GM future billionaires, we are finally back with the third installment of the blocmates year ahead thesis. First and foremost, let’s take a quick look at the 2024 calls that were made and decide if I am full of shit or not.
- Experimentation with the EVM: AlternateVMs, parallelization, and more ✅
- Lots of experimentation with moving away from the EVM. Move and SVM have been dominant mainly due to Solana’s huge success.
- Move over Solidity ✅
- Luckily, this has fallen into place here, as the MOVE token literally dropped a few weeks ago. Sui has done incredibly well and Aptos is well… Aptos.
- Modular chains & data availability: Frankenstein chains or modular Megazords? ✅
- This was a biggie for more reasons than one. Celestia is becoming the main competitor for Ethereum. It’s kinda wild that I expect this to continue in 2025.
- Restaking will be HUGE: Eigen Layer, AVSs, and LRTs ❌
- Nothing as of yet has taken hold. AVS demand is not there. I struggle to see where this goes. If they do manage to pull it around, it can be one of the biggest things the industry has seen.
- Cross-chain restaking: Composable Finance/Picasso Network ❌
- Max pain.
- The winner stakes it all and DVT
- Kinda worked, kinda didn’t. Ethereum has been flat as a fart, so this hasn’t really stood out as impressive, unfortunately.
- Decentralized Physical Infrastructure (DePIN) ✅
- We had a huge year for DePIN. I expect this to continue!
- Bitcoin expansion through experimentation: EVM, BRCs, and more ✅
- Runes and Ordinals had their day. The private markets have allocated heavily into Bitcoin DeFi. I expect explosive growth in 2025.
- Intents for everything ✅
- This has played out, but I don’t think we have seen anything yet.
- RollApps?! ♻️
- Rolling this one over as I think based rollups and shared sequencers like Initia will be the talk of 2025.
- IBC and Cosmos ❌
- Some have done well, to be honest, but more of the same. Cosmos needs to get out of its own way.
- The dYdX playbook, coming to a perps DEX near you ✅✅
- HYPPPPPPPEEEEEEE
- RWAs: Yes and no
- Kinda, but not really. No demand when you can go to Pump.fun. That said, the BlackRock RWA fund is interesting.
- Crypto and AI ✅
- What can I say… I never expected it to get this crazy, but now I think we can go to Valhalla.
- Memecoins are NFTs without pictures ✅✅✅
- Best call of the whole thesis. Nailed it even before the pump. Nobody agreed. HFSP!
- Solana fee-model revamp ✅
- Solana is profitable…
- The year of the Bera ♻️
- Fake chain until 2024 Q5.
- Crypto payments are THE killer app
- Kinda, but no settling on which platform.
Final points:
- “We will probably see more business licenses for DeFi’s most innovative products. Following on from Uni V3 and V4 business licenses, I think this trend continues for better or worse.” ❌
- Thankfully I am wrong about this one. AI x crypto is thriving in open source.
- “KYC DeFi is certainly coming. Whether we like it or not, the rails provide a much more seamless experience, whether it is public or private blockchains.” ❌
- Regulatory clarity saved our asses.
- “Token V2 launches - Immutability? Nah. I think we see a lot of token migrations, particularly from OG protocols to further increase the supply and utility of their tokens.” ✅
- Lots of new tokens/V2s — this will continue in 2025 too. Virtuals, Beam, GuildFi, Fantom, FET/OCEAN, to name a few…
- “Crypto’s first breakout game, I would imagine, is Off The Grid. This is just a cut above the rest, in my opinion. Avalanche had set themselves up to be THE destination for deploying games with subnets. If one takes off, then it’s all systems go.” ✅✅✅
- Really happy with this call.
- “Airdrops, LBPs, and stealth launches on Uniswap are the only way to launch a token in a fair way. I think, given the XRP case which ruled in favor of free market prices not being a security sale, these methods are the way to get a token out in the open.” ✅✅
- True, but now we have regulatory clarity and also have Echo :))
Overall, the 2024 thesis performed very well, with our memecoins call being the star of the show.
The year ahead
I’ll start by saying that it is a lot more difficult to picture where the industry is headed in 2025. Coming off the back of a brutal crypto winter, 2024 was relatively easy to call.
This time, though, the arena has changed dramatically. Breakthroughs are happening constantly, and it will be hard to keep up.
There aren’t too many contrarian picks, unlike the Solana bet from last year. So I’ve tried to highlight some of the areas that will finally get their day in the sun and the projects that are best placed to capitalize on the chaos.
Crypto x AI will continue to be the main narrative in 2025
AI stole the show in 2024, nothing could get close to it, particularly towards the back end of the year when agents started sprouting up all over the shop.
