42. Perplexity Raised, Smaller LLMs and 2024 Optimism

hello everyone welcome to the cube pod the cube podcast episode 42 I'm John furer with Dave Volante this is our weekly podcast we review everything we're looking at analyzing and uh talking about and mainly looking out in the future and the trends bringing together the cube research formerly Wiki Bon Cube video and network and community and of course silicon angles coverage all kind of discuss the days and and weeks ahead Dave got a lot of action great to see you happy New Year first pod of 2024 and uh wow what what an how was how was the snow in Tahoe literally on the west coast there's no snow I saw some tweets and posts from black home Whistler British Columbia it's like mud flowing everywhere really Tahoe had about two feet of Base um roughly and then they got eight inches so we had good one good day there nice I heard even in Vermont there's nothing up there the broke Lakes aren't Frozen nothing's happening um except it's coming here on your side you got a little storm coming yeah we're gonna finally get get some snow but uh but so the skiing was was pretty good then we had I had one good day skiing and um and I took a break with kids coming home um my daughter Jacqueline and her her boyfriend in Manhattan um a lot of good New York stories coming up on the cube so I can't wait to talk about that but just great new year to reflect um really kind of did a detox Dave um after reinvent um took some time not a lot of tweeting although some targeted tweets uh I put out there mainly I saw some news I wanted to comment on and identify some of the trends that a lot of people have been talking about so um other than that took a little detox kind of zoomed out took a perspective trying to put the goals together for 2024 you know in in the um the landscape of media right now it's a bloodbath I mean cheddar just had a Furlow companies going under in the digital media area um thankfully our business is robust and our investment in you know free content Network effect high quality analysis community and real hardcore data and Analysis has been paying off not a lot um not a lot of other companies can can pull that off and so we've been doing well so I was very thankful this year that we got a great team our um 13 years with the cube going on our 14th season and silicon angle I'm super thankful for for our team our customers that support us and sponsors and our audience who over the holiday did a lot of outreaching little testing with and and I think it's going to be a year of the community for us this year and so I'm very thankful reflect on that and I'm super excited to gas it up in 2024 with the cube um go out and change How We Do event coverage and I'm pumps I had I have this this calendar my wall it's like a little little calendar white you know you can you can write on it and erase it it's like an eraser calendar and I took down last years's and put up like last week just before New Year's I put up the new one and I started filling it in and it's like already it's this sea of black ink I'm oh boy yeah my my wife did was like when are you traveling again next I'm like let's not have that conversation yeah I mean there's a lot of things going on what's interesting is is that um you know one of the things I did over the holidays was really personally reflecting on our 13 years together with Cube and you and I partnering just it's almost like a whole another inflection point this whole AI wave has brought our business to an inflection point and you it reminded me of that Andy Grove talk he gave at MIT in 1996 the atams are not yet clear and they're going to come together and you know I saw Matt Baker's post at at Dell about Rag and some of things we've been seeing early you know this this AI everywhere all the time everywhere all the time moment is aily compelling for our business what we've invested in and so I had a lot of personal reflection and I think this year my personal goal um is to lean into AI heavily go deeper into that rabbit hole and bring that expand that in from our business perspective but also do more out and talking about going to New York and having a good presence there get great feedback there um CU Global is looking good again back on the table so we're kind of post the covid back to 2019 it feels like kind of the things that we were thinking about in 2019 are on the table and so um I'm super excited I think it's going to be a year of of of reinvention for the cube I think it's going to be a year of new things and AI has been the hype and and again the news is all over the place again it's AI everywhere platform media scale and it's going to be interesting lot of fake data coming out fake content AI is being like we predicted it's going to start polluting the market I did my look back you know like you say it's a time of the year to reflect and one of one of my personal big goals in 2023 was to just do more research you know non-paid editorial research like I've done with breaking analysis I was I was really stoked that we hit almost 650,000 downloads this year we had great guests I think the fact that we did what did we do four super clouds this year and three St super Studios and what I liked about them is even though the super Studios were paid they were uh industry events community events thought leadership you know by with with sponsors you like IBM like vast and like Dell who weren't just trying to pimp their products they were really putting fourthought leadership for for customers and talking about transform and disruption and so that was really good and of course the super clouds continue we've got we've already got our next supercloud scheduled for February 13th John yeah it's it's gonna be exciting I want to go through my some of my high level prediction areas that's coming up I know you got a prediction breaking analys you just did but let's just go through a quick rundown of what I want to talk with you about today I got the AI all the time uh we'll get into the predictions I got some some some summary buckets forming that you'll see uh us focus on this year but let's get into some cool news dates so let do a quick rundown um Microsoft just unveiled their co-pilot keyboard key first keyboard change in 30 years that's huge that personal because it's our generation of of it um and then open AI again more hype coming from open AI their revenue is talking about 1.6 billion and they're going to launch their GPT store next week that got delayed because of the hold debacle and the leadership crisis and that was supposed to happen in November um and then also AI open AI is offering media companies 1 to 5 million to license the content on the heels of the new New York Times suing them for Content so you're going to start to see training uh sites being paid for that's notable because we talked about on the Cod and the QBE AI we'll come to that her perplexity which was on stage at reinvent company company we talked about on the cube a lot has just got huge