here is what i sent the @anagramxyz team after DAS
it was an interesting one. defi protocol founders have ditched the tie dye shirts and are wearing full suits now. Pretty clear that if your company does not have an institutional partner (i.e. big bank or trad fintech) then they will struggle to sell their product. A lot of those partnerships are being formed now but in the next few years they will be ossified and banks will likely stick with one partner. Its time to get selling.
it was great to see friends and portco founders but on the venture side there are not many innovative defi solutions/companies being built. the majority of deals are either packaging trad yields into an onchain vault, derivatives/HIP3 play on a specific asset class or AI Agent Commerce/Prediction Market or other ideas that are 'hot' right now. i can see many of these businesses making money but few generating venture style returns. i think this is reflective of a larger trend in new company formation. higher velocity, lower capital requirements & single/double outcomes rather than homers.
there is a massive opportunity to zoom out and think very long term about the next innovation in crypto. everywhere you look companies are in a rush, ai has compressed new product releases to days. Founders are now looking at the landscape with rose colored glasses. Everything can be built.
patience and prioritization are incredibly important. especially more now. ai productivity boosts have made the cost of scope creep/new prod development effectively zero. its already difficult being a ceo, but being a ceo with an ai-enhanced engineering team means you need to have maniacal focus. i also believe that startups must find ways to bring ai productivity gains to sales/marketing, i asked every company i met with and not a single company has cracked this yet. there is still a tension as engineering output has 10-100xβd and sales/marketing has 2xβd.
its clear to me that ai productivity boosts are better harnessed by second or third time founders. they basically just gained a free chief of staff at the earliest stages of the business & they know what to aim it at.
a lot of funds are struggling to raise capital. FoF's are struggling & the broader LP base is struggling which is trickling down into the venture ecosystem and has been for some time. i think in the next 6 months many funds will close shop driven by this inability to raise. the situation does not seem to be improving anytime soon and the coins remain down. I expect to see mandate creep, I.e. we do robotics and crypto as a way to appeal to a broader lp base. as with all things, you need a definitive edge to succeed if funds go multi-domain.
the positive for startups is that trad vcβs have now rotated back to crypto as βcrypto is fintech nowβ. on average, these investors are higher quality and have a comparative advantage given their connectivity to the broader fintech/trad finance sector. for our portcos raising in the stablecoin/payments space it is imperative for us to form relationships with these firms to help founders access this capital.
i think reality has finally caught up to crypto's printing money out of thin air via smart contract. A few months ago, after the movement/rushi thing broke, i had a friend ask me an interesting question: "Do you think we are wasting our time?". he was referring to the fact that movement, a company with zero users/revenue and was a known scam, traded at a $1.5b market cap. i remember this conversation distinctly, because it was one of the first times that i actually did question if i was wasting my time in this industry. movement is used as an illustrative example to represent the countless protocols with similar dynamics. the market agrees that these tokens are in fact a waste of time and many of the low float high fdv tokens that we have grown to love/hate will not be returning.
our liquid book should try to find assets earlier in their life cycles. with metadao and other fundraising platforms we will start to see better startups launch tokens earlier in their lifecycle. if we are willing to write $2m tickets into private companies with a liquidity expectation of 8-10 years then we should be jumping for joy that we can half our entry valuation, invest in great early stage businesses and have instant liquidity.
i have not seen any great businesses launch on metadao or other fundraising platforms but we should watch this space as the trend seems to be picking up and this method of fundraising feels much more in-line with the structure of ai native businesses. if your idea can be built in 2 days why would you waste a month doing a venture fundraising roadshow. just launch on the blockchain, get instant capital and continue to irerate.
if canton wins i will be miserable. it is the most bizarre ecosystem that operates in an antithetical manner to the free markets that crypto was built to proliferate. yet somehow. they seem to be crushing it & the product resonates with many institutions.