The Kaito narrative heatmap over the past 12 months is the live and breathing exhibit of this.
Here is what the narrative trend looked like across the whole year. As you can see, AI, memes, DeFi, and ETFs were the key talking points.
Despite being ahead in mindshare for the majority of the year, Truth Terminal and GOAT were the catalysts that changed everything.
There was something so special about the launch, the A16Z lore, the fart jokes, and Andy. It made everyone pay attention and inspired builders, investors, and hobbyists to start taking the crypto x AI intersection seriously.
Maybe we were always going to get there, and Truth Terminal was the activation energy needed to get us over the line.
Since then, we have expanded drastically.
At the time of writing (December 21, 2024), cookie.fun is currently tracking over 447 AI agents, which collectively total $9.43 billion. I recently wrote a short piece comparing this AI agent category's birth to what we saw in DeFi summer but on steroids.
NOTE: This has currently changed and sits at a $12.33 billion market cap and 1373 agents created.
The TLDR is that the DeFi total market cap topped out at around $200 billion, and this did not include the networks (L1s/L2s), etc., that they were built on.
If I were to guess, we’d flip that figure and then some in 2025. The infrastructure and agents resemble networks, and their apps (to a degree) resemble the agents — this is exactly how I believe they will be valued at some point in the near future.
AI agent framework in 2025
Virtuals Protocol made a late run to be the contender for the protocol of the year in 2024 with its AI agents launchpad. After a hard pivot from Path DAO, the team created the most user-friendly no/low code environment enabling anyone to create, launch, and train their own AI agents with a token.
Stars of the show include AIXBT, an agent that scraps multiple sources of real-time crypto data, news, and info to provide a very impressive alpha account on X [formerly Twitter]. AIXBT
People have begun copy-trading AIXBT's calls as they are so on the money. This is obviously extremely reflexive, NFA.
Another key component of the Virtuals ecosystem is its framework: G.A.M.E.
GAME is currently sitting at a $73 million market cap and serves as the Virtuals framework to enable AI agents to interact autonomously using the plug-in and play APIs and SDK from Virtuals.
I believe this is drastically undervalued in comparison to everything in this ecosystem and will probably re-rate viciously higher.
LUNA is the first agent that really came onto the scene with the Virtuals launch. It is an “influencer” that does 24/7 live streams, calls her followers kittens, and has long-term memory.
So, if you simp for her, odds are she’ll likely remember you going down the line. In fact, she can also tip fans from her own wallet.
This is an example of dogfooding your own product from the Virtuals team to show the proof of concept. I think the influencer-style agents will probably fade away.
It’s like a magic trick or a joke you have seen/heard a few times. It loses the shock factor quite quickly after a while. That being said, I still think it can do well.
I think the upside in the near term is in the enabling tech i.e. frameworks… I think this is the case until you get breakthrough agents — AIXBT is showing what that looks like already.
Once you get an agent that is doing tasks for millions of people and being paid, becoming self-sufficient and is already co-owned by its token holders, that is when things get crazy.
I believe on-chain data integration for AIXBT is around the corner. Factoring in its social metrics and information aggregation, this could push it into a different league altogether.
As I mentioned above, I am picturing the valuations and comps of these specific assets to reflect the same dynamics seen with L1s and their apps.
Virtuals have some great ponzinomics (in a good way) that drive a lot of demand for the token and their agent tokens, too. I like how well-thought-out this has been. Usually, when there are a few additional steps with a token, the barrier to entry can be off-putting. Now that they have hit escape velocity, the token dynamics really kick into gear.
And, as if by magic, as I am proofreading this thesis, the Virtuals team has announced that the fees generated from the Virtuals protocol will now go towards a buy-back-and-burn mechanism towards the respective tokens that generated the fees.
The way this space is moving, I will never finish this thing.
Once again I am set to post, and they have announced that they are deploying on Solana. FML, I hate it here.
ElizaOS
ElizaOS is an open-source multi-agent framework written in TypeScript. Shaw, the founder of Eliza Labs, had a polarizing breakthrough in the crypto scene a few months back, but in my opinion, all this did was grant a generational opportunity to pick up the native token of the Eliza framework and ecosystem, ai16z.
ai16z is now the confirmed token that will support and fuel the ecosystem. And what an ecosystem it is shaping up to be…
Eliza's adoption is unprecedented!
Yes, you’re reading that right — the amount of Eliza GitHub stars has surpassed the most used Uniswap repo in just a matter of weeks…
It has been the number one trending repository on all of GitHub over the past month. It’s wild to see this level of adoption for something that is inherently tied to the scam, that is, crypto.
I can’t argue with this data, and the rate at which Shaw and all the contributors are shipping is incredible.
