The State of AI Agents

When I first mentioned AI Agents, no one was talking about them—valuations were dirt cheap, and user adoption was minimal. Over the past weeks, these tokens exploded in popularity, delivering significant gains for early entrants. But everything in crypto moves in waves, and I believe Wave 1 is ending. In this post, I’ll break down what happened, why I’m shifting my approach, and why I’m still convinced AI Agents are the most exciting narrative heading into 2025.

(Remember: This is not financial advice. I’m just sharing my personal opinion. We’re just a random Instagram page after all.)

Some Bragging

Excuse me for bragging here, but I want to capture this moment with you guys. Here is an overview of performances, from entry to tops of each asset mentioned:

1. VIRTUAL: +1,600%
2. MISATO: +550%
3. SEKOIA: +8,405%
4. VU: +1,680%
5. WSB: +320%
6. VIRTU: +1,160%
7. WEBSIM: +425%
8. SAINT: +400%
9. WAI: +4,075%
10. DEGENC: +1,405%
11. DOC: +255%
12. TAO: -20%
13. EAI: -47.5%

Average Performance: +1,555% (16.55x)

Hit Rate: 84.62%

Note that these performance are calculated from valuation at mention all the way into their all-time highs. However, until today everything remains positive with the exception of SAINT, TAO and EAI. I consider FOMO a failed investment for me, because the team is still not allowing presale buyers to claim their tokens even though valuation is not down from entry (hopefully that changes).

Breaking Down Wave 1

To understand where we’re going we have to first look back at what we had in AI Agents.

A big part of Wave 1 was capturing attention—think of it like building a massive user base before focusing on sustainable revenue. Most projects relied on trading taxes to generate quick income, which isn’t a real revenue model. Eventually, that has to evolve into commission-based services, subscriptions, or other tangible sources of revenue. That pivot is what I expect to see in Wave 2 and beyond.

In my view, most AI Agent to date fell under one of three main categories:

1. Market Influencers & Insights: These agents promise to simplify the market, provide advanced analytics, or curate insights for traders and investors. Unfortunately, most are just ChatGPT wrappers pulling in unstructured data. True innovation is rare. One notable exception is AIXBT, which offers fairly decent insights—and the market rewarded it with a valuation near $1 billion. DEGENC is part of that category too.
2. Fund Management & Trading: Some agents position themselves as automated VCs or trading bots. In reality, humans are still making the decisions behind the scenes. While I don’t believe AI Agents should be fully autonomous from day one (founders do need to guide them), I’ve seen certain projects exploit their platforms for cheap tokens in exchange for marketing. This “predatory investing” is neither sustainable nor transparent, and it gives AI Agents a bad name. The good ones are SEKOIA and WAI.
3. DeFAI: One-Click Crypto: DeFAI aims to simplify crypto investing by allowing you to “tell” the AI what you want done (e.g., bridging, staking, swapping). The AI then handles the transactions for you with minimal slippage. I see enormous potential here—this is a practical, revenue-generating use case—but it entered the market at a tough time. As the broader AI Agents sector corrects, DeFAI projects will likely get hit too. VU is part of that category.

More categories must emerge.

Why Wave 1 May Be Ending

Over-Investment & Vaporware

The huge returns we saw early on attracted low-effort builders who churned out projects purely to capitalize on hype. Many simply copied the “memecoin playbook”: Twitter shilling, telegram promotions, and inflated claims. This “quick buck” mentality oversaturates the market with fluff.

Straight-Line Gains

A lot of AI Agent tokens rocketed upward without real corrections. That’s not sustainable—markets need regular pullbacks to weed out weak hands and projects with zero substance.

Potential But Uncertain Timeline

I’m not saying this is the exact top, no one can nail that and we might see another sudden spike if, say, a major exchange listing or big partnership is announced. But, we could also plunge into a quick mini-bear phase that lasts 3–8 weeks, flushing out the nonsense before the next major wave of innovation. No one can predict the exact timeline, but the sector is definitely looking overheated.

I could also be completely wrong.

My Strategy

I captured most of the upside from Wave 1. With prices and hype at a peak, I’m shifting to 70–80% cash in my AI Agents portfolio. I already hold some cash after sells to recoup initial investments but the gains have been so big that the cash became a small part of my portfolio.

That’s not a call to panic-sell; it’s simply a risk-management move. I don’t believe in “roundtripping” profits (i.e., watching them disappear in a crash).

I’m still holding 20–30% of my positions. Why? Because there’s a chance we might see further upside before a correction. Markets can stay irrational longer than we expect, and I’d rather have some skin in the game in case that happens. This section will consist of the agents and infrastructure (VIRTUAL/EAI) that I already hold.

Long-term, I see AI Agents becoming massive—bigger than nations, replacing countless jobs, and amassing “populations” of billions. Once this correction shakes out the fluff, the best teams will emerge with genuine revenue models (commissions, subscriptions, etc.). Wave 2 could dwarf anything we’ve seen so far.

Until then, I prefer to preserve capital and then redeploy into the strongest players—those with actual product-market fit and stable revenue streams beyond trading taxes.

Conclusion

Wave 1 of AI Agents has been a remarkable ride. If you got in early, you likely saw unbelievable returns. Now is the time for caution: the hype has drawn in opportunists who are turning the sector into a minefield of vaporware. While there might be more upside left in the short term, I believe a correction is overdue.

Once the dust settles, I expect a few truly innovative AI Agents to stand out, leading us into Wave 2—a cycle marked by real revenue, genuine user adoption, and long-term sustainability. That’s where I plan to go heavy again. Until then, capital preservation remains my priority.

(Again, none of this is investment advice—do your own research and make decisions you’re comfortable with.)

Frequently Asked Questions (FAQ)

Q: Why not stay fully invested if AI Agents could still go higher?

A: Crypto markets can be unpredictable. I’ve already seen huge gains and I’d rather hold onto a significant percentage of profits than risk losing them