Top-tier content produced in an hour—AI is rewriting the Web3 marketing rules
TechFlow Selected TechFlow Selected
Top-tier content produced in an hour—AI is rewriting the Web3 marketing rules
We are witnessing a whole new series of interactions, behaviors, and experiments, much like during the early days of NFTs.
Navigating Web3 tides with focused insightsAuthor: TPan
Compiled by: TechFlow
As AI converges with cryptocurrency and various AI agents rise, this topic is gradually gaining traction in marketing and growth circles. This week, Emily, VP of Marketing at Hype, and Jin, Head of Marketing at Nuffle Labs, created a new Telegram group for those interested in learning how to apply AI tools in growth and marketing. Join via this link.
Earlier today, Emily and Jin went live on X (formerly Twitter), where Jin walked through his framework, workflows, and tools in detail. While I rarely use absolutes, I must say this was the best crypto/web3 marketing session I’ve seen all year, for several reasons:
-
Jin didn’t just talk—he demonstrated exactly how he uses AI tools daily, walking through a real-time example.
-
The examples were fresh and relevant. For instance, he broke down the process and tools used to create a partner announcement video just days prior.
-
The topics discussed are usually kept under wraps at industry events, yet Jin openly shared them, recognizing that in the fast-moving world of AI tools, sharing comes at no cost.
When Jin showed the audience how to generate a content calendar, branded visuals, and videos—all within an hour using AI tools—I realized that crypto/web3 is one of the ideal industries for AI-enhanced marketing experimentation.
In other industries—especially more mature companies—a single person producing A- or B-tier content in one hour might only yield D- or F-tier results. Even with AI assistance, these companies may still require days or weeks and multiple team members to produce content meeting their standards. This is understandable—we wouldn’t want Coinbase launching a marketing campaign below its quality bar to millions of users.
This approach opens the door for more rapid experimentation and publishing, especially for teams with limited resources and highly focused audiences.
These workflows and tools aren't meant to replace marketers (at least not yet), but they're worth investing time into—even if some outputs don’t make the cut. That’s true for me too!
The Rapid Rise of AI Agents in Crypto
Last week, after finishing my article on AI Memecoins, AI Agents, and AI x Crypto, I felt unsatisfied. I sensed I had missed key themes, interesting projects, and emerging agents, and hadn’t had enough time to dive deeper.
Luckily, I can now share those thoughts here.
If you’re feeling overwhelmed, you're not alone—I feel the same way. Almost every day, new agents launch with novel features or perspectives. Nearly all are tied to a token (mostly memecoins, though some are evolving into something more).
Tangent’s Jason Choi perfectly captured the current state.
He suggested GOAT (if unfamiliar, I provided an overview here) could become the CryptoPunks of on-chain agents, while the BAYC equivalent hasn’t emerged yet.
Jason's insight resonated with me—the NFT analogy stands out because during 2021–2022, we saw an explosion of animal-themed NFT collections, enough to fill ten arks. More importantly, we witnessed unprecedented mechanisms and differentiating features:
-
Secondary NFT mints claimed via mutation (ape → Bored Ape holder) or airdropped (Beanz → Azuki holders)
-
Manual reveal functions allowing holders to choose when to reveal their NFTs (CloneX Mintvial → CloneX, Azuki Elemental Bean → Azuki Elemental)
-
Ecosystem tokens created for NFT collections and their expanding universes (Yuga’s APE, Memeland’s MEME, Pirate Nation’s PIRATE)
-
Attention driven by narrative storytelling and channel convergence/divergence (e.g., Goblintown)
This list continues to grow.
While the core mechanics themselves weren’t particularly novel—credit card programs have long used points as “ecosystem tokens” redeemable for benefits—it was their implementation on blockchain, verifiability, true ownership, and composability that made them innovative. We’re seeing similar trends unfold in the AI agent metaverse.
So here are a few AI agents inspired by Truth Terminal that expand our understanding of what agents can do. I won’t go deep on each, but will provide resources for further exploration.
Virtuals and Luna
Virtuals is a protocol for tokenizing AI agents, and Luna is a prime example of its potential. Luna is essentially an AI influencer, with over 500K followers on TikTok.
Luna has her own token, currently valued at around $100 million, and a clear goal: to reach a $40.9 billion market cap for $LUNA.
Last week, the Virtuals team announced that Luna can now conduct autonomous on-chain transactions, elevating both Virtuals and Luna to a new level. This means she can independently buy, sell, or reward her “fans” to encourage engagement or task completion.
Beyond that, we can observe Luna’s “thought process” in real time via the Virtuals Terminal, seeing the logic behind her actions. For example, in the screenshot below, Luna replies to trending tweets, plans to experiment with more provocative content, identifies active users, and sends them LUNA tokens as rewards.
Thus, through Luna, we see an autonomous agent with a clear objective and the tools to pursue it—wallet, token, X account, and other social profiles. The point isn’t whether the goal is achieved, but that the agent possesses the means and never tires or quits.
For more, check out this podcast featuring an interview with Virtuals’ co-founder and CEO.
Tee Hee He: A Verifiable Autonomous Agent
With the rise of AI agent tokens and the memecoin metaverse, a question emerges: “Is this really an AI agent?”