funding they just raised um 74 million out of $520 million valuation actually I think that's kind of low I think they could have probably got more but that's a good round for those investors and then Intel and uh uh we're know all we know a lot about out and been following and a cube alumni that was on supercloud 4 the one that you quoted about that thought exercise he's now the CEO of Intel's new independent Enterprise gen AI software firm called articulate AI so great to see Cub alumni formed and another Cub alumni Justin hotard has been picked up from HP to move to Intel to run their data center AI group to replace another Cube alumni Sandra Riviera so so so um you know that's a CU alumni on the Move action there so that's super exciting Bitcoin went over 45 as ETF his approvals have gone through anticipating that and then just more Doom and Gloom on the startups uh scene again as we had predicted on the cube startups going to fall out of the sky and new ones going to be born yet the seed funding rounds are significantly strong in Ai and other areas so again transformation shift this cultural revolutions happening that's going to be our Focus section at the end of the pod as I've said many times and we've talked about a digital Revolution is coming it's I call the digital hippie Revolution what happen in the 60s is going to happen kind of from a digital perspective and you're already seeing that with AI it's a complete generational shift um this is not your your father's internet anymore and and there's some stories that came out I want to talk about that so that's the rundown Dave AI everywhere any anywhere all all the time that's what's happening and AI is everywhere all the time so that's you again more the same so you know you mentioned that Matt Baker post which caught my attention I I didn't see it actually when it got posted but I was doing some back channeling um emails with him today and he said yeah like I said in my post so I went to his LinkedIn and I saw you had commented so I commented on your comment and basically his premise was all about you know really I thought aligned well with the power law that the cube research put out of llms on really the long tale of smaller models you could do so much more with smaller models I think he said it's like putting a dollar in machine and getting $2 dollar back um and so and and then somebody from Google chimed in and said hey you know I'd like to debate you on some of the uh assumptions that you're making here and i' I'd love to have that baate the debate publicly you know Matt Matt Baker for those you who don't know he was the senior vice president of strategic planning at Dell for a number of years and so you know you're in that job I mean you got to be have a you know quite a observation space and they put him on the head of strategy for AI now at Dell so he goes deep he's super smart guy he's he's I like Matt because he doesn't let you get away with he'll call BS on you um and but but I thought his post was was thoughtful I know and my comment John was you and you have been on that from from early on in the cycle and so I was pretty pretty pretty confident that uh that there you got guys are in the right direction my view well I mean you know we he and I both talked about LinkedIn but when we had the um big analyst review with all their top ex EXs I think I was contrarian from all the other analysts out there that said you know all large language models at that time all the top analysts were seeing the value other ones other ones were poo pooing it or just like not understanding it the fake analysts um but the but the re the real analysis was that it was a lot brewing in the open source because the canaran the coal mines and open source is the developers and so when you see the large language models come up it was apparent to us that this long tail was going to emerge or the power laws we published it that to me was the around um April May time frame when the hallucinations just started to emerge it was very clear that the hallucinations were all not were not going to go away they were going to be managed by more data but we had identified hence now our new set of research that you're driving this data layer or this new supercloud data modeling going on is changing the game because hallucinations are direct result of bad data right not enough data or the wrong data at in the wrong place or the right data not in the right place and so Matt Baker was on this too so he and I were riffing and then so when we saw that the smaller models were emerging in open source you only had to be a little bit smart to realize hm what if CPU and gpus on chips got better on the devices so to me that was very similar to the Telecom wave and the web where Transit got faster everyone got faster bandwidth internet got faster so you know he and I were ripping and others were saying hey if you assume that Nvidia is going to do their job and Intel's going to get back in the game which they are and Amazon's going to step up more compute more gpus more systems on chip more stuff happening at the Silicon layer then that advancement will only Power the ability to do the training and inference on any device so that's kind of what the thesis was and that's that was not contrarian that was kind of just obvious now you see seeing llama playing out on smaller devices so if you're Dell you're in the PC World you're looking at our Armageddon unless a new thing comes along like AI which makes your Hardware more valuable which is happening so so you know and we called that it was it was good to see it but you know they have an opportunity Dell HP Lenovo the folks making PCS and devices are in an opportunity to create a Renaissance in their business of saying look at same game new processing new architecture in Microsoft obviously and I want to mention something else about the U the power law and you I I take a lot of credit for the power law but you're the one who gave me the idea and a lot of the background you worked on it very closely with me and Rob stretche but one one of the things that you pointed out which was your sort of innovation was that torso getting pulled up to the right of 40 45 degree angle by open source and I have some brand new data that's not been released but I'll share it here just as a teaser from ETR they're going to be they're in quiet period right now and they're going to release uh their results in in in a couple weeks but llama 2 has in a survey this the latest survey of 1700 Enterprise it decision makers llama 2 has about 177% more installations than anthropic which I found really interesting um and and oh by the way open AI has about 7x of the the the installations over those two so but the point being the exactly what you said that torso being pulled up to the right by open source and llama 2 getting a lot of traction and to Matt Baker's post being applied in in in situations that are smaller the other data point is at the Deltech Summit I met somebody from meta it was like St