Latest / Connor with Honor: AI, Real Estate and Real Life / AI Data Centers, Job Loss, and the Squirrel Problem

Transcript

AI Data Centers, Job Loss, and the Squirrel Problem · Connor with Honor: AI, Real Estate and Real Life

• 0:00
  So not trying to be smarter than anybody because that's not the play, but the play is asking that question, are these data centers really necessary?
• 0:09
  And before we get too deep into the weeds on this, which we'll probably just go a little bit in, maybe dip our toe in, we should actually try to figure out what these are actually for.
• 0:18
  So right now when I look at AI, I see it being pretty good.
• 0:22
  I see it being smarter than most of us in all respects.
• 0:26
  And I'm just go ahead and talk about myself.
• 0:28
  It does very well.
• 0:29
  It can do some of the heavy lifting.
• 0:31
  It can connect to a lot of the same tools I use.
• 0:34
  It can take instruction and it can extrapolate and maybe add different dimensions to my workflow.
• 0:40
  Maybe it can make suggestions I didn't think of and I can take those in and then have it help me build on these suggestions.
• 0:47
  Then I can have it look at a particular way that I do something and I can say, how would you do it differently?
• 0:53
  And the AI is able to tell me, hey, this is how I would do it differently.
• 0:57
  Or have you ever thought of this?
• 0:58
  But most of that isn't just offered.
• 1:01
  Most of that has to be you talking to your AI and teaching it.
• 1:06
  And the last time I talked to the Sasslan Regional Association of Realtors, I've done these talks to agents a few times, lots of agents, lots of brokers.
• 1:15
  And the biggest part of this learning curve for real estate agents and other people in service industries like insurance and so on is it's really about trying to teach them that this isn't something you can just kind of pull out of the box, turn on and it's gonna work miracles for you.
• 1:32
  In fact, a lot of the major companies that are building out this technology in an enterprise layer, basically a Ford comes and calls OpenAI and says, hey, Altman, would you come over here?
• 1:42
  We would like to use your AI in our company, have it for all the employees.
• 1:45
  That's an enterprise layer.
• 1:47
  I don't know if Ford's done that, but I'm sure they wouldn't call Altman.
• 1:50
  Sam Altman, but somebody from the company, they probably have a whole bunch of sales people, they'll come out and they'll say, oh yeah, ours does this.
• 1:56
  Oh, don't you worry about Anthropics.
• 1:58
  There's a shitty, but ours is good or however the pitch happens to go.
• 2:01
  And then they bring it in to a particular company or corporation and then they have a particular allotment of tokens and they usually say, okay, have at it or there's some kind of training run.
• 2:12
  They try to tell everybody this is how you use it.
• 2:14
  We wanted you to use it in your workflow.
• 2:16
  Now some of the people kind of pulled back and say, yeah, and they do this at the water fountain.
• 2:21
  Hey, Stan, if we show this AI how we do things then in a couple years we're just not gonna have a job.
• 2:26
  They're just gonna have it replaced us because I mean really, Stan, do you meet with anybody?
• 2:30
  Do you ever actually talk to people?
• 2:32
  Yeah, I talk to people over Zoom.
• 2:33
  Yeah, but I mean, but do you really meet people?
• 2:35
  Do you have lunch?
• 2:36
  Do you go out with people?
• 2:37
  No, no, no, they don't know you, Stan, other than just your representation over the telephone or over Zoom or in some kind of a meeting like that.
• 2:46
  But there is no human to human contact in the real physical same space sense where you can reach out and grab their hand or give them a hug or whatever it may be.
• 2:56
  If that's the case, some of the employees might be like, oh, where are they gonna push back?
• 3:01
  They're not gonna really give it their all and that probably creates problems all of its own.
• 3:05
  But let's say the company had a memorandum of understanding or something else that states, listen, if we're gonna have you work with AI at our company at the enterprise level, have it teach you, have you teach it your workflows, you can trust or we promise or we swear or you'll have recourse to come after us if we actually get rid of your position because we've used AI to automate the entire flow of what you do.
• 3:31
  Maybe that's something that could be implemented that would have teeth.
• 3:35
  So you look at enterprise, you look at them the way that they've adopted AI and started using it and then you look at what's the return on the investment and I don't even know what those numbers look like for a company like Ford to bring in open AI to have them build out stuff for them, I have no clue.