crypto foundations will need to follow in the footsteps of solana and operate in a top down manner. they need to either buy or build first party products for fintech/trad bank customers. if they don't then tempo, arc or canton will and there will be less innovation as a second effect. if I am an institution, would I rather deal with a single counterpart (i.e tempo) and receive every feature/product i need to handle my blockchain deployment (stripe/bridge/privy) or would I go to Ethereum and be directed to 5 different startups that each solve a portion of my problem but none are built to cohesively fit together and the smart contract risk is amplified + steeper learning curve.
anagram is looking good here. i think the ability to build and deploy capital in the ai native world is incredibly important. i expect many other venture funds to start doing 'incubations'. its just too tempting/the tools are there + getting a high dollar on dollar return via incubation is mathematically easier than venture.
i was very impressed with the blockworks team ability to ship a high quality event. if crypto continues to grow they have the potential to build a strong business.
the institutions treat crypto like a dirty word. its almost never said.
crypto is dead. long live blockchain.
Reactions and replies to this article.
Emilie ai/acc βπ§π»
@emilie4rio
@darkmarketio A few thoughts after reading this: 1. a lot of DeFi-related crypto financial innovation already belongs more to fintech now. You can clearly see the signs of the industry embracing tradition: the hoodie-and-T-shirt geeks are now trying to blend into the world of full suits. But why isnβt this also a more top-down, and in some ways more efficient, path to mass adoption?π€ͺBTW, this is also an interesting signal for event managers.π· 2. AI-driven productivity gains in engineering are not just changing the path of founders. They are also changing the return path of capital. The old playbook of getting in early, waiting, and exiting at a huge multiple may no longer hold. VCs will either have to build themselves, spread their bets across more baskets, or find ways to exit earlier and more flexibly. They will need to adjust their expectations and strategies accordingly. Founders, too, should adjust their expectations of VCs, and the way they negotiate with them. 3. by the same logic, if the low-float, high-FDV path no longer works, that almost tells you that the βget rich from token investingβ opportunity is unlikely to come from VC-backed projects anymore. Either the upside comes from short-lived meme token sentiment, or these VC projects can only use tokens as a tool for community participation and incentives β even if that makes them far less attractive than before. The other role for tokens, as the author mentioned, is as a parallel fundraising path that broadens access to capital. 4. I could not agree more with the point that AI has improved sales and marketing far less than it has improved engineering. You can already see that reflected in the composition of these jobs. This is not to say that people with a BA/MA degree cannot vibe code. It is simply that the problems sales and marketing are trying to solve have never been the kinds of problems AI is naturally best at solving. Maybe that is also part of the hope AI still leaves to humans. These functions are fundamentally about trust. And because of that, brand effect and impact may be even more important than people think. 5. Crypto may or may not be dead, but todayβs AI and yesterdayβs crypto are, in many ways, just another dose of founders' drug, maybe with different duration. No matter how great the invention or how meaningful the progress, once we zoom out across a bigger time horizon, none of them are truly singular or beyond comparison. The underlying business logic, the allocation of productive forces, and human interaction have never really changed. Crypto only wrapped a revolution in distribution in the sugar coating of computers and finance, and gave our generation a cyclical opportunity. So no, distribution does not need a token. But tokenomics maybe did make a difference in our generation.
Takisoul
@takisoul
@darkmarketio excellent piece but kind of worrying that metadao is mentioned by name multiple times as the go to launchpad, considering how it starts to err into "insider trading is fine" side of things. if anything, ai native businesses will definitely turn to infra/launchpads with programmatic enforcements and transparency.
Metasal
@metasal
@DanJablonski @darkmarketio pretty accurate - i say Iβm in the blockchain and web3 industry because saying crypto is too risky to bring up
Greenpill Dev Guild
@greenpilldevs
@darkmarketio We stopped saying crypto a couple years ago, let all negative notions live with that word and we keep blockchain and others pure.
Maika Isogawa
@maikaisogawa
@darkmarketio blockchain moved beyond crypto while crypto was running out of narratives long live blockchain
Jim
@jimchang
@darkmarketio had the same thoughts for the past 6 months, but you articulated it way better
Kydo β’ 74K views
Dougie β’ 50K views
Wintermute Ventures β’ 101K views
Chris Dixon β’ 1.1M views
Ryan Watkins β’ 361K views