blockchain in the age of AI
Web3 is doing just great, so they say. We’ve got:
- The outright scams and pyramid schemes facilitated by blockchains.
- The Serious Businesses with a level of funding that that only makes sense in a low interest rate mindset. (This isn’t unique to blockchain, see pets.com and WeWork.)
- The digital minimalists and tech contrarians that keep banging the drum of: “Why does this project need blockchain? Why not just use a traditional centralized database?”
It’s not controversial to say that the new generation of AI has created more value in the past six months than blockchain has in an entire decade.
And now we’re seeing a whole lot of pivoting.
We live in a post-AI world, and the air — and funding — is being deflated from the blockchain balloon. Now, more than ever, we need to be realistic about the future of the space.
Scott Alexander provides a reasonable defense of cryptocurrencies in his blog: ‘Why I’m Less Than Infinitely Hostile To Cryptocurrency’. Which is good, because I don’t want to be written off as an advocate of replacing our decent financial system with cryptocurrencies!
As for the state of web3 in the popular zeitgeist… lets just tactfully say that the current story of web3 doesn’t capture the true potential of blockchains; there is no coherent narrative that appeals to those that make decisions informed by politics and philosophy. The optimist of the world, those mythical people who know how to leverage their free time and goodwill to change the world, have yet to be sold on web3.
I don’t say this to be contrarian or to grift engagement. My intention is to propose a path through the quagmire we find ourselves in.
centralized databases will always enshittify
Myspace, Digg, Facebook, Instagram, Twitter, Reddit, etc. — they’ve all followed the lifecycle process of enshittification detailed by Cory Doctorow in their blog post: “Tiktok’s enshittification”. (2023’s word of the year no doubt!)
It’s not just something that happens to social media companies. It just hurts more when it happens to social media companies because they are uniquely valuable to their users. It’s our data; we created it, sometimes out of status-seeking, sometimes out of goodwill for strangers, and often out of our need for connection and love of our friends and family. Watching a platform you’ve invested a decade into light itself on fire… it sucks. But here we are.
the value of data
In the past 30 years of the nascent information age the strategy for startups has been to capture as many users as possible, as quickly as possible, while burning through seemingly endless mountains of cash. Only once the platform becomes a monopoly is profitability discussed. Low interest rates have fueled this strategy of monopoly-at-any-cost, but this is less a strategy and more a business requirement for a business built on information and data.
Ultimately, the only valuable assets media tech company has are it’s users, the users connections, and the content the users generate. Therefore, the only moat possible for a business built on data is to safeguard that data. This is accomplished by creating walled gardens — parallel Internets accessible solely through specific apps and narrow APIs.
If users want to maintain access to their data and connections, they must stay on the platform. Users are effectively held hostage by their own investments into the platform. Even if a user wants to break away and leave it all behind, the monopoly of network effects means they’d be doing it alone
you are the product
If this is not the inherent nature of all businesses, it’s a common enough destination that entrusting a centralized database with our memories and connections is too big a risk. Even if we can’t articulate this, we’ve all come to understand it intuitively. That’s why over the past few years, we’ve collectively grown pessimistic of social media and tech in general. Five years ago, it was rare to find individuals without an online presence. Now, it’s increasingly common to encounter people who exist only anonymously on the web.
Does your startup have a strategy that looks like this?
growth -> monopoly -> enshittificationWell, sorry to tell you, but no amount of low-interest-rate mindset funding will provide the runway needed to succeed in capturing the users required to create Yet Another Walled Garden in 2023. We, the users, know the game, and we’re not interested in playing it anymore.
and then came AI
The modern AI tools are almost completely trained on the internet and it’s corpus of user generated content. There is tangible, real-world value in the work users produce on social media platforms like Twitter, Reddit, Quora, and Stack Overflow.
AI poses an existential threat to these platforms. Why bother navigating through Google’s labyrinth for an answer, wrestling with the ads and the ̶t̶i̶m̶e̶-̶w̶a̶s̶t̶i̶n̶g̶ …engagement driving dark patterns, when an LLM can provide the answer without such hassles? And you might think this is only the case for search engines, but perhaps that’s just because LLMs as a replacement for search are the only use case we’ve seen yet. LLMs likely pose a risk to all businesses built on content aggregation and discovery because content is what drives engagement on social media platforms.
In rapid response, these platforms have raised even taller walls around their gardens. Not just to maximize their revenue from LLM training, but also to reinforce their moats; their profit driven motive means they must do everything possible to ensure that their value — our data, our content, and our connections — remains trapped within their platforms. Even if it means amputation.