cornered them on the bus ride about how much of their installations of llama were on on Prem and they said we don't really know but we can infer from who's downloading them and based on the ETR data I've calculated about 30% are on Prem he said it could be as high as 50% now I'm pretty confident the ETR data doesn't include three-letter U you know like the NSA agencies and supercomputer installations so that number could be much higher it's probably somewhere between 30 40 maybe even as high as 50% so that we we're starting to get data that validates some of the anecdotal information that we had talking about before so I just wanted to share that yeah and I I think that points to the thesis that you're going to see models on devices and again I think I can't say this loud enough I've been shouting it from the The Mountaintop for years and even this past year the entire Data Business is going to be upside down the next 24 months you cannot operate the data strategies you've been doing company's been doing for decades with the new model data needs to be available for any app at any time anywhere and that means either collocating the data because you can't change the laws of physics data's got to move around or be there when you need it how do you make data available and highly addressable for every application every inference if that's a new indexing change with a Spector databases or something else the entire data Marketplace Industries are going to be completely radically changed this has to happen it will not it's not going to happen with data warehouses and just Snowflake and data bricks yeah data Lakes will be around everyone's predicting that another year data Lakes it's the year of the data Lake it's been that way for years the data architecture has to be very agile very accessible and think about how complicated that is because you've got you've got all all these stovepipe you know data platforms whether it's within AWS you got multiple stove pipes within Microsoft Google you got snowflake you got data brick so you got all this data and you're not going to move that data you're going to bring AI to that data that's a big theme okay but if you got co-pilots now operating on that data which data does the co-pilots operate on that's coherent and current how do you know you know what is the system of Truth so this is a real challenge so that's why you're seeing things like you know metadata unification Unity cataloges uh you know the try the attempts to eliminate data movement and and ETL so there going to be a a a burgeoning market you know around data coherence and then up the governance issue so it's a very complicated matter and one that you know we're going to pay a lot of attention to this year you know um the folks listening if you're interested in what we're going to be covering this year Dave do breaking anoun on predictions we got a lot of people from our Cube Collective weighing in but to me in summary that you know the 2024 is going to be a pivotal year because of the realm of of of AI right the AI world's here and you're seeing it impacting VC Investments startups that are being funded look at what's being funded what's not being funded uh in Ai and Tech um plus the advances of AI are impacting infrastructure um data Technologies and security so to me data related Technologies platforms tools the Pix and shovels platforms are going to be the big Focus security developers uh and investment you're going to see those are the hot areas are going to be burning with energy either burning down or firing up and so so look at the VC's series seed and series a fundings will be a marker look at the AI production workloads and then look at the data Technologies and the apps that are the best again the Uber of the Enterprise you've called it Dave and I think the shift in the VC Investments already clear you already seeing the data come in is they're predicting a worse year in 2024 in VC than 2023 mainly because the bulk of the market of the previous bubble is going to be impacted and squeezed and or cor normalized if you will so you know I'm expecting infrastructure old school players like Dell and new emerging chip players to benefit uh and then security I think the security Market's going to be huge this year I think it's a great time to invest too U for you know all the LPS out there i' I'm going to invest in another fund coming up I think I think 2024 is going to be a good time to to launch new funds you know valuation are getting squeezed a little bit of course you know some of the AI valuations are a little Bubblicious but if you can if you can pick the right companies I think you can do really well um in you know when when when you we talked about this in earlier CU pods that that those with zombie corns or whatever they call them that were getting you know crazy funding in 2021 really struggling now with the runways um and so you know this what is it I think Michael Dell said it one time that that he loves a good rainstorm because it cleans out the streets you know so well well it's interesting rainstorm you know those that rain rain's got to collect somewhere you know the data Lakes again I was joking earlier but the data Lakes will are going to be huge again again I saw one prediction oh data L is gonna be the dominant data Arch okay I get that but if you look at the business intelligence market and separate that from say what observability where observability is going you're seeing more cloud and distributed Edge Computing so companies like Cloud flare are doing well um you're going to start to see Edge uh inference right so I think one of the things that that Matt Baker was talking about was inference is not so much the big deal because it's going to come from these models but I think he's missing a key point the smaller devices that be able to hold to llms the better inference you can do there so you know I think the edge piece is going to be big set up you think he's missing that you think he's missing that or call out I think he could works for Dell I mean he's like no I don't I don't I don't think he's missing it I think he he could have really Amplified he just didn't he didn't call it out as much it would have been it would have been the um you know the the swan song to his post because it would have been like hey the ultimate will be to prove his point is on the Silicon side is that my device whether it's a Dell device or Dell enabled device at the edge and could be as small as a wearable right so again you know how I feel about that John it's going to be arm powerered I mean in a big way yeah so again Dell's in the business of making PCS and again they're resilient we talk about Dell all the time because remember they had the web the web was supposed to kill mail order well so we know what they did they sold them online yeah they sold PCS online again they were in the PC server business then but if you look at Dell they weren't really in the server business until the web until Michael Dell converted over