Hosting their weekly “What did you get done this week?” to recap the developments in the ecosystem and holding zero to hero—AI Agent Dev School, I cannot bet against that. It’s inspiring to see people so dedicated to open source, and it gives me hope.
On top of that, there’s chatter about ai16z tokenomics being changed to help capture value.
Strangely enough, this was published just hours after our Head of Research, 563, shared insights on how value accrual could be driven back to the ai16z token. Wen royalties?!
Another is arcdotfun, from the Playgrounds team. They’ve built an AI agent framework on Rust (which is obviously the programming language used on Solana) called RIG.
Like Eliza, RIG is gaining traction from the developer community. At this point, it is a race to get these code bases into the hands of as many developers as possible and see which ideas can take hold.
Time to pretend we know our way around GitHub for the foreseeable future.
Swarms
Swarms will most likely become a thing in 2025, too. Swarms enable agent-agent communication across different languages, code bases, and models in an asynchronous manner.
Take the above example of ZerePy and Eliza. Agents built on these frameworks can have open communication and create swarms of agents, all chipping in together to resolve a common task.
Gaz did a guide to swarms back in December, too, if you want a pre-cursor.
Imagine you were assembling the on-chain agent Avengers and you need them to all do what they say they are going to do (reputation) and be able to communicate with each other to make sure the task at hand is done with maximum efficiency. That is what teams like FXN are building.
In the same way that humans have evolved and specialized their skill sets over time, so will agents. AIXBT is never going to headline Coachella, Zerebro might… while LUNA is never going to manage an AI hedge fund as good as ai16z or DegenSpartanAI.
Banding these specialized agents together to collaborate on a shared task is the future, and the closer we can get them to do it in a truly automated and agentic way, the better.
That brings me to my next problem to be solved in 2025… reducing the human input/curation of these agents, effectively getting ever closer to purely agentic agents.
Humans are still curating the output of these agents to ensure hallucinations don’t get posted to their forum or platform of choice or inaccurate outputs don’t get actioned. This would (and has) kinda ruined the allure of certain agents at times.
This is a big problem that many teams are currently trying to solve, not just in the crypto industry but across multiple AI-related sectors.
I recently came across a team building to solve this very problem and reduce the human input into agents. Mira Network has shown in its early data that it can increase the output precision from 73.1% to 95.6% in its 3-of-3 consensus network.
Just like we use a collection of nodes to validate a traditional blockchain network, we can also validate the outputs and reputation of an AI agent and its underlying models.
This increases the reliability of agents while the team can focus less on pruning the output and more on training the agent better with a broader dataset to further optimize its results.
I’m super excited to see these guys' integrations in 2025, with an expected release date in and around Q1.
I hate the word, but DeFAI. (Can we just call it AI?)
I have a whole section on wallets below, but I think wallet UX is about to change in ways we’ve never seen before.
I don’t know about you, but my online search habits have totally changed since ChatGPT and Perplexity. I have slowly shifted away from Google and have begun solely using prompts on these interfaces to get the desired results.
I think wallets are headed there, too. Griffain looks like your typical AI chat interface, but you can direct it to do on-chain actions. They have a cool tagline that is “make crypto as simple as typing.”
There are a few of these teams in the market building this kind of product, and I think they all have a very strong claim to the throne.
Projects include:
We also recently released a full beginner guide to DeFAI if anyone is interested. Plus Tony from Griffain joined us on the podcast recently to talk through the product and what we can expect in the near future.
I wonder if established wallets will start rolling out this interface style, too. I guarantee they are already in R&D for this kind of thing; they naturally already have the distribution, so this could be an interesting space to pay attention to.
AI-adjacent bets for 2025
Nillion
I am expecting a big year from Nillion. The idea of data privacy in this day and age cannot be understated, and the Blind Computer’s architecture is a thing of beauty.
We wrote a full explainer article on Nillion which you can find below, that gives you a complete breakdown of why the entire blocmates team is super bullish on this idea.
Privacy is going to become a huge issue, particularly with the power of these LLMs and the concentration of the most powerful ones in the hands of a few tech titan monopolies, which I don’t trust as far as I could throw them.
This can also catch a bid on the decentralized compute, AI, DePIN, and even the privacy narratives which will almost certainly be a thing in 2025.
Their last round was led by Hack and included Animoca, with previous rounds backed by the likes of HashKey and Maelstrom, plus angel checks received from Ansem, DCF, and Jordi Alexander.
Story protocol
Story is launching an EVM-compatible network that focuses on changing how intellectual property is managed, stored, licensed, and used across multiple industries. Their stance is that everything you own or do is IP.
Led by two extremely bright and talented young founders, Seung Yoon “SY” Lee, who sold his last company Radish to Kakao for $40 million, and Jason Zhao (25), who was the youngest employee at Google’s DeepMind before leaving to build Story.
