringand that the majority of the stuff won'teven be outward facing it'll be justbehind the scenes andand then this idea of consciousnessConsciousness is a liabilityyou don't want your car to be consciousyou want to drive you don't want to bewary about whether it should major infinance you want it to focus on the roadso there will be advertising AIS ascautious free dumb and obedience yesexactly extra thirty dollars per monthexactly soI would say a couple things one is Ithink AI overall is under hyped but thecurrent version we won't even call AI in30 years we'll look back and that's thatwasn't it and so it means that there'sno AI experts right nowbut in the short time we're probablyoverestimating this idea likevast unemployment stuff not in the nextcouple of years for sure so everythingyou've said makes sense tools will getspecializedthey will become so embedded that wewill cease to think about them hopefullyand hopefully right in the same way thatyou waved at the lights you know we haveall sorts of Lights in here butit's not like we walk into any room withartificial light and we thinkgood Lord what is this miracle ofengineering and human ingenuityI think most folks be like okay okay sowhy is it under hyped right well whatshould surprise people or what what arepeople not appreciatingso I was involved with the internet Iwas living online for at least 10 yearsbefore 1992-93 when wire startedand in a certain sense it's likewe couldn't get anybody to take itseriously it was dismissed as teenageboys stuff and it was kind of that'swhat it was butI felt like no this is like this isreally significant this is reallypowerful and what changed it was aninterface change it became visual forthe first time in the web was picturesand stuff and that's when everybodySwitched at it most of the AI happeningtoday has been happening with all thesechap has been happening for years what'snew is that we now have an interface wehave the conversational it's this ideaof the large language model we have aconversational interface and thatsuddenly was likethe power that's been there for yearshas now suddenly accessible it's likehaving the pictures on the weband so we're suddenly thinking about itsuddenly in people's faces in the sameway that the internet was completelyFringe and then when the web came alongit very rapidlybecame mainstream I remember the firsttime I saw my gas station the pump therewas like a URL I was like oh my goshthis is like it's here this is I havethe same feeling right now happeningwith the chatbots and the and the imagegenerators is these capabilities havebeen around for at least a decade butnow what's new is we have aa language interface a conversationalinterface with them and the powers issort of completely in our faces now andso where do we go from there I think weare going to then start to apply this toeverything right it's going to be I meanas we speak every day there's peopleembedding this and they're going toembed it with this interface so I thinkwe're going to move to thishaving a whole nother level ofinterfacing with this machine withlanguage andthat's very very powerful we'll just gothrough the whole thing it's liketake Xand thelanguage interface to it that's reallypowerful