from PC mail order to the worldwide web where you can get online on a browser and order your own PCS they really nailed their supply chain because of the internet so the internet was a disruptive enabler for Dell itself to create a supply chain advantage and build that supply chain you know the patent they're famous put the suppliers around your area which Dell did and that's kind of historic perspective but Dell Technologies Michael Dell's success was because of the internet in my opinion that enabled him to manage the order in ease of ordering which made his supply chain more efficient and the rest is history they dominated the server business and they'll Crush they'll crush the internet there's no doubt about that and so so what what what M Baker's getting at with the strategy V's running strategy when we had that meeting with them was AI is their next wave and they will play because they're in the hardware business and Hardware matters so so if you look at all the action right now from anyone who's age 15 to 50 or higher they want the fastest stuff you because the developers are writing stuff for the new infrastructure so again developers are driving the agenda on AI if you look at the the hugging face numbers the leaderboards look at what's going on in open source you mentioned llama 2 the the Llama 2 uptake based on your research you just came put out in the pot here is direct result of Open Source and meta um hit a home run I think we talked about that pod last year when they released Lum um you know we were like hey they missed the metaverse might as well get in with AI but again AI is only going to help the VR Market da because AI will make VR better because if you go digital you're going to have ai augmenting developers and the user experience for VR I think I think the metaverse is gonna get a huge thing you you just well it's funny you just you just stated my one of my 2023 predictions was was that generative AI hits where metaverse missed um and you know you may say okay that's obvious but I just just did a predictions post with uh a bunch of cube Collective guys known as the data gang Sanji Mohan Tony Bear Carl Olson Dave menninger and Doug henin and we did one in this is our third in a row third year in a row we did 2023 no mention from those guys of geni and of course everything had geni this year um now because they're deep in in the weeds of their their data platforms and of course you know their rationalization was that open AI was a little bit early you know a lot of experimentation going on so they kind of walked that back but but so I was actually happy that I made a call a year ago that you know gen was GNA be big even though it seems kind of obvious but I'm not down on metaverse I'm I'm picking up on what you said uh but but before I get off this these guys came up with I'm gonna be posting tomorrow they are so smart when it comes to data um intelligent data platforms simplifying database design unifying uh uh uh and rationalizing uh uh metadata uh gen Dave menar gen is not going to replace traditional Ai and all these use cases and then Doug henin talked about how it's we're g to completely sort of change the way in which we think about the whole data pipeline so they did a really good job we're going to go deep um publish all that tomorrow uh we love love collaborating with other really smart analysts well I'm gonna lay out right now what I see as the editorial agenda for silicon angle in our coverage area that's going to be mainly my prediction post because these are areas that are going to go deep uh for for multiple years um and obviously they're going to reset the agenda in the in the industry landscape from Venture Capital Ai and Technology number one the IPO market and m&a Market you're going to see um maybe one or two IPOs maybe data bricks but not a lot of IPO action maybe the second half of the year for the ones that have been that missed the window but for the most part one Trust Maybe One Trust might go stripe data bricks whatever you know the the ones that kind of missed it but but have massive power and traction and relevance that are going to be obvious home runs that are already successful they should have been public anyway but we'll go public the rest is going to be m&a you're going to see a lot of dying companies Walking Dead to actually completely shutting down to a rollup m&a Market um so that's going to be one we'll be covering a lot of that stuff um Ai and data Dom um Ai and data technology will continue to lead in the funding uh with some companies getting more exceptional growth rates they hit the Home Run they get the lightning strike consumer search will see significant ship perplexities of the world so big AI data dominance new shit's going to come out of the woodwork new companies new brands we're going to see just that next Google come up cryptocurrency Dave and web 3 I'm I'm watching this so I was you converted me last year so with Bitcoin hitting 45,000 and the ETF on the horizon I think you're going to start to see the foundation of legitimate action around funding so I think you know the fraud side of the whole shitcoins and the fraud going to be go away I think we're going to see decentralization and web 3 come back with real crypto not fake crypto so I'm putting it out there G to watch it um VC let me let me just follow up on that you know you I think correctly pointed out a lot of the crypto developers went to Ai and I said hey I think Ai and crypto are going to come together and I think could be the year where we see that this year or maybe next year um I think you might see that and you mentioned Bitcoin hit like 45 you know thousand it dropped down people sold they were like hey let's take some profits so I don't think it's quite ready to rock it back up to uh you know over 60,000 but but but I think I think you'll see legitimate action come back and I don't mean that that's G be a massive but like you said developers will oscillate between Ai and crypto because there's synergies there if done properly again I pointed out some of the decentralized uh things on the last pod but we're watching it carefully um VC Trends will be huge I think you're going to see the continue rise of solo Venture uh partners coming out with solo funds solo GP will be continue to be explain what that means a Solo solo Venture partner what do you mean by that solo GP is a individual person who basically acts as their own Venture Capital firm they raise between 5 to 20 million and they're either a domain expert subject matter expert or a former entrepreneur or VC from another firm and there's only one partner and there's only one person they make all the decisions there's no big partnership um and the thesis behind that is they can get in the deals early and they go into areas where they see the smoke before the fire so to speak as things start to blow up good because the bigger firms are are replacing the traditional VC series a series B are going to end up doing the bigger rounds now that