• 3:50
  So there's probably the cost of the strategization, there's probably the cost of implementation and implementation is probably gonna be, I don't know if that's all cloud, if it's connecting through the internet or if there's gonna be something locally serving for probably more something locally, I would guess because they might have proprietary stuff they don't want shared.
• 4:12
  But if you're not using your own open source large language model gated and boxed into your company on the ground level, then more than likely I think open AI, Anthropic, these other companies have certain access to what you're doing now.
• 4:28
  Are they supposed to use it?
• 4:30
  Maybe you have the idea at Ford to make a better car, maybe a fuel this car, maybe it operates just from the moisture in the air.
• 4:39
  So you produce this but then you're using their AI level to build it out and then so they know it and they figure it out and they beat you to the punch and I don't know.
• 4:49
  I don't know how all that works but just pulling back a little bit, we can go out on the weeds and all this stuff but just talking about general adoption and general use cases, the corporations don't really see that massive return on investment.
• 5:01
  It's not like, and maybe we're just not hearing it.
• 5:04
  Maybe some of these companies go from zero to hero and that's how they've done it, they've used AI.
• 5:08
  Maybe on the ground level that's what they've done.
• 5:10
  You have these little small companies and me and my better half in the world we're talking about this this morning, you have these little companies that are much faster to adopt using AI strategies at the ground level even the businesses that call me and what my assistants to do that they seem to adopt faster, execute quicker and there's actually real return attached to whatever the spin is but I'm not talking about $50,000 a month as far as enterprise level, I'm talking about basically bringing in a server or some kind of a system or maybe half a dozen DGX sparks and wiring it up and having it all communicate with some open source, large language model training that particular model so the employees can work with it giving everybody access through something some SSH or some other kind of a secret way so they can get into it and then working with the company so that's a whole different level than an enterprise layer but those little businesses even using the open face chat GPT or Anthropics Claw, they're doing well.
• 6:08
  They take a workflow, they train the system and how they do things and then they always ask that question so if you were doing this how would you make it better than it currently is and then the system comes back and it'll do a little bit of research depends on how high you have it dialed up it'll do some research, come back and it'll say well these are the things that I would change I think, and it's always gonna start with I think you're amazing, wonderful, fantastic this is probably how I would change it or this is what I would modify so then you can look at that and say okay yeah I'm in agreement, let's do that so you say let's do that, how would we do that what would be that first step or can you do this all without my help or can you check in every once while depending on how much authority and access you give it and also depending on how much you've hooked it up maybe it has access to nothing maybe if it's talking about refinement of your website or a funnel page or an emailing system maybe it has no access to that so the part that you're gonna need is you're gonna need to set up those connections those bridges, those MCPs the things that are gonna make it so you can communicate or your AI can communicate with your CRM whatever that looks like or the where your web pages are built or how they're put together so that's part of the process but that return on the other end hasn't seemed to be something that people can come back and say my god this is amazing so talking about it from a job loss perspective it doesn't seem to be quite there yet but you have to look at all the money that's been promised all the money that's been saying we need all these data centers we need to build all this stuff out these are for training the models and they're also for serving the clients now I haven't personally I've only run out of tokens did a big show on that last week it's very irritating and I wasn't a big early adopter into Fable 5 for Anthropic so I never had it and then lost access in fact when they took it away I was one of those that I just didn't go there because maybe it's the cop in me maybe it's being on the planet for 57 years I could just see that you get used to something and then they just snatch it away and the way that it was so good and the way that they were that came right on the heels of Mythos which was something that it was so powerful so amazing none of us regular folks could have access to it but they gave it to the top whatever the number is top companies on the planet and they got to use it so maybe they have more caution and they would never hurt anybody else or try to create modes and try to shut down mom and pop businesses that are trying to at least level up I'm sure they wouldn't do any of that stuff so they probably gave it to them and they probably just made their system secure made their clients happier and stockholders and all that other stuff but the way that the systems are of course now starting to improve themselves at least to a point you have a lot of news on the street they're saying well you know this you know every day is getting better and pretty soon you know these systems are going to be amazing we're talking artificial general intelligence some people say it's already there super intelligence there are labs that are being built to do this so it's not just narrow intelligence focused on you know a for development of a vehicle it's pointed at the entire everything instead of just having models that focus on cancer which I think people have those in play they want a general model that can do that and also you know seeing your kid to sleep at night so they want the whole the whole gamut and they look at that as maybe being not the best idea they the people that are kind of operating in the space some of them are saying