Not only is the content we create the product, with the revenue generated from it going solely to the shareholders, but the content is also under the complete control of financial interests working on short, often quarterly, timelines to increase revenue. This incentive structure often results in stewardship that can be incompetent and out of step with the community it governs. For instance, Tumblr was sold to Yahoo! for $1 billion, and it quickly shed it’s user base and relevancy. In it’s self-destruction it wiped out years of effort users had put into building a community. It sold again for just $3 million only a few years later. Twitter is seems to be treading this path, and as of June 2023, Reddit is speed running it.
In short, the beliefs that encourage altruistic social media behavior have been slowly drowning for years; We no longer believe any tech company will be a good steward for our digital lives. And AI is the stake through the heart.
profiting from, not for, the public
The idea that our communal efforts, that something created for friends, family, and strangers alike, belong to the public domain, is not a new notion. Yes, posting this blog on Medium/LinkedIn benefits my career, but it also, hopefully, benefits everyone. The free flow of knowledge and information efficiently moves us forward. Everyone should be able to access and read this, and any LLM should be able to regurgitate the ideas presented here.
Giving Medium/LinkedIn control in perpetuity defeats our goals in content creation. If Medium places this behind a paywall, it slows down the flow of information. LinkedIn confining it within a walled garden, isolated from the open internet, makes it undiscoverable for search engines, LLMs, and end users. The moats that businesses construct to safeguard their revenue introduce inefficiencies to our economic system and slow our progress as a society.
We don’t create to increment a hedge funds balance sheet. We make cool things to impress potential mates and, occasionally, because we love each other. Social media is just arbitraging that value at our expense.
all this has happened before
The Free Software movement began in earnest in the 1980s with a similar idea: software isn’t zero-sum like physical goods. It can be used by many, and thus, for the sake of economic and social efficiency, it should be available to all. The movement didn’t gain momentum until the late 90s, and open-source tools like Linux and Git didn’t become commonplace until the mid-2000s. Even if it took decades to get there, the results have been profound: every interaction a user has with technology benefits from Free and Open Source Software (FOSS).
The success of FOSS has led to a democratization of tech, but over time, the Internet, and the world at large, have still become increasingly monopolistic. Adam Smith, the man who coined much of the verbiage of free markets in the 18th century, opined that a free market can not stay free without intervention. Similarly, in the digital age, data monopoly approaches one while competition approaches zero. It turns out that while software is important, it is less important than the users and their data.
ideas iterate generationally
What we believe dictates our political systems. And what we believe changes slowly. So slowly that I often wonder if our world views, and therefore the world itself, only changes when a new generation comes to power. Young people in the 90s went on to senior leadership roles in tech and took with them their beliefs in FOSS when building those tech companies. But we need to move faster than inter-generationally.
Unraveling the perception of blockchain and web3 as an incoherent libertarian movement of scam artists and gamblers is a challenge. The pro-democratic and classically liberal intelligentsia class that traditionally stalwart for progress are now reflexively against blockchain, web3, and anything associated with the space. However, they are exactly the people who need to advocate for free and open data as a requirement for economic and personal liberty.
Change takes time. FOSS took more than 20 years to come to fruition. We can and should move faster with free data.
blockchain is no longer sexy, and that’s ok
Over the last ten years, blockchains have come a long way. They may never out compete a centralized database running on AWS in terms of speed, scale, or cost, but they are now fast enough to provide a good user experience and scalable enough to handle projects with millions of users. And they’re only getting better.
Even if some of the glean and shine has come off of blockchain, I’m still a believer; if not of ‘blockchains’ and ‘web3’ as we currently sell it, then of free and open data by any other name or technology:
# Call it:
public ledgers,
global state machines,
hashgraphs,
fediverses,
decentralized databases,
very advanced linked lists,
etc...But open databases won’t come to fruition through the path followed by traditional business. Promoting competition is inherent to the structure of open databases, and competition is fundamentally antithetical to traditional business. Therefore, it’s difficult to imagine VCs funding the next blockchain based media super-app. And retail investors aren’t going to the moon next week investing in these projects either. Thus, the standard marketing campaigns and social media hype shilling won’t be pushing free and open databases forward.
Adoption will ultimately will come from:
- advocates selling the vision,
- builders building it,
- and believers adopting it.
With the future of the internet in your hands, what will you build?