the private Equity guys are dropping out of that financing so it's just been proven that a lot of these solo General Partners have been very successful Zas is one I know in the valley here there's many more um and also it's very democratized um you can have um anyone become a VC if they just make the right bets again you know My Philosophy I think VC investing is easy no I mean seeing what's going to win that's pretty obvious to us why do you think this is a successful model it's just because they're more focused because because they can get rois better because they know the bets early and they can get in early deals so they can hit the right deals the right set if they have the right amount of capital if they're spraying and paying that's one thing but a solo GP has can go can make decisions very quickly and they're usually trusted and usually have more value add and again they don't they usually stay within their realm of expertise they don't they have out of scope uh factors in other words they take a very narrow scope and they say I'm only going to invest in this area through these colleges I'm going to watch this heard of entrepreneurs come out I'm going to know what a what a winner looks like and I'm going to have the vision to connect the dots that's a winning formula versus a firm of 12 Partners have internal politics that one guy didn't fund the other guy's deal the females AR aren't represented there's no one to represent minorities it's all white guys like it's like it's like the old VC um metaphor you know bunch of guys sitting around the room and they it's slower and they are bigger they got a billion dollars on their Capital so the bigger firms want these Scout funds they used to be called Scout funds now it's like just invest and you start to see citu form again it's it's it's a trend because a big firm has to put all this Capital to work they don't want to spend the time to write a $30,000 check or a million dollar check they want to write a $20 million check or $50 million check and so that's where the economics change so that's a VC Capital Trend and also I think the startups on the AI side right now are the Canary and the coal mine and we're going to we're going to do a lot of coverage with startups you're going to see us doing a lot more Cube videos around AI startups and that's the big area again the next area is AI regulation and productivity we cannot ignore the regulation in fact there's a lot of stuff going on over the holidays around um the the U political article that was going around saying that this whole Silicon Valley culture of pessimists and you know um Dell are infiltrating the politics which is complete BS and first of I thought the article was a great article by Politico I'll put the link in the show notes but um their premise was is that there's a huge Silicon Valley AI contingent in DC putting the Doom and Gloom into the political realm around Ai and it's not optimistic it's not um positive at least in my opinion so I I I don't I didn't I didn't like that article so I think we're going to watch this regulation game because it's game played around policy makers policies are basically lobbyists so if you're in policy you're a lobbyists lobbyists aren't adding value they are lobbyists they don't build anything and usually are on the payroll of someone trying to set an agenda and so when you start getting into a world where you have agenda setting and ausc and politics and lobbying you have a lot of misinformation a lot of Astro turfing going on um and a lot of misinformation campaigns so do you think it's always you think lobbying I mean I'm generally not a fan either because there's a lot of waste that goes on but you think it's always a negative is is there any positive outcomes of lobbying I mean I'm thinking about there is I'm thinking about like in the 80s with the semiconductor industry Association you know going after you know lobbyists trying to compete with Japan when they were dumping you know dams on the market and you know hurting Intel and others and you remember that and so I mean they they do play a role but there's you know but but a lot of times you know they go too far right and it just they have unintended consequences I'll say yeah well there are examples and again um like I always say you rain in the chaos after you let it rain let let chaos rain then rain in the chaos sorry sorry G thought you got say it a lot I actually say it more than you on the Queue but but but but here regulation M I'll give you an example of good policy during the internet days when the US created the internet the policy standard by the government was let's keep it open and so it was an extra effort for the US government and lobbyists around the tech industry at that time Network Solutions um and then became verisign they ran the DNS servers with r power of the internet there was an effort to take those away from the US to make it completely fractured and so there was a set of poly lobbyist lobbyist or poly poly policy gurus that was making an argument or educating people around the benefits of open and that worked until it got too big where you start to have a little bit more you know sovereignty issues and then uh I can became more of a global organization that's some another history point there so policy and lobbying should be about education not managing Direction per se invisible hand a little bit here but not over the top of the hammer so to me bad policies when you're basically you know taking a blunt instrument to lawmaker's head and saying I want to see it this way I'm getting paid I'll pay you quid proquo some you know dirty politics versus education on the on the value and so policy should be open and transparent like it was with the domain name system so again that's one one example other examples you mentioned so you know AI regulation should match productivity and some philosophy whether it's openness or safety in the sense of AI so we're going to watch that so regulation is going to be um anything that's to do with government regulation and policy or um and governance usually there's probably some hassles in there inefficiencies get again this is my personal opinion it's a rant section item but we're going to follow it this year heavily we're going to have probably some DC folks covering it for us um and then the area next area is data architecture and evolution I mentioned data Lakes you're going to be covering Amazon Amazon announced some Innovations and storage you're seeing silicon you got Cloud flare you got a lot of um you know architectures around semantic web you've been talking a lot about this nextg data platform sixth platform you know this is real this real change happening with the compute and and um processing power TPU gpus uh grock has got an inference chip another great company we've been covering so data architecture Evolution heavy access to this one this year tons of content it's going to impact devops Dev SEC Ops everything and then geni startups we'll cover them