AI Super Cycle: Unlock the Next Wave of Decentralized Innovation | Faris Jebara - InfStones

good afternoon everybody I'm Ferris here from M Stones uh and uh here to talk to you guys about AI super Cycles unlocking the next ways of decentralized innovation Pros cons and tbds so we're going to first start off the presentation with a question very simple one here are we in a web 3 AI hype cycle whatever it is curious about and US specifically at M Stones really bringing out a concerted effort to understand this and we have teams studying this around the clock for the months and months on end and uh for this presentation we won't be here to bore you with heavy technical details we're going to be looking at this more from a macro effect and so in order to do that I think it's going to be really fair for us to establish a quick Baseline that Baseline is narrative drives price action any concerns you're probably not going to like the rest of this presentation next Point here a little bit maybe a little bit more controversial but still a good hot take crypto fundamentals however sustain price action and this is what mstone bases its model off of where we decide to deploy where we decide to act how we decide to partner now in order to look at the narratives it's important new narratives it's important that we look at the trends of the past and um try to glean something try to see what we can learn so we're just going to kind of quickly do a little bit of a refresher who remembers this top one up here at called metaverse it's a fun one right and then you have this new Enclave except for nfts we also remember that uh and um but it's important to know that not all narratives are built the same so we have to begin to unpack that and I think understand the defining features of these narratives so what are they right four quick ones retail speculation buzzwords lots of pivots everybody's now an AI company and uh most importantly the next billion users if that hasn't been said we're pretty sure it's not an uh a hype cycle or a promising narrative and uh to quickly hone in on some fun pivots who could forget Zuckerberg back in 2022 20 uh 21 launching off of metamasks coiles renamed their whole entire identity uh to meta to only uh jump ship soon thereafter with mass layoffs to boot little bit more of a plausible extent Nike's real attempt to get into the nft game I don't know if anybody saw the headline peep that 23 24 rather December they shut that down so it's important for all of us to understand narratives especially on the infrastructure level because infra underlines it all and we need to understand it so that's what we're here to to kind of cut in some data on and see land somewhere hopefully AI and blockchain back on track okay first stop token landscape the macro landscape this is quick coin gecko data there's 626 AI tokens with a 30 billion market cap I cut into it our team of researchers cut into that 30 billion market cap that can't be that's pretty big who is it what is it benser is only three billion what does that mean right so we cut in there's a lot of folks tagging themselves companies tagging themselves as a who are not truly Ai and for fear of the ATS on Twitter and x and wherever it is we won't be calling them out today but there are tons of them you can check out that data everybody is now an AI company it's fun why not so hype still a question mark we're not sure next point the rise of AI meme coins now this one's even better still because AI meme coins is a a meme coins and anything else is a entirely individual thing to to just meme co uh to just AI there isn't rwa meme coins categorization on these big uh schemas or any it's just AI so it's really feeling hypey and our team of course grows concerned when we see these things we can't build custom infra we can't consert our Capital our Partnerships our hundreds of millions of raised our billions of Stak if there's indicators like this but take a look at this 2.7 billion in market cap across 383 million coin 380 million 380 coins I think we're starting to see the Wayan the waning promise of getting rich quick on Meme coins it's quite the diluted amount of capital next so now it's important to look at what web 2 is saying interestingly enough web 2 is actually tting a 16 trillion doll market cap and of course within there you have loads of companies like apple your your big Mega caps and uh of course within that as we mentioned before with every narrative in hype cycle there are tons of pivot pivots the difference here is that these pivots are actually a little bit more believable because they are not feature functions I mean they are they are they are rather they are feature functions not so much backend we use AI to facilitate the product to to for example Apple's releasing their AI meta is releasing their AI Etc Tesla everybody's actually has an AI that is truly touching the end user so this is actually believable and this is where we're here set to focus our presentation on this trend 92% of these companies plan to even increase their their um AI Investments and 8% however have started to look into it and they're starting to wind down or stabilize a in a growing and research-based technology so I think that's very interesting there's something going on here we're going to go back to that uh statistic in a little bit so uh next important thing does this cycle feel different we got to look at the funding rates some really interesting stuff going on here 2024 we were the fourth AI was the fourth out of five major key category Trends fourth out of fifth they weren't the leading you would feel that it might be something different however this might be attributed to to these companies being newer on the scene seed preed series we're not sure right there's a lot of data to point it a lot of different places more discipline VCS more the capital being diluted between uh X or traditional Enterprise uh web 2 versus web 3 so forth and so on so it's difficult to discern but it is interesting to note that they haven't taken the line share of Investments and uh it's lagging only behind refi and I'm pretty sure a lot of us don't know what refi is I had to maybe we'll be talking about that next year here hot take no hype so the question is to before we jump to that it's really important to note here 131 billion invested by VC capital in AI in 2024 it's huge look at the disparity 1.8 bli we are a drop in the Hat an absolute droping out and and yet all Enterprises are going in so the question becomes how do we gain more market share where do we go from here every company can be a web 3 company and this is a trend that we need to Cal uh uh capture capitalize on so another quick data point this is important this is something that we're starting to see at the very Earth and Foundation of this now ai has been around for a few years now Chachi BT made its debut in 22 Fortune 500 now two years later are seeing AI as some sort of fundamental business risk fundamental business risk that's a 400% increase so it took them two years to realize that from last year this risk in it of itself talks about the proprietary ownership of certain things they talk about risk in um data accuracy all things that are base level statistics any con uh data scientists could tell you