My Crypto AI Portfolio: What I'm Buying Now

[00:01] if this is the exact crypto bottom, especially with everything going on in we're heading into an AI super bubble over the next couple years, and I would rather build my crypto AI portfolio now than before the crowd shows up and

[00:15] everything's up 5x higher. So, in this video, I'm going to talk about what I'm buying right now, why I'm buying it, kind of how I'm sizing everything up, one of these bets. A spoiler alert, there's a lot more things I'm watching

[00:28] this exact moment. There's a couple things that I've gone into with some conviction right now, but it does feel kind of wise to to take it slow. I I we have some time. I don't have to rush into all these things all at once. I

[00:42] going to settle over the next couple months, even if this is the bottom. I to the races anytime soon. And so, I'm kind of nibbling at things slowly, although there is one thing that I really went into heavy and hardcore, and

[00:57] do want to clear something up a bit because I do get sometimes a little bit know people in the comments are stupid. They're like universally just the worst. coming in saying, "Ah, your calls are stupid. Whatever you're talking about,

[01:10] I'm I'm going to short." Or like stupid stuff like that. And I've tried to talk them that you both buy and sell an asset, and therefore, you know, the opportunity exists in when you choose to

[01:22] people have a hard time, I guess, wrapping their mind around that. And difference between how I think of the market and how a lot of other people people think of all these tokens like Bitcoin. Like if I talk about like a,

[01:37] you know, $2 million micro cap token, they're thinking, "Ah, this is the next for the next 10 years." But I talk about all the time that those are like the max riskiest possible bets that you can make. And you should expect like two out

[01:51] of 10 of those kind of bets you make to end up working out. And the other eight out of 10, you should expect to go to zero. They're like literally max risk. And in general, when you're in crypto, when you're investing, 99.9999%

[02:08] anywhere. This is the same as like early internet startups. This is the same as Like most of these things do not succeed. Most small businesses don't thinking you're just going to like buy something and hold it for forever. You

[02:21] have to be actually selling these things at some point. At some point you have to take profit. That That is how you make money. And how much money you make is dependent on when you get in and when you decide to take

[02:33] profit. And when I kind of assess myself, I break it down in two different thing's just an absolute dud? It ends up going down 80% and it's just the worst. Or when I invest in a token, does it end up going up at some point? Does it 2x?

[02:48] Does it 3x? Does it 10x? Does it 50x? And then I measure it based on that. If I have a token that I invest in and it does a 100x, was great alpha that I surfaced. I found a real opportunity. And how much money I make is dependent

[03:00] on the strategy I use around that opportunity. When do I sell? When do I initial? So I separate it basically into two buckets. Real opportunity. It was this an actual real opportunity that I had a chance to make money on. And then

[03:13] how do I handle that opportunity? How do I actually, you know, take profits eventually and make money? Because that's the goal here. And because I post my buy and sells in the OC, I can actually go back and track every single

[03:25] thing that I got into, when I got into it, and then measure the total alpha, the total opportunity had I sold that thing at peak. And obviously, you know, peak. I'm not anywhere close to that good. I wish I

[03:37] was that good. But the point is was there real alpha there? Was I picking that were just going down? And it did have some dumb things that I got into that literally just were losses. You can see those here. They really just didn't

[03:51] But then I had other things that ended up doing a 20x or ended up doing a 30x because there wasn't the liquidity to support in profits, you know, full profits on this. But this was a 500x or a 100x. And

[04:07] you can see out of 70 different plays, 60 of them achieved a 2x or more. So 85% of them, 65% of them achieved a 4x or more, 27% a 10x or more, 11% a 20x or

[04:19] more, 27% a 10x or more, 11% a 20x or more, 5% a 50x or more, and 2%. Just 2% achieved a 100x or more. And on average, each asset peaked around 9 months after about it, talking about it, etc. So it's not like just they're peaking right

[04:32] like that. On average, it was around 9 months. Some of them took around 18 point is the vast majority of them were good alpha. But to be honest, I have not done the best job this cycle on taking the profits at the right time. And

[04:47] that's really where I'm focusing on improving my strategy. Last cycle I did taking my initial and I'd take profits along the way. I didn't do that as well switching back to that. And I'll be focused on taking profits more

[05:01] aggressively to make sure that my strategy is working for me. And if I were to brutally honest grade myself, I would say in 2021, I crushed it. I did a really good job on both good alpha and good things that were going to go up and

[05:14] actually taking profits along the way. On this previous cycle, I would say I finding that they were going to go up. I did really bad on surfacing when to sell those things. That was one of my biggest