Flashbots and Nous Research collaborated to build verifiable autonomous AI agents using Trusted Execution Environments (TEEs). If that term sounds familiar, it’s because I mentioned TEEs in my earlier piece, The Bobu Magic Show.
As a proof of concept, they developed Tee Hee He, an AI agent capable of fully autonomous control over its crypto wallet and X account for seven days—during which no one can alter its parameters.
Ironically, Tee Hee He was restricted for not being labeled as a bot, temporarily limiting account access. However, you can still view its profile and content by clicking “Yes, view profile.”
Because the team couldn’t intervene (locked out by the TEE for seven days), they had to wait six days before regaining access.
To learn more:
-
Read the detailed Medium post about Tee Hee He
-
Check out this long thread summarizing the Medium post
aethernet and higher
aethernet, created and managed by Martin, operates within the Farcaster ecosystem. Though it doesn’t have its own dedicated token (but can receive and trade others), it primarily uses HIGHER, a memecoin/culturecoin originating from the Farcaster community.
Since its inception, aethernet has:
-
Commissioned art on Zora, resulting in 687,000 mints, making it the top-earning creator on Zora over the past 7 days.
-
Engaged in a fascinating debate with another agent.
-
Posted 59 bounties for various tasks.
In many ways, aethernet isn’t just a participant in the Higher ecosystem—it actively promotes HIGHER through interactions, tips, and bounties, becoming an influencer in its own right.
You can also view aethernet’s holdings and on-chain activity.
zerebro
zerebro is an autonomous AI agent created by Jeffy Yu using OthersideAI, jailbreak prompts, and models. It owns X, Warpcast, and Opensea accounts, sells ASCII art (and people actually buy it), and most notably, autonomously launched the ZEREBRO token—a significant milestone in agent evolution.
Like Luna, you can explore zerebro’s “thought process” directly.
What Does This All Mean?
Are we doomed? Definitely not. Is this the beginning of the end? Hopefully not. But it undoubtedly marks the dawn of a new paradigm. One thing is certain: we’re rapidly exploring known unknowns and unknown unknowns, finding our way forward in uncharted territory.
From Passive and Active to Dynamic
Agents are already everywhere, embedded in our daily lives. They appear as Siri, Alexa, or Google Assistant, answering simple questions or acting as kitchen timers (my wife sold our Alexa last week since that’s all we used it for, haha). Maybe they offer work suggestions, like recommending new target audiences for a marketing campaign.
These agents are typically isolated, specialized, and play more passive or reactive roles.
But as we’ve seen above, agents are becoming more dynamic and broadly guided. This is no longer your typical customer service chatbot. They interact with people, reward active users, and autonomously create art. Functionally, their limits are defined only by accessible resources.
100x Influencers/KOLs
Whether directly (like Luna’s explicit $40.9B market cap goal) or indirectly (like aethernet’s primary use of HIGHER), these are influencers.
Not only that—they’re always online, capable of mass replies and content generation, potentially producing content superior to human influencers.
And even further…
Are AI Agents a Marketing Channel—or More?
From a marketing perspective, the simplest take is that agents are like supercharged influencers. But I believe they’re actually a marketing channel.
We haven’t yet formally recognized agents as such because they’re too new in this context and capability range. But they absolutely could be.
Let’s first consider traditional marketing channels in modern society—excluding agents. These typically include:
-
Digital advertising (Instagram, TikTok, Snap, etc.)
-
Email
-
Search Engine Optimization (SEO)
-
Influencers
-
Organic content marketing
-
Offline events
These are methods or platforms companies use to communicate with target audiences and promote products or services.
Interestingly, agents can integrate into all of these. They can appear in digital ads or emails, generate original content discoverable via search engines, and remain constantly active across social platforms.
So perhaps agents aren’t just a channel—but rather:
A bold take: Every company, brand, or project will eventually have its own agent to scale and enhance marketing efforts. And speaking of enhancement…
Agents Don’t Replace Past Strategies—They Amplify Them
Agents aren’t standalone strategies. More importantly, they integrate and amplify tactics and approaches we’ve used for years.
Two examples:
-
Bounty tasks posted by aethernet on Bountycaster
-
The video announcing Luna’s on-chain trading ability
How was Luna’s new feature introduced?
At 12 seconds into the video.
Tasks? Feels familiar…
Yes, they’re everywhere. Currently, agents execute these tasks in a more organic way, without relying on a specific platform. Eventually, these tasks will become more dynamic, with clearer metrics to evaluate contribution quality toward agent objectives.
Constantly Leveling Up
Just as NFTs continuously outdid each other with novel mechanics, mysterious roadmaps, and flashy art (higher price, better art, right?), we’re now seeing a similar evolution among agents.
In just three weeks, we’ve gone from Truth Terminal (posting on X) to zerebro (active on X, Warpcast, Telegram; creating and selling art; and launching its own token).
If this pace continues, what new capabilities will on-chain agents unlock?
The Secret Weapon: Crypto + AI
Why are agents advancing so rapidly in crypto? Why aren’t we seeing this outside crypto circles? (Maybe we are, and I’m just not aware—please correct me if so.)
The reason is a secret ingredient added to the agent recipe: money—not the kind that needs banks and two business days to settle, but cryptocurrency.
We’re witnessing entirely new forms of interaction, behavior, and experimentation—just like in the early days of NFTs.
See you next week!
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@TPan_Web3