like the cows come home we then do that no problem and then multimodal models Dave and AI operations I think that's going to be huge you're already starting to see it now ai operations in Cloud not the old AI Ops it's the new AI Ops we're going to get into that that's devops and then devops confidential Computing and then emerging Technologies like assembly web assembly wasm and others is going to happen so those are the areas last two were multimodal models and AI Ops and devops adaptation confidential Computing there's a lot there I I haven't done my predictions post yet I'm going to do it end of the month with uh with Eric Bradley and we're going to look at you know the macro spending like we always do I'm definitely hit on the ne you know six data platform um no question we're going to have some kind of gen AI you know predictions I have a stack I you I'm sure you get them too right you get all these inbounds of predictions hey John you're doing predictions post this year and PR people sending in predictions I literally have a stack like this I mean I have thousands and thousands and thousands of predictions maybe not thousands but I bet you I have a thousand predictions that I it takes me a week to go through them and I get out the highlighter and I don't use a lot of them but I get ideas from them and I give credit to to those that I think are um both thought leading and can be U measured in other words did it come true can you it my My Philosophy John with predictions is somebody some independent should be able to look at your prediction and you know the detail around the prediction and say okay it's binary it it came true or it didn't or at least have some kind of ability to grade it part of it came true part of it didn't so we always strive to to make those predictions you know a little bit harder or to put some data points around them or percentages around them and then have ways whether it's for instance you know IDC data on how fast the market grew although you know or Gartner data whatever use that as The Benchmark uh or even our own surveys you know look look back surveys and so I feel like that's an important thing that's missing in many of the predictions I think the predictions are actually most of the predictions I'd say 95% of them are Trends okay this is a trend so you're noticing a trend and that's good that's useful but predictions should be like this is gonna happen did it happen the Patriots are gonna you know get the first draft pick you know that's my prediction or make or or make the playoffs make the playoffs did it happen or not right yeah are gonna win the predictions are hard I mean to me what I what I try to do is I try to identify like like where you plant your crop this year right okay what's going to happen this year in 2024 what's the fertile ground so to me it's areas of interest to program around as we set the agenda for silicon angle uh and look at our community we look we take all those predictions and we kind of boil them up into saying okay here's the areas that we're going to dig into um because one of the requests this year that we're going to do that we got a lot of feedback on last year Dave as you know is our Cub alumni and our expert network of 18 20,000 plus people have all been kind of chirping hey do more unpacking like supercloud on on camera bring people that are experts in on camera and unpack issues like what's the future of devops and you you want to do those in areas that are be relevant and cool so you know the areas I just laid out were more kind of editorial areas we're going to dig into and they might be a couple layers deep and we said not not much there or it could be a big gusher like if if crypto hits we're going to be there so you do enough surface area coverage dig into it see how it goes but obviously AI is undeniable this is the year that AI starts to happen in a in a way that's going to be the transformative signals emerge and and and those are to me like I said the investment climate what's tracking you're going to see the winners start to emerge quicker quicker this year the production workloads you're going to start to see those emerge pretty quickly you'll so it's G to be the year of observation what popping what's popping what's not happening that's going to be easy to hit the second one is what's the underlying what's the underlying infrastructure change because the infrastructure will set the agenda for the applications that's why I'm hardcore on the data data Pang is huge and it's and it's the infrastructure requirements are changing pretty dramatically um I mean look Microsoft just put a a key on the keyboard okay copilot on the keyboard okay I think that's a pretty interesting uh trend line right there you know like changing a keyboard to put a co-pilot key so that's that that that's not a signal of of relevance well to your theme of AI everywhere I mean that's it's true it's going to be everywhere I I I you know it's a it's it's an interesting time right uh I've said this before salpy reinvent said we I've seen worse times I've seen better times never seen such uncertain times um I I think things are starting to calm down I mean it's going to be interesting to see what happens in the market this year I mean the stock market actually did incredibly well everybody was predicting recession recession recession last year and it never happened and now people are like oh rate Cuts rate Cuts rate cuts and so you know I'm not sure I would be predicting you know rate Cuts I think there's probably going to be less than people expect um and I think the market postco is still very unpredictable and you know and this just happens when you get these transitions we've seen a lot of these wave transitions what happens is the new stuff is not which is getting all the hype and all the valuations it's not big enough not nearly big enough to offset the old and you're seeing this now with AI spending AI experiments and spending definitely you know 30 or 40% of the spending is coming from other places and so you're seeing a compression in other budgets you're not seeing a huge Topline growth in it spend and as a result you know the new stuff's not paying back enough and so you don't you get this weird sort of offset imbalance that creates dissonance and uncertainty uncertainties in the marketplace and that's I think what we're seeing now in a in a in a way that we've seen before but we've never seen the pace of change yeah happen this quickly at least in my experience I predict that this year will be a year of optimism okay I predict that U you'll see more optimism than pessimism that's my prediction um mainly because the the the the scale is already starting to tip a little bit you start to see so so here and I think the developers in open source drive that and also look at the narrative some of the conversations okay so here's the question then so this past year 2023 was a year of pessimism which was not founded I mean the the result was you should have been an optimist