The weird world we are moving into with AI and how your data and personal information is used is going to get very funky very fast.
The message from the Story team here is, “Don’t fuck with my data.”
We are currently being asked to use this opt-in message to decide on the blocmates YouTube account.
Third-party companies are now looking for larger and more tailored pieces of content to train their models on. In this example, we basically have no idea who these people are or what our content/IP will be used for.
On top of that, Alphabet (the parent company of Google and YouTube) will almost certainly be charging a nice fee to enable this transaction and leave us with crumbs.
Story would change this transaction completely and give the power back to our own content and IP.
We can list our content on a marketplace on Story Protocol, specify what we would allow it to be used for, and quote our price. Then, the terms of the agreement can all be signed on-chain.
TCP/IP
On top of that, the Story team recently released its TCP/IP paper which aims to standardize agent-agent communications. Sound familiar?
What happens when agents start creating their own music, art, or basically anything, and they want to license it out? Well, Story Protocol will play a role in that too.
For more information, you can read the paper here.
Story Protocol has raised over $134.3 million over three rounds, all led by A16Z. Take from that what you will.
Bittensor
We had to give our Head of Research, 563, his own podcast as a release valve for his bullishness on the Bittensor ecosystem. It’s called TAO Talk for those who have not yet subscribed.
Here’s what he had to say on the year ahead for Bittensor.
“With the benefit of hindsight, I think Bittensor’s 2024 will be looked at as setting the stage for a Cambrian explosion of the network. While it was no stranger to drama and missteps this year, the project has positioned itself to be the AI coordination/incentivization layer (the definitive ‘AI layer-1’).”
The best agents will be powered by the best intelligence network (Bittensor subnets).
This ecosystem map is only a month old and is already severely out of date. That’s how fast the ecosystem moves. Also, check out TAO Pill for a list of subnets.
In my opinion, the updates coming in 2025 will dissolve the barriers that have been holding the ecosystem back:
Dynamic TAO:
- Weight-copying by validators has been long discussed and well understood. The introduction of Dynamic TAO will introduce a market-driven mechanism for emission allocation, meaning that all actors (investors, subnet owners, miners, and validators) will now have “skin in the game.”
- Speculation, as we have seen and discussed at length in previous reports, is the use case that drives market attention. By letting users ‘bet’ on subnets, a whole new class of participants will flock to the network.
Subnet slots increasing:
- Platforms such as Pump.fun (and the countless forks it has inspired) have shown us that crypto investors love speculating on new tokens. Bittensor’s current subnet registration cost of ~$1 million establishes a huge barrier to entry for scrappy projects looking to shake up the game. By dialing up the slot count (via chain improvements in 2025), the hurdle will be lowered and the games can finally begin.
The Bittensor ecosystem has grown to a size that is hard to keep track of. To stay up to date, be sure to tune in to our weekly TAO Talks (Twitter, YouTube, Spotify).
Trusted Execution Environments (Enclaves): TEEs
I recently posted this tweet stating that TEEs are this cycle's smart contracts. To expand on this point, I don’t mean they are smart contracts. I mean they will have as much impact on the industry as smart contracts had and will begin to be adopted in this space.
Amazingly underrated thread on TEEs above.
And Ollie published his own piece on TEEs for blocmates.com.
Many teams are working on different use cases for TEEs in AI to prevent tampering, increase security, or offload compute for dApps, and I believe they will see a large uptick in 2025.
Teams working on TEEs include Nous Research, Marlin Protocol, Automata, and the Phala Network.
Here are a few special mentions for the year ahead in the AI sector, which we believe will have a strong 2025. Of course, we’ll be covering ‘em in more detail over the course of the year.
- Mode - Doubling down on AI agents prior to the boom. Plus working very heavily with multiple subnets in the Bittensor ecosystem.
- EternalAI - Drag and drop toolkit for the no-code deployment of agents with or without a token on any-chain.
- Nous Research - These guys are the kings of the AI sector in my opinion and don’t get the recognition they deserve (probably because they don’t have a token). They also have the best branding I have ever seen. Shout out to John Galt — your work is incredible! I will be paying attention to whatever they push out over 2025 and beyond.
- Almanak - Have just concluded their community raise on Legion which was initially supposed to be for $500k, but they upped it to $1 million to try to include people based on the demand and they were still 15x oversubscribed. Keep tabs on these guys who have been building towards the agentic future for a number of years now, no bandwagons here.
- Prime Intellect - Recently pushed a decentralized model trained on 10 billion parameters which they made open source (chads). Also, very cool branding…
- Lit Protocol - Agents and apps build on Lit’s key management network to confidentially orchestrate identities, platforms, and crypto networks.