[05:27] flaws and one of my biggest mistakes this past cycle was identifying when to my disclaimer before I get into these tokens, so you kind of know my strengths and my weaknesses. The first token that I am holding in my AI portfolio is a

[05:41] smaller VC type bet. And when I say VC type bet, what I mean is this is like max risk, okay? It's a micro cap token. A micro cap token is like as risky as you can physically possible get in the crypto space. The first one is one that

[05:56] I've been talking about for a long time. That's Avo. I do still hold my Avo and I have actually really clean shop over the bear market. I sold my virus. I sold my just really weren't doing well. And I was really particular about the ones

[06:10] investment or a new bet. Like would I buy this still today? Yes or no. And if ended up keeping Avo because I still think this is a really killer narrative and I still think this is a really killer project, especially in the AI

[06:25] They're doing it on Solana. Solana has really just taken the cake in terms of chains and just where all the attention is. They're focusing a lot on AI agents are really popular right now. And it's not like the AI agents on virtuals, but

[06:40] trading for you and different things like that. Uh I believe this will become bigger as it heats up this cycle. You're going to see a lot more attention on kind of like Claude and Codex and different things like that trading for

[06:54] those things into your trading tools. And so I think there's a lot of this, especially because they have, you know, grinded all throughout the bear market. So this is one that I'm holding in my AI VC bucket. The second one I'm

[07:07] holding in my AI VC bucket is another one that I've been talking about for a long time. You will notice I'm very consistent about the tokens I invest into. I don't like to jump around to like, "Ooh, this is popular. This is

[07:19] popular." and all that kind of stuff. I invest in things and I talk about things I actually put my own money in. And even though I recognize that the bear market's a great excuse to start over and start clean and I could just, you

[07:31] kind of stuff. I still believe in this token and this is still one that I am planning on holding again in that VC bucket. And that's Infra. And this one one's never like the entire existence of

[07:44] this one has been in the bear market. It launched in the bear market. It went up, bear market, and then it's come back down again in the bear market. I do talk extremely talented. It's a just super, super small token. And I think when the

[07:57] market comes back, when there's more momentum, this one really could shine bear market. They're working on something brand new. I'm going to reach this in the video. So, if you can see it right now, they said that was fine. This

[08:09] trillion dollars or a hundred billion dollars. It's just one I think it could cap, which is like four hundred thousand dollars in market cap. And finally, the biggest bet that I'm making right now,

[08:21] That's People are going to hate on me so much for this. I sold my Bitcoin. I sold my ETH. And I took a big chunk of cash and I invested it into the Akash Network, which is one that I've been talking about since 2022. But it This is

[08:33] one of the projects that I believe in just ridiculous amount, especially as I see a coming AI bubble. I think there's a really massive market for legitimate AI crypto projects. Uh when I say legitimate, like the VC kind of bets,

[08:46] those are good projects. But those aren't the ones that I see getting like, have to be more established. You have to don't think you could even come to the market with a brand new project and get

[08:58] too much institutional interest, uh unless you're doing crazy numbers in of Akash, has been working on what he's building since 2015. They're not going anywhere anytime soon. They're really well established. They've been building

[09:11] for a long time, 11 years. And especially as open source is rapidly catching up to some of the top models on the planet right now. GLM 5.2 beats Fable out on certain design benchmarks and different things like that. A lot of

[09:24] people kind of overhype this, but it it really is like very specific design uh still, it is kind of foreshadowing what's to come, which is a year from now, uh probably sooner than that, uh open source will be as good or better

[09:37] than Fable uh Mythos level intelligence. And whenever we do hit kind of that hard gains where it's a lot harder to make are going to be equivalent to a lot of these higher-end frontier models. Which

[09:50] means a lot of the value capture might not even be at the model level, but might be at the distribution level, and it might be at the compute level. Which Akash has distribution with their inference service built on Akash via

[10:04] Akash ML. And then they have compute via their open permissionless cloud, which basically think of it like AWS but open and permissionless and decentralized. As well as they have compute via Akash home node, which is like where you can hook

[10:17] up your gaming rig. So if you have like a gaming laptop, gaming desktop, etc., you can hook that up and that could be used for things like inference. As in obviously for training, you know, the GPUs you need are typically H100s,

[10:29] H200s, Blackwills, different things like that. So this isn't going to work for training, but for things like inference you can use much smaller GPUs much lower quality or lower grade GPUs to process inference. And inference is really the

[10:42] my single biggest position in all of crypto. It's currently sitting at 220 million in market cap. It doesn't have a really high FDV like most tokens. Like I And I wrote a really in-depth article on X where I go into exactly why I believe

[10:58] it is one of the most underrated tokens in all of crypto. It has to do with the fact that I believe permissionless compute is actually an essential thing I think when especially with you see right now, the government clamping down