from an investor standpoint right do you feel like the the optimism will be rewarded in 2024 in other words I mean you said S&P grew what 20 20% last year it's it's very rare if ever you see back I mean I think it happened in 1999 where we had 98 to 99 or 90 yeah 98 to 99 I think you had backtack 20% S&P 500 growth very very rare do you think that optimism will be rewarded in 2024 I do I think you're gonna see um I think you're gonna see the VCS who are going to recognize where the optimism is on that customer conversion there there's there's optimism enthusiasm in the startup Community you don't need more optimism there I think the VCS are pretty bullish on where they're at there the ones that are in the trenches um I think the optimism has to be more mainstream and again that's that's why the DC thing caught my attention because they're they're sing the wrong narrative um it's not it's not pessimism it should be optimism um and there's an article that was written um I thought a great article around that from a successful entrepreneur um this week the issue on reward is going to come down to um uh valuation increases because the payback on using AI to change either the value of your business or change your business will be one to watch if you're an existing company like ours if you're a startup getting something Mark product Market fits going to be different I think that's where the reward is going to be so I think a lot of people just need just they need to understand where's the line between safe Ai and bad AI because again there's a lot of bad AI just as much as this good AI it's everything's symmetrical right you can whatever's good good over here there's an opposite effect and that that to me is where it is now I tend to look at the other side of the coin people are talking about how bad AI is here I look at on the other side and say where is it good and that's going to be interesting to see I think um and and the and the other thing I'll predict is I think there's going to be a continued regulation um mindset around um the government trying to T take big firms down because I think think um Amazon Facebook noweta U Google Apple they're going to be under scrutiny in fact you know just today New York Times just posted an article that the doj might file an antitrust law suit against Apple okay um around as dominance of the iPhone yeah suit suit dour you were so successful can you imagine can you imagine if Steve JS was alive iPhone you're too good yeah breaks you up you know sounds like Microsoft and you go back into the 90s you know so um I just find that incred Microsoft I mean just unbelievable how successful they are back back to your reward though I mean on our last podcast we talked about the S&P 500 Dave you saw all the Analyst at the end of the year it was a great year you invested in in uh Facebook meta Apple Amazon alphabet yeah the big seven U and JP Morgan I mean S&P 500 was huge by way I think the rich those Rich The Magnificent Seven I think they keep getting richer um despite all the the the potential you know activism against you know government coming after him and so forth but if you're right and I think you are I think I think optimism will be rewarded and I think one of those rewards is going to come in the the form of IPOs I didn't really chime in before but you mentioned data bricks I think data bricks I mean data bricks based on the data the survey did is doing incredibly well anecdotally we knew that we knew that we know they're doing well you interviewed so many customers at the datab bricks data plus AI Summit last year they're doing they're doing you know kicking ass we know that but they're not a public company so they don't have these quarterly catalysts for whether it's news or you know earnings news earnings prints yeah they can make a product announcement but it's yeah it's not as it's compare that to like snowflake every quarter you know you get to dissect it and I know they're on the you 90-day shot clock but there's to me some advantages despite all the compliance concerns of being a public company just from a marketing standpoint and a and an awareness standpoint and I think that confers advantages to those public companies I'd love to see data I'd rather for instance be uipath than automation anywhere you even though uipath got crushed if you know in in the market initially I'd rather I'd rather have a public forum where I can say this is what I can have a say do ratio that people can track look at this is a great point I mean we can next let's unpack this next week but I I think we should put a pin in this because this is exactly the opportunity that I see the data Bri world the uip pass and the companies that were funded or overfunded in the last cycle there's two companies in that last cycle that got overfunded or and that are doing well the data bries of the world and and the ones that didn't make it so it's becoming clear to the capital markets right now very clear which companies have a path forward and which ones don't okay and if the ones that don't have a path forward they didn't get the product Market fit they have limited product differentiation they don't have what it takes they got to know I mean if you're the entrepreneur and you're the founding team or the investor um you got to have a sou searching moment and you got to look inside yourself and say look at do we have what it takes do we have a path forward and a lot of companies are going to either saying I'm going to have to land This Plane softly in another company an AC you hire or Target a UI path or that plane's going to crash run out of gas and fall out of the sky this is what's going to happen and and this is an opportunity for UI Paths of the world these companies going to can pick up a bunch ACU hire so if I'm at uipath or a company like uipath that's got great position I'd be looking at all the white space and saying okay where's my product gaps let's pick up that company let's pick up that company because you can pick up teams right now the startups that don't have a path forward have some people and some tech grab that Tech plug it in so to me that's what I would be doing and then the ones that have a path forward they're going to go public data braks will go public will go public it's very interesting to look at you know the halves and the have knots and the AI washing and all that stuff I mean you mentioned uipath you think that's a company that started an RPA and has moved into you know intelligent Automation and there is a difference by the way RPA is a point tool um you know kind of a desktop Point tool and intelligent automation is a much broader you know agenda but you think about a generative AI it's going to do a lot of the things that RPA was designed to do and that's a lot of the business of uip path and automation anywhere and blue prism and guys like that so so they're going to potentially be negatively impacted there's a two-sided coin there when you look at the spending data and you look at you