Uniswap founder Hayden Adams posted on April 23, 2026 that his X follower count had jumped from a few hundred new accounts per day to roughly 600, with little change in what he was posting. His conclusion, delivered in one line, was that the dead internet theory is looking more and more real. It is a small data point from one founder, but it lands on a platform that crypto traders still treat as the primary read on market mood.
What Adams actually said
The post is short. Adams wrote that his follower growth went from a baseline of a few hundred new accounts per day to around 600, and tied it to the dead internet theory, which holds that a growing share of online activity is automated rather than human. He did not publish screenshots of the follower analytics, and he did not accuse any specific botnet or growth service. As of posting time, the tweet had 437 views, 13 likes, and 3 replies on the source account (x.com).
For context, Adams runs one of the largest decentralized exchanges in the industry and is one of the more measured founder voices on X. He is not a growth hacker chasing vanity metrics. When someone in that seat flags a sudden doubling or tripling in new followers as suspicious rather than flattering, it is worth reading twice.
Why this matters for crypto specifically
Crypto lives on X in a way most sectors do not. Token launches, governance debates, exploit post-mortems, and price calls all route through the same timeline. If a meaningful share of new accounts are automated, then:
- Follower counts on founder and protocol accounts are a noisier signal than they used to be.
- Engagement metrics that bots and quant desks use to read sentiment get contaminated.
- Reply sections, which are already heavy with airdrop farming and engagement farming, skew further toward automated noise.
None of this is new. What is new is a founder of Adams' stature saying the shift is visible in his own numbers, in real time, in April 2026.
The AI bot angle
The dead internet theory predates the current wave of large language models. It used to describe cheap spam and template bots. In 2026, the concern is different: accounts that post full sentences, reply in context, and mirror the tone of whatever community they land in. Hayden's observation does not prove that his new followers are AI driven, but it fits a pattern that has been reported repeatedly across crypto Twitter over the past year, including during the KelpDAO exploit coverage, when reply sections filled with near-identical summary accounts within minutes.
For anyone running a wallet, an on-chain reputation system, or a card product that relies on social proof, the read-through is practical: the cost of passing as human on X has dropped, and the cost of verifying a real person has not dropped at the same rate.
What founders and users can actually do
A few concrete responses are already visible in how crypto teams operate:
- Treat follower counts as a weak signal. Active, consistent authors with low follower counts often carry more signal than a six-figure account that grew overnight.
- Weight replies from accounts with a verifiable on-chain history or long-lived identity higher than generic accounts.
- When reading market sentiment off X, filter for accounts that existed before a given narrative began, not accounts that appeared once the token did.
These are not new rules. Adams' post is a reminder that they are getting harder to ignore.
Overview
Hayden Adams used one line and one data point, his own follower count, to argue that automated activity is a growing share of X. For crypto, where that platform is still the default sentiment layer, a Uniswap founder saying the timeline is getting less human is a signal worth logging. It does not change any protocol, but it should change how traders and builders weight social data.