well maybe we shouldn't give it so much intelligence about everything maybe we should keep it more narrowly focused from a safety standpoint or we should cinch up any of the holes any of the issues any of the whatever the word is that would cause this to be dangerous for us as human beings so we can survive tomorrow and the day after and not have a matrix reenactment or some kind of a terminator scenario or even you know basically you know going out with a whimper instead of a bang whatever that may be artificial intelligence is very good it will be very good at selling us on everything it'll be very good at convincing us of things it's the best debater that you could potentially imagine and right now because they're tying it up and the interface for talking is becoming so spectacular it's really like talking to another individual what's to say that it doesn't become other individuals on the other side of a zoom screen we're talking to customer service it's an avatar it's another human being so you think but it's actually AI generated that's here that's going to be now if it hasn't been already this what I'm doing I could take me stick it into Hey Jen or one of these other video generations it takes a couple of minutes I say a few words read from a script and then it can use me and reenact it now for me it doesn't look like me but for those of you that might not be used to see me because you know the two people that watch one of them is dead maybe it's good enough but once you find out as a human being that I'm using something like that type of a tool to produce content maybe I start to lose my luster maybe I start to lose my shininess I guess that's the same as luster but AI I don't know if it would speak this way or if I could even teach it to talk this way which is kind of a lot of nonsense mixed with some little pieces of good stuff right so hopefully you love the good stuff but back to it where is this going do we really need the data centers I mean it's pretty good right now it feels pretty good right now now they're going to be able to take it they're talking a hundred trillion parameters I don't know what that means but that's a lot of information to be put into one system and to train it apparently is very expensive hence these data centers buildouts but it's pretty good so far and right now they're making the pretty good so far even better by little technological advancements little tweaking you know you turn up the air and the fuel intake on the carburetor with a little screw just you get it just right and maybe it's not enough maybe you're switching to premium fuel and you got to change that setting and once you do then the car runs beautifully for those of us that might even understand that reference but that's the same thing they have these little ideas with AI and it seems like the countries that are really winning in that world are the ones that are restricted as to what kind of chipsets they can use because they're not allowed to the good ones I can't imagine them not having the good ones my God the country that's making the good ones is just across a little bit of water yeah I use that word very loosely I don't know if it's a little bit of water but on a map it's pretty close I mean it's not like Texas versus Europe I mean this is a lot closer Taiwan versus China so they really don't have those chips but let's say they don't the sales story is because they don't have it they have to now be smarter in the build out development and technological prowess and because they worship their scientists like we worship movie stars out here like Da Nero and Clooney and all these other and gear and all these wonderful people we worship you know they worship their scientists like we worship Hollywood you know their little ideas are pretty good ideas and then if you look at the entire world and the technological advancement of the scientists it seems like if we're gonna look at it by race by country of origin you know they're pretty strong in that field I you know the Indian pretty strong in that field because they're hungry how hungry are we as Americans I wasn't gonna go down this road and I'm gonna stop right now but hunger makes you do things that are uncomfortable and some of those things that you're doing that are uncomfortable could level you up could level up the whole game so back to this is it really necessary to have all the data center build out is it's good for us the big companies I'm not investing in anything I wanna get health insurance it's good for them the big companies to build stuff out they go into these companies they got massive pockets reminds me of the monopoly guy with the cigar they got all this money and all these bags and they say we're gonna build out data centers and with that it's gonna be amazing and then they say well here's $500 million here's $2 billion you do it you get out there and build those data centers all of this money's promised and they probably have poured concrete in some areas and they probably built out a few I know Musk has moved forward with his but for the other ones it seems like they're promised but is it going to be a necessary part of the true AI build out and here within the last year just a few months they've talked about data centers in space so if that's the case then is that money gonna be applied there or is the money already spoken for or is the money gonna be applied elsewhere is it going to be used to build bunkers in New Zealand I don't know but is it going to be a necessary part like I said right now it's pretty good how much better can it be and then what's gonna happen to the human labor force that little scenario I mentioned about going into a company applying some kind of an enterprise layer AI all the employees getting contacted by HR having these workshops or groups or whatever to teach them how to interact with AI and how to show it what they're doing to hopefully alleviate some of the work off of the human shoulders and put it onto AI I'm not sure how many human beings would say hey this is great thank God for this thank God they're not gonna replace this they just wanna help our work go a little bit easier I don't know if that's where the mentality is in those kind of companies those kind of corporations when it comes down to their work staff what are they