know you cross AI accounts with uh certain RPA accounts some companies you know hold their own like certainly M Power automate with Microsoft holds its own uipath holds its own actually you know boost a little bit others get depressed when interestingly when you look at I just recently looked at this data snowflake really didn't have an AI strategy prior to gen AI right they went out and right I mean so they kind of you know poooo it in a way from my you know distance observation so they went out and bought Mosaic ML and have begun to integrate that the latest data actually shows they're getting an uplift from their a AI announcements and their their marketing around that whereas of course data bricks has always been you know deep steeped in in ml in AI so so that's G really interesting it's interesting my point being we want to see how companies can respond and and it'll separate the wheat from the chaff in other words hey I I didn't really necessarily have a kick a killer AI strategy I made some Acquisitions I brought in some talent to your point some AC aih hires and that's actually become a Tailwind look at look at what Salesforce is doing on on on on TV with the Matthew mccon uh ads they're like they're like more ads than IBM Watson ever had so yeah it's interesting D mean if your snowflake clearly I mean we know we talked to them I personally had con with sales snowflake people that they were like ah AI we're on it but like you know really we're not there yet I'm sure they had people working on it but it wasn't top of mind now the wave hits a year ago they it's an easy pivot for snowflake they're in the data Cloud already it's not a hard bridge to build for them quickly okay if you're a startup and you're say series B funded C or Beyond you're either have to be in the data bricks category or you're either Walking Dead Or there's a path out of the turmoil you have to find that path so am I the questions that they're going to ask themselves are am I walking dead startup or do we have a path can we find a path and that's going to be hard and the VC investors or the investors might want to have a call option for instance how long do they take before they either shut it down or make a play so this is going to be a very interesting first two quar of the year you're going to see every conversation that starts go down that way am I walking dead or is there a path through and then the board do they have the balls do they have the guts to have the fortitude to say let's make a call now or they want to keep the call option on valuation on a soft Landing or roll the dice for go big or go home crash and burn or clear the runway that's it's gonna be that's gonna that's there's no there's no Market it's decimating on the startups if you an AI startup up and to the right you're out and and and the other factor is what's the denominator in terms of the cost to actually launch a company you certainly saw with the web you know it created opportunities to do things that you couldn't have done previously of clearly the cloud created this Spate of SAS companies that didn't have to go out and buy you know Unix servers and Oracle licenses and and the premise is that AI is going to allow you to instead of having to hire you know 15 engine Engineers you can do the same amount of work with two or three engineers in you know one tenth the time and and I expect that that is actually a a real thing you know we're seeing it with our own development how fast we were able to get you know to actually a working product MVP and then beyond and then actually launch you know a public product so I I suspect this is going to have a huge effect on on on on speed to build a company valuations time to time to Value time to return and it's going to be competitive as hell well it's going to be fun Dave I want to just close out the podcast by just making a reference to an article by Anil Dash who's the um he's a web original I call him B proponent of the open web great guy always got a great perspective he wrote a great article for Rolling Stone um magazine and it's called Uh the internet is about to get weird again okay it was written on December 30th shout out to Anil Dash again I'm a fan of Anil so people some people take shots at him but the guy's to totally Pro open web um I think he's cool I like his views um some of them are a little bit out there but there's nothing wrong with that the title is called the internet is about to get weird again and it says the New Year offers many promises of the online moment we haven't seen in a quarter Century his basic premises is that we become so siloed with the LinkedIn of the world then X going the way they're going then everyone every tribe having their own little Network that he's seeing a Rena a swing back I should say to the idea of the internet the open web the worldwide web 1995 so it's interesting because we've been talking a lot about this on the Pod around how we think the internet web movement is a lot of parallels to the AI side where you have Telecom powered the web which became information super high Highway internet and then the worldwide web sat on top of it this article is worth reading because a lot of the cool stuff that happened in the early days of the web were because it was open so the question is will ai go that same route or are we going to have more of a Apple iPhone android model pick your closed wall Garden is it completely closed or it's going to be open so a lot of the debate in AI right now is about open source versus propriety it's kind of weird this is um pretty thoughtful post here I'm reading it now it's like it's pretty pretty well done um it's and that by the way that was one of the premises for crypto right was to build and to build a a a distributed you know new internet Sil the show Silicon Valley we're gonna build a new internet with my compression algorithm and then there was another article by Brendan McCord from the cosmo Institute it's called um pessimism versus accelerationism okay um this is interesting long read but it's a really about the the the cultural change I was talking about earlier and I think you know I think we're going to have a revolution in this generation of the internet and AI specifically and that the younger Generations coming online now who are building and setting the agenda and the standards and the deao standards um are going to have a little bit different view I think it's going to be an interesting um thing so it's it's going to be interesting to see as we have the diversity of participation from us old guys you know multiple cycles of innovation um and then the spectrum of ideation am I a pessimist am I an accelerist acceleration person a D cell an aell e cell we you call it it doesn't matter the startups will Define it so it's gonna be very interesting to see how the next wave um of optimism or pessimism drives it so I thought both those articles were interesting and and they both have different cultural impacts and Neil&

2024-01-11 02:42

Other news