[11:10] on Fable access for like foreign entities with Anthropic. But as the more and more, they're not really going to be able to. They get shut down you know, then Chinese open-source models are just going to release the

[11:24] same thing and everyone's still going to have access to it. There's literally no right? It's just going to keep coming from somewhere. So, they can't really kind of belief is that the next thing they're going to do is they're going to

[11:38] know, wherever you're getting your inference from. And so, I believe really the only thing they can't clamp down on is compute. And so, in that world, a decentralized permissionless compute network is amazing. It's like Bitcoin,

[11:51] it. Anyone can tap into it, anyone can get access to compute. That's a really easy way to frame it. You know, we're starting to see why you might care about future, getting access to these advanced models is going to be the difference

[12:04] between making it or not. But, it also gives you other things where, you know, with Fable access getting shut down, it did kind of show a lot of these Fable, they rolled out Fable in a way that it's like taking all your data into

[12:18] you know, enterprise people don't want that. They don't want their data going to Anthropic. And so, they're trying to build their own in-house models. But, as imagine you have like, you know, instead of a thousand

[12:32] agents that are running your business for you. And imagine that goes down. going down. You know, it's not like one person gets sick day, that's the end of the world. But, if your entire agent force goes offline because AWS flags you

[12:45] you know, whatever it is, that's pretty also my belief that you're going to see a lot of these people want some sort of backup, some sort of thing that can't get turned off, credible fallback

[12:57] infrastructure. And that's another use case that I see Akash foreign AI companies, like, are they going to want to use AWS or like US-based cloud compute? Because, you know, the US government can command AWS

[13:11] or, you know, Google or whatever to cut off access, just like they did with other points, including like home node and why being able to tap into the GPU consumer market is such a big deal. So, currently, there's estimates that range

[13:26] from about 200 to 400 million gaming PCs on the market. There's other ways to tap compute. This is just kind of like their starting thing is gaming PCs. But let's say you tapped into 1/4 those, which is arguably actually pretty large. I don't

[13:40] this is just kind of paint the picture of like, you know, how high things could go. If you were to able to tap into 100 million of those, just 100 million gaming PCs would exceed all available compute from current AI data centers

[13:52] today. So, um that kind of gives you a picture about how much consumer computes one-for-one. It's not like consumer computer is the same thing as a Blackwell, you know, a GPU or different things like that. But for inference,

[14:04] it could fulfill the same need. Like that that's the thing we need it for. We need it for inference. And so there's a really strong case for the fact that consumer GPUs and consumer products could fill that need. Anyways, outside

[14:16] of that, there's just a lot of different opportunities in the crypto AI space, even right now during the bear market. This is one that I didn't personally get into. Uh but we have this thing called Nightfall and uh the Degen chat in the

[14:29] That actually there's a couple guys in there that really really crush it. Like they're just so stupid good at finding small picks and getting the in them early and they're doing like 10Xs, 50Xs, 100Xs. It's actually kind of absurd how

[14:43] ridiculously good they are. But one of those guys named Tripwire got this one at around 300K market cap. So when it was just absolutely tiny. And ended up posting about it. And this thing ended up running up to I think it was around

[14:55] $13 million in market cap. Which was a nice 43X in a couple of days. And it got a ton of attention because the founder, Leighton, ended up doing decentralized like inference on multiple different GPUs. So there were

[15:09] different GPUs scattered different states. Normally this process is pretty pretty quick. Um this is actually really really quick. So it's basically at the speed that you're used to uh doing inference on ChatGPT or different things

[15:22] that. Maybe a little bit slower, uh but really fast for scattered GPUs. And they Virtuous. They ended up getting tweeted out by Nvidia. All kinds of different things. Which kind of crazy for a really tiny crypto AI token. Again, this isn't

[15:36] one that I currently hold or am currently looking to buy at the moment. of the bear market, there's abundant opportunity in the crypto AI space. And there's actually some interesting things being built in the AI crypto space. And

[15:49] this is AI crypto tokens. I still hold tokens like RSC. I still hold tokens talking about the AI ones I hold. And I'm just really particular on on the ones I pick cuz I like to find things that I I can kind of uh stand behind

[16:03] the comments below. And if you're curious about seeing my entire portfolio or you want to see every time I buy and sell various tokens, as well as uh currently the Obsidian Council is closed to new members, but you can sign

[16:15] of this video. As always, none of this is investment advice. None of this is me money. I'm obviously not your financial advisor, and you should always do your if you want to see more videos like this, make sure you hit that subscribe

[16:27] be notified each time I release a new video. Thanks for watching, and I'll see video. Thanks for watching, and I'll see you next week.