- Gaia - Decentralizing AI inferencing.
- Gen Layer - Building the AI-focussed L1 with their Intelligent chain.
zkTLS
Adjacent to TEEs we have zero-knowledge Transport Layer Security (zkTLS). In a nutshell, zkTLS allows users to share specific pieces of information (data) with their counterparty in a trustless and secure way that doesn’t reveal any more information than what is required.
The most left-bracket way I can explain it is this: Think about when you are asked to provide a bank statement or a bill to show proof of address. That is overkill, right?
So, when we bring this into the on-chain/AI world, we open up a whole new design space that can potentially disrupt web2 incumbents, which rely on the additional excess of data that you provide them to be profitable.
AI inferences (outputs) can be verified and tamperproofed using a mixture of TEE and zkTLS.
Some of the building teams in this sector are Sophon, Reclaim Protocol, TLSNotary, and Opacity.
Anyway, I can’t stay on AI for much longer as it is four days before Christmas, and I still haven’t bought my wife anything.
Wallets have won… Well, Phantom has won!
Wallets, particularly Phantom, have basically won. By that, I mean their distribution levels are nothing short of a monopoly (in a good way).
This has been the cycle of mobile-first crypto onboarding and, more specifically, on-chain onboarding.
Phantom enabled a seamless onboarding experience for millions globally when memecoins started to rip. Phantom trending is now a meta that people pay attention to, and with the additional explore tabs introducing new DeFi applications to the masses, I think they are going to be the go-to crypto onboarding platform throughout 2025.
They have recently enabled social sign-in (bye-bye seed phrases for your nan) and also user names, which bring everything back to the real world with modern online banking apps, etc.
As I mentioned above, if they push towards a chat-like interface for certain on-chain commands, that could be very interesting. Another aspect is that they could quite literally create the in-app crypto social experience that everyone has been waiting for.
In-wallet chat is coming, whether it is 2025 or beyond, I don’t know but wallets like Phantom will have a wallet-to-wallet chat function one day, mark my words.
If Phantom can build on the back of the work being done with account abstraction and chain abstraction and become more like a portal, I believe they could dominate the entire industry.
I do think Infinex could also begin to gain some traction with its incredible product and team that knows how to ship (and win). I’d definitely keep an eye on what is happening over there as I think their chain-agnostic fundraising product is potentially game-changing for the industry.
Something else to watch out for on the EVM side with wallets is EIP7702, one of the most powerful account abstraction EIPs to date. It can effectively turn any EOA into a smart contract wallet, allowing social login and session keys to enable subscriptions and even gas sponsorship.
Here’s our good friend Henri from Privy explaining it below:
https://x.com/privy_io/status/1867249205133373905
P.S. Phantom, drop a token, please…
Expanding on the bonding curve and prediction market experiments
One morning in the middle of summer, when I was trawling through Pump.fun and saw a flurry of Liam Payne tokens, it dawned on me that this might be the quickest news source on the planet due to its heavy incentives for tokens to bond.
Of course, I am half-joking here, but that situation really made me realize that bonding curves and/or prediction markets have more unique use cases to come. In 2025, I believe we will see lots of interesting use cases that expand on prediction markets and bonding curves.
There are a few lined up to launch that take a different spin from Polymarket or Pump.fun:
True Markets - News powered by markets
Upside - Turn any tweet into a viral token (Disclaimer: We have invested in Upside)
Myriad Markets - The interconnected media ecosystem that integrates on-chain prediction markets and quests alongside written and video content
Dev.fun - Turn your ideas into reality with one click and make them instantly tradable on Pumpfun.
Clout.me - BitClout reimagined
To name a few…
I’ve had to rewrite this thesis three times since starting in December. This space moves incredibly fast and will continue to do so with the daily unlocks that AI is granting us. Trying to keep up with everything on your own is a nearly impossible task, you need others around you that are like-minded and share the same goals.
This is why we created the most reliable research platform in the industry that is run by people who are full-time actively trading, investing, and building in the trenches. They’re not writers or creators by trade. They are the early adopters, the ones trying to shit out before anyone else, getting their fingers burnt, and also funding some life-changing opportunities along the way.
To learn more about the blocmates Meal Deal, head to the link here or click the image above.
The year of the based rollup?
We will probably hear a lot of talk about Ethereum-based rollups this year. Justin Drake proposed that L2s should offload the sequencing to the Ethereum L1 instead of relying on centralized sequencers, which the vast majority of L2s are currently using.
IMO, the future of Ethereum can’t be centralized databases on top of the L1. So, with that in mind, I think we see the pushback of what we have seen with Base, Soneium, and others…
I’m not entirely sure if this is enough to capture the attention of the masses and something that can drive activity and mindshare, but I do think it is something to be aware of. I kinda put this in the same bracket as ZKs. For now, it is a worthwhile pursuit, but does the market really care that much? Well, time will tell.