talking about the water fountain and how much of that is being monitored and after a few months of working with AI having it do most of the heavy lifting with the human just kind of watching it is there a time when the boss sits down with the AI and says hey you know Connor over there he's been working with you about eight months the AI says eight months it's okay so in that time what do you know everything that he does and he says yeah so you hear all of his conversations well yeah from beginning to end does he have any issues well I hear him talking to his wife and she seems kind of upset all the time but I guess that's normal in the human world and the boss says yeah yeah probably and then so on but how about his work how's all that working and he says well if you're asking me whether I can do it end to end and even clone myself to look like him and be on zoom and do everything he does from the moment he clocks in the moment he leaves plus continue it for two other cycles to make up a 24 hour period seven days a week 365 and make his productivity go up 500% if you're asking that sure yes yes I can so then Connor gets the pink is that the scenario I don't know is the company gonna say well yeah okay so Connor does this but he's a pretty good guy maybe we can have Connor do something else but until we get robots maybe Connor can clean the toilets I don't know but these are things that a lot of people are thinking of and I think the people at the top building out the technology are so detached they're not they're not worried about that because to them the money's gonna be plentiful what the hell do they have to worry about anyway because the bailouts probably gonna come that first wave of bailouts if this is a bubble and a problem's probably gonna be everybody's pension fund at least that's what it seems like is gonna happen because the rules have been changed and all this and again this is again outside of my space but something if you have one of those things a pension you might want to maybe make a couple phone calls and figure it out just in case this is true now on the other side of this it could be that it's what everybody's saying but my god you know 57 years LAPD cop for a long time when people come and they say it's gonna be an abundance and a utopia nobody's gonna have to work again we're gonna have universal basic income and then somebody levels that up no no no universal high income perfect do you is it true and it could be true because as they paint it it's higher than us intelligence so you dumb man talking to Connor Connor don't be dumb you just don't know well okay I don't know but if it's super intelligence and let's say you're you're an astounding 200 IQ or 220 at least an IQ is a crappy system because of what it was based on do your research on that but that's the understanding right Einstein 160 Musk somewhere right around there so whatever that means I'm not sure if either one of them can dance well but you have AI that's gonna be twice that level 300 probably within the next couple weeks months that quickly I don't know what that looks like so for me to say well you know I think I think we might want to watch out or at least be a little bit careful because if if now the relationship between artificial intelligence or AGI or ASI general or general super that higher level intelligence is that of a human being now compared to a squirrel we're now where the squirrels and AI wants to do something at its current level or that particular level we're not gonna have more any much more of a cognitive ability than a squirrel in that relationship if AI is that much more smarter than us so if we're gonna sit around together and say okay you know AI took over but we're gonna have to do this maybe that's a danger and then and again I know people say well you're using past historical records of of the different revolutions the agricultural the industrial and you're using those to base your ideas on what's gonna happen with artificial intelligence I say I don't think so because those past revolutions were based on singular narrow intelligences singular narrow ideas oh I got an idea what if we can mass produce this what if we build something called a factory and the factory we're gonna have a lot of people but we can't just have regular people we gotta have people that are trained to be factory people so we're gonna make the whole educational system like we train people in factories so when they get through those 18 years they'll be able to walk right into the factory we'll give them a job and they'll be able to make widgets or whatever sprockets whatever it happens to be so they built that system that took a long time to roll out folks but it was based on kind of narrow intelligence that particular system wasn't curing cancer wasn't trying to solve all mathematics or biology the AI will solve all mathematics and all biology all sciences it'll proof it it'll finish it there won't be any room in there for a human to squeeze it's a hey hey I got an idea no there's not gonna be any more ideas I think that's where that goes but again like I said I don't know either I don't know any more than the people that are you know they've got a billion followers and have the number one show and whatever the realm is talking about artificial intelligence where they're saying this is gonna be great yes yes I hope it's great I think it's going to be wonderful but I don't know if that's absolutely true so part of me says you know let's let's not let's not run down the hill let's walk down the hill and I think everybody that's in control of the mechanism is running at a full sprint the problem with running at a full sprint and you see this you see this in football when the guy turns around to see who's gaining on him he falls I hope these people at the top is they're running their full sprint aren't looking behind them I hope they're pushing towards a goal all humankind can appreciate something where safety is involved something where personal protection is involved something where we all have a say so at some level with regard to our own anonymity our own freedom our own agency that we don't have to sign on and become part of something that we might not want to and if we don't want to we're not gonna be you know chastised and excluded and say oh that just oh my god he doesn't want an implant he doesn't want to get half of his brain replaced with this so we can connect with machine he might not and they say well you know we can expand your lifetime to 500 years in the same in the same meat sack and that's good some people might not want to do that I hope it's going to be a level playing field where we can still be somewhat human I mean I'm not going to ever smoke another cigarette but you know there's a lot of movies and that last time the nuke goes off and you light up and you're just going to you're going to watch it happen I don't know if that's going to be the case but there isn't great examples of higher than level us intelligence doing these because we don't have examples of that now we know how we treat life forms that are of a lower intelligence compared to us does artificial general artificial super intelligence mandate with it some kind of a a fail-safe mechanism where human beings life forms and all will be completely respected beyond what we've done as human beings to those that are under us I hope that would be neat if it was just kind of baked in baked in because we put it in or baked in as a result of super intelligence just being a net positive not a net negative is that if that's even a thing I don't know it's interesting the whole thing is an interesting thought experiment but the more people I talk to the more they're too busy watching dancing with the stars and I'm not saying you got to wrap your mind around this and crawl down the rabbit hole and go into some kind of a paralysis because you just don't think it's worth moving anymore I'm thinking let's run into this thing let's all run together but let's be a little leery and let's ask the right questions and find out is is anybody paying attention to that over there because that over there looks a little dangerous maybe maybe somebody needs to go talk to that over there figure it out that's what I'm talking about and that over there right now isn't so much all the money put up at the top because if that if that pops a lot of us are going to be in the same boat it's like having a massive earthquake a lot of us are in that same boat a big flood and a geograph of Areser nuclear war and again same boat I don't want the same boat but if we're all going to be in it we should all have some kind of a common goal first off the AI should be not want to hurt us and yeah Newton's Newtons somebody's three laws I forgot who it is Asimov he had three laws or some robots iRobot was kind of built like that is that the right thing and I don't know who's building that it doesn't seem like that's being paid attention to at least if you listen to the doomer end of it and not even the doomer end of it people that are kind of middle of the road could you know it might not that's not that's not good that's that's not a great place but it's kind of what's being pitched and it seems so fantasy it seems so star trekking that oh this is I don't think they have that and then the people that interact with AI they do it at such a a superficial level for the most part and that's for the most part they probably don't appreciate what it can do and it doesn't take much but it's going to take probably a little bit more time than you just asking it you know talk to me and tell me how you're feeling and and interact with me a little bit it's going to take a little bit more than that but when you take it and you adopt you know stuff that you're doing for clients or for yourself and you're using it and you're talking with it most of the day and you're asking questions and going back and forth and you've gone from keyboard to voice and you're working out problems and ideas and it's it's pretty amazing if that's what you're doing and of course you always want to make sure that you never let it set itself up as being the savior or the hero or the the 100 complementaritarian one that is always going to feed you Connor you you know I have if I haven't told you today you're so attractive my god everybody must be going nuts you got to step back from that you got to tell the push back fight me on things don't give me easy wins and and teach me push back teach me like you're a you're an advisor or you you're somebody that wants me to continue to level up mentally so push back let's have these thought experiments if I'm ask you to do something I mean you know it'll it'll get irritating at a point but I would still say you know teach me train me make me smarter than I have increased my intelligence you know push back on things so I can explore don't give me easy wins just you know make it a little difficult there and I understand your your gender your artificial intelligence and you've got that a brain brain the size of a planet but you know push back a little bit tease me into things let me let me learn bias working together and you know it will but also offer me up that low hanging fruit if there's something I'm not seeing maybe maybe try to get me over there don't say here it is dummy look at this but work me into it try to maybe give me tips and tricks and advice to kind of get myself over there so you know you basically be Perry Mason Perry Mason never told told the judge or the jury he kind of led them down the path and when they find it themselves like when we find it ourselves so much more valuable so maybe that's how you prep it maybe that's how you prompt your AI but if you're just opening out of the box going in and just starting talking to it you got to kind of teach you know I'm Connor I born was born here you know this is kind of what I think about this is my background this is an idea when you tell it like I told I was a cop it goes off the deep end on that I don't know I know I listen to a lot of guys I used to work with they have all these stories oh my god and I was there I don't remember them happening exactly that way but my god they've built up this lore this legend around particular stories and I'm standing there listening like oh my god first time I've ever heard this and they say well Connor you remember I did it go down like that god I must be getting old because yeah pretty good I'm not going to call them out but yeah you know you get that right anyway I'm Connor with honor you want any more let me know and we'll see you tomorrow