about and things that we rifle with so how do we gain more market shares by leaning into these very issues that they're trying to solve we have to lean into the cost efficiency the transparency the democratize access the Open Standards Etc and we also have to solve the infa challenge because it is heavy right as an INF provider what we're seeing are really complex infrastructure outlays I promised we won't I promised you earlier we won't be getting into this for this presentation but we will be talking a little bit about these infrastructure challenges now back in 2017 when F proof of stake first came on the scene bunch of providers raised tons of liquidity we were the narrative back then hundreds of Milli billions of dollars in to support proof of stake and what's now going to become soon proof of liquidity to some extent and uh it feels like we're going back to proof of work with the GPU consumption and these h100s and and uh we're just not a lot of providers just are not ready for that change right their their Finance teams are't understand the way they go to market how many networks they support it's not something that it's catching a lot of folks on the back Hills obviously we're here to talk about how we're ready and we're there and and what this actually means and what we've done not all will be coming with us so at M Stones we have three different verticals quick shill first and foremost is your node sale where we support these products we've supported aor Gaia autonomous we're supporting autonomous incredibly exciting Tech going to be coming on the scene in another couple weeks here uh and we're supporting them from their no to earn project fundamental enabling the uh sovereignty and the decentralization there excl and in the second vertical here we're getting involved in bit tensor subnets and bit tensor validation and lastly next maybe in the next few months we'll be excited we'll be able to share with you our chat and our more research heavy features within AI so more of a bit down on the Deep dive here's the uh node cell vertical that we're running over there's the more bit tensor stuff and inflective AI another very exciting partner of ours we'll talk more about them soon so what are some key infrastructure Chens there's been a decline in infrastructure provider supporting the number of networks however we are remaining strong we are now up to 90 networks we have not decreased the number of supported networks but overall there has been an intentionality to decrease and increase partnership we're able to do both continued new networks however on the same time at the same token there are now even more networks right obviously bar chain where their humongous launch monad with their humongous launch rise and just countless others tons of AI blockchain and so nobody's slowing down and what does this create it creates the need for Value streams and differentiation and uh what you're going to end up seeing is more of a balanced decentralization Cadence do you need really 50 descent uh validators in your validator set it really is calling into question that can you really expect all five major infrastructure providers to be in every blockchain everywhere right and because what and what's happening now is there is going to be needed an intentionality right a real with this liquidity fragmentation this skills fragmentation this time fragment right we're starting to dilute ourselves however this is It's really just making it very clear to the foundations that they need to be very intentional about the Partnerships they call which brings us to some really strong players who are solving this problem and that one of them is inflective the global data layer for AI their mission is actually just simply to bridge this all they see all these players they've sat back and said there needs to be a solution that speaks to all of this reflective excellent Partners excellent technology two additional beacons Aether and bit tenser we all know about Aether they've gone about this in a radically decentralized way managed compute really Bringing Down the cost already hit first ones on the the scene maintaining lead already hit amazing traction bit tenser as well again they've taken on the compute question from a subnet approach and on that subnet level they've uh managed to launch Federated tokenomics same solution to different ways of getting up the mountain deployment Solutions now ai has now coined a new world called aios two amazing companies taking this on autonomous coming on the scene in the next couple of weeks zerog G as well who's been there who uh launched earlier this year both excellent technology and uh really providing the full Suite to for uh solutions to launch and deploy looking at it from a data access point smart contract execution the whole thing so now a little bit more of a uh head scratcher right uh where where we're still trying to figure this out uh AI has its s and one of those key problems are bias right and uh it's ironic in that there's going to be this validator set trying to understand and validate the data that they process and and say yes this is true but if inherently the data is biased whatever you really done we've just dra brought cost efficiency and maybe a little bit of transparency in the cases that the day on chain which by and large it's not but those other providers were trying to do that still we run into the problem of just kind of business as usual just a technological efficiency so I'd like to ask the question we're like we'd like to bring up this point a TBD here how can AI systems create a more fair request or response from these chat bots in order to understand that I think it's going to be important for us to just quickly be educated on how the data flow Works comes in from for example any of these major marketplaces right Sahara or bit tensor first there comes the request from the model trainer itself next could be from any data major company Facebook whoever it is who's looking to train their model they put this request out there into The Ether The web3 Ether the web then that's where the web 3 Market uh AI data Marketplace picks it up and uh the miners act upon it but what's missing from here is that that data request inherently might be biased they might leave out certain key words for example LeBron versus Jordan very easy one let's focus on LeBron not Jordan let's focus on new articles as opposed to Old articles these games are played all the time and we know them that's going to be a quick one that's not too politically charged but quick search all online we've heard these and it is scary we're relying on AI to be the source of Truth and we need to know that there are issues and you know I don't know about you guys but the reason why we we're here at least much of us are here is to be that democratizing force to be that decentralizing measure and and so I think this is a real area where we could step up and Champion our ingredients we have the right stuff we have slashing we have Dows we have decentralization we've self-governed oursel for years now now maybe more successfully than before but uh it's it's something that I think that we should seek to do so question here are there too many protocols and is this hype resounding answer from MST no thank you