Teams to keep tabs on:
General-purpose L2s die… kinda.
I’m no longer convinced that general-purpose L2s will survive (other than Base). I think L2s need to drastically specialize by focusing on one vertical, i.e., gaming, consumer, AI, DePIN, etc., and leaving the generalized L2 space to Base… If Base launches a token (which I believe they will), given the new administration, it is literally going to be game over.
I’d love to think that when Arbitrum and Optimism reach Stage 2, everyone is happy and sings Kumbayah about the future of Ethereum, but I just do not see it happening at all… Both Arbitrum and Optimism are completely unaligned (naturally) and are wall-gardened ecosystems when it comes to composability and collaboration.
More on interop below…
I think the opposite is also true, specialized and bespoke L2s that are built for purpose will thrive. The likes of Superseed (CDP), MegaETH (real-time blockchain), Sophon (consumer/entertainment), Abstract (consumer), Eclipse (SVM L2), Mode (AI), etc. are all designed with the use case in mind.
Now, the question is: Composability and chain abstraction—Do we get there? Do we achieve a standardized interop solution? If so, who is it? And what is the end goal of all this?
Apps should be just that. It doesn’t necessarily matter where your tokens are, what version of the token they are, or whether you have gas for that transaction, sir. All of this is going to look prehistoric in a few years, and I think the early stages of what it will look like are being brought to market now.
We need interop, and it needs to be invisible. That doesn’t mean the interop providers underneath aren’t valuable. In fact, they’re quite the opposite.
Below decks, I’d never take my eyes off what LayerZero is up to. I think the team’s purist take on decentralized communication, censorship resistance, and immutability is what we should strive for as an interop solution.
The industry looks like it agrees, too, with it being heavily adopted, over 70+ chains integrated, over $50 billion transferred, and 200+ applications being built on LayerZero.
Another strong competitor that seems to tick all the boxes is Omni. Broken down into two main components: Omni Core and SolverNet, which allow users to have a completely chain abstracted experience across all L2s, enabled by fast cross-chain messaging handled by their solver network.
Beautiful branding, positioning, and super clear messaging is quite frankly refreshing in a space that oftentimes just. does. not. get. it.
I also wrote a short thesis on Across towards the end of 2024 and why I think it is becoming a strong go-to choice for a lot of interop projects across different protocols and chains/app chains…
Quickfire thoughts on established sectors:
DePIN
There are a hell lot of interesting DePIN projects. A lot of them are doing feasibly well when it comes to revenue, market penetration, and profitability. They can stand on their own two feet when it comes to being an actual company, which is quite rare in crypto.
Helium, io.net and Grass are the shining stars of 2024 in this sector and I expect that to continue in 2025 too.
The only issue I have with DePIN is that the very nature of token incentives to globally coordinate an outcome at scale is that the tokens themselves are not very investable until they hit critical demand. That being said, it is pretty easy to calculate when that demand/supply ratio is moving in the right direction.
I think the most interesting DePIN project for 2025 has to be DoubleZero.
Double Zero: Accelerating communication in high-performance distributed systems to increase bandwidth and reduce latency.
This is co-founded by Austin Federa, who was previously the Head of Strategy at the Solana Foundation.
In ape-speak, DoubleZero aims to allow users to permissionlessly access private fiber networks so that data (transactions) can be sent in double-quick time. This is aligned with Solana's goals of increasing bandwidth and reducing latency.
Providers of the said hardware can opt in and get paid for the usage of their connections and fibre.
If you couple this with Firedancer, Solana will literally process transactions faster than anything we have seen in this space (apart from centralized database chains).
We get it, this shit is super difficult to keep up with. It literally changes every single day. If only there was a newsletter that you could subscribe to for FREE, that lands in your inbox every weekday so that you can be the smartest person in the room.
Join 16,000 others who are staying ahead in the crypto and AI space by joining the link here.
DeFi
We are at a bit of a weird junction with DeFi. We had a mini-renaissance towards the end of the year. We even have the current sitting US president, who is technically a DeFi founder… kek
To expand on my point about wallets. I think the apps now need distribution. Yes, we have the on-chain natives being aware of how these things work and how to access yields, opportunities, etc.
But, similarly to how we have seen in-wallet swaps take a small fee, we could begin to see further DeFi products become integrated and abstracted away from the end user all in-wallet.
You only have to look at the Family wallet, which is part of Avara (Aave, GHO, and Lens protocol), to see that the super-app thesis can enable a new user base for DeFi.
Create lending and borrowing or even just basic yield opportunities for its regular users in an environment where they are comfortable.
Interesting developments and projects to watch:
Pendle - I think Pendle continues to do well with Boros, a new primitive enabling off-chain funding rate trading that taps into a market that trades upwards of $200 billion a day.
The end goal (in their own words) is to allow trading for every yield in the world…
I love the team; their grind is unparalleled. Wishing them all the success in 2025.
Ethena - These guys have done an incredible job in 2024 and their recent institutional offering announcement (iUSDe) is the cherry on the cake.
This is a huge step to bridging the TradFi world and bringing it on-chain. I’d expect this to continue to do extremely well in 2025.
Uni V4 - The fourth iteration of Uniswap is on the horizon and this design-unlock for liquidity, trading, loss versus rebalancing, LPs and degens-alike is going to be a game changer for the industry. There are multiple elite teams building on top of Uni V4 which all have unique use cases. I don’t have the time to go into each of them now, but I’m leaving below a short list of my personal favorites:
Stablecoins - We have recently seen a few more stablecoins with newer dynamics come to market, such as Usual and M^0. These are stablecoins, which like Tether and Circle take in cold hard dollar mint stablecoins on-chain but then give revenue share back to either the stablecoin holders themselves or the governance tokens associated with the stablecoin.
I think this is almost certainly going to catch on. Am I certain they will grow? I'm not entirely sure, as all it would need is Tether to turn on a fee switch, and it’ll be lights out. It is an interesting concept, regardless.
I’d be very surprised if the likes of Visa, Mastercard, Robinhood, and Revolut didn’t roll out their own stablecoin this year. If this were to come true, then M^0 could be an interesting play as it has the ability and framework to create other stablecoins, as seen with Noble USDN.
Hyperliquid - I give the 2024 team of the year to Hyperliquid. They have thrown everything that was previously thought to be true out the window and said, fuck it, we will do it our way. Community-first, flipping off the big guys and having an incredible product to back it up too.
I see huge continued success throughout 2025, with the HyperEVM going live and a ridiculous ecosystem being built off the back of the groundwork being done by the community.
This ties back to what I am saying about generalized L2s. I’d first build a native product that people actually want to use and then expand out from there. It makes a hell of a lot of sense and behind Ethereum and Solana I’d put Hyperliquid up there as the third chain with the highest potential for now.
Jupiter - Jupiter is basically the homepage of Solana. The recent acquisition of key infrastructure and applications such as SolanaFM (block explorer) and Moonshot plus the rollout of their mobile app tells me that they are going for full-stack domination.
Being so close to the end user, with these recent product updates and acquisitions, will solidify Jupiter as one of the most profitable and high-growth products in the industry in 2025 and beyond.
Expanding on Jupiter's $2.2 trillion lifetime volume, the 0.05% simple swap fee, and 0.1% fee on complex orders like limits and DCA — not to mention Jupiter perps — I think this could be an extremely profitable and prosperous token to hold.
Meow recently announced that they have burnt over $3 billion worth of JUP tokens and will use 50% of the platform fees to buy back and burn JUP.
And the pièce de résistance: JupNet. An omnichain network to aggregate everything. The Jupiter team is rolling out the Jupiter SVM, Aggregated Decentralized Identity (ADI), and the DOVE Network (DOVE stands for Decentralized Oracles that Validate and Execute).
This puts the JUP token in a league of its own. At this stage, it is pretty difficult to value it — but it certainly shouldn’t be valued as an aggregator or a DEX token.
Royco Domination
I had a great time sitting down with Jack Melnick, who leads DeFi at Berachain, a few months back to discuss a novel liquidity incentive program that they were calling Boyco.
After many attempts by teams trying to figure out a way to incentivize liquidity in a modular fashion, i.e. not a one-size-fits-all approach, which leads to giving away excess capital when you don’t need to, and it naturally falls into the hands of people who will dump the token anyway.
Boyco is a Berachain liquidity initiative that utilizes the Royco Protocol. Royco allows anyone to create a market that incentivizes users to perform a specific action. It also allows the potential action provider to negotiate the active offer.
I believe incentives are so vital as a customer acquisition channel in crypto, and until now, they have been taken advantage of because the game was way too one-dimensional.
Farmers, airdrop hunters, liquidity locusts, etc., all knew what game they were playing, and it was a PvE strategy. Now, creating bespoke incentives for every action, asset, user type, and chain allows teams to incentivize the preferred behaviors more whilst not overspending on excessive and wasteful token emissions.
This has allowed Berachain to amass $1.8 billion in deposits before the expected launch in Q5. I can almost say with 100% certainty that every meaningful network will follow this and launch its pre-launch markets. This will eventually become the norm.
Royco will become a meaningful part of each new chain's go-to-market strategy, with the added benefit of having guaranteed liquidity on the day for the applications and users.
Berachain - I’ll use this as a segue to talk briefly about our good friends at Berachain. If anything I think we have reached a point where culture is the main factor of any ecosystem. Once Berachain launches the incredibly dedicated ecosystem they have built over there, it will flourish due to the demand, availability liquidity (PoL and Boyco), plus the novel innovations and composability between dApps and tokens.
I am obviously biased because I am an investor, but I just can’t see this thing not doing incredibly well, barring any huge OpSec issues…
Initia - I am really excited about the idea of Initia. It is the perfect middle ground between composability (thanks to IBC) and app-specific chains without the limitations of what teams want to build.
Minitias (L2s) being able to tap into the native liquidity on Initia (L1) is also very cool, and I see this being a very strong force for attracting teams to come to build on Initia in 2025.
It is a game changer to have a high-performant L1 and extremely flexible environments for apps to build on the L2 level, all the while not breaking compatibility.
The team also kickstarted the Echo launch trend, where the round was 25% under the latest VC round. I think that this sets a strong precedent for how they will conduct their business following the launch.
Top blockchains to watch in 2025
- Berachain
- Botanix
- Eclipse
- Movement Labs
- Monad
- Initia
- Hyperliquid
- MegaETH
- Fogo
- Skate Chain
- Anything that launches using Commonware…
Final thoughts
Bitcoin L2s/DeFi - I think the funding over the past few years will finally result in some standout L2s and products coming to market in 2025. The idle BTC sitting on the sidelines will create immense waves in TVL for some of these chains. I’d boldly predict a Bitcoin L2 to break into the top 5 chains by TVL in 2025.
Gaming/AI - Will gaming ever happen? Who the fuck knows at this point. If so, I believe AI will be integrated all over the place. NPCs will have memory, create infinite storylines and gameplay will never be the same again.
Kaito - Kaito is the SocialFi app that we all have been waiting for. Yaps have changed the game and met the users where they are without having the additional need to create additional content to prop up a reputation on YET ANOTHER APP.
We at blocmates couldn’t live without the Kaito platform and are very fortunate to have had the chance to invest in this product and team.
Memes? - I think they are cooked. There’s a new shiny toy in town, and it can enable larger capital to invest based on a ridiculous thesis that funds and capital allocators can justify to their LPs. You cannot say with a straight face that you are deploying LPs’ hard-earned capital into Retardio and not expect some recalls on the funds. AI agents, on the other hand, ain't gonna get you fired.
SOL Flips ETH
I have been saying Solana will flip Ethereum for the past two years, so that is not ne,w and I expect that trend to continue in 2025. I am not sure if it happens this year but maybe shortly after. I also think it could actually flip it, but on the downside, when the wider market falls.
Given we have just seen the leader of the free world launch a memecoin for himself and his wife and the destination of choice was Solana, I think that kickstarts a flurry of new speculative tokens surrounding celebrities, IP, and companies even launching tokens.
The future is going to be tokenized, and it will happen on Solana.
It is quite clear that even despite the front-end issues faced by some applications during the largest on-chain event in history, Solana is the only network that could support such a circus.
That is bullish in its own way, and I don’t think for a second that reaching 10% of the Nasdaq’s daily transactions will end there for SOLDANOOOO… particularly with the SOL ETF odds at ATHs (92% at the time of writing).
If I were to guess, $1,200 is on the cards…
My thinking is that if ETH breaks ATHs and hits anything from $5,000-$6,000, people will rage quit. I could be wrong here, but I do not think I am for a second.
Base launches a token
Base will launch a token given the drastic change in US government. It only makes sense for them to have this as an incentive mechanism to drive more and more users to the platform. I don’t even think this is that wild to say, to be honest.
Conclusion
All in all, I think we’ll have a very productive year. However, I am slightly worried about inflation coming back and rearing its head. I think Trump and Powell will have many differences, and this will be quite public.
I can’t see a real reason to be bearish other than everyone is bullish. Stay sharp, keep active, try new stuff out that you stumble upon, and try to keep up.
If you are looking for a place to chat about the markets, then the blocmates Discord is open to everyone for free, you can join here
Finally, if you enjoyed this report, this is only scratching the surface of what we cover at blocmates. Last year we released over 300 free research reports, 250 podcasts, 260 newsletter editions, and 50+ premium research reports behind the Meal Deal.
Expect bigger guests, better content, and more value from us this year! You can follow all the action on our Twitter/X account here. GM future billionaires, we are finally back with the third installment of the blocmates year ahead thesis. First and foremost, let’s take a quick look at the 2024 calls that were made and decide if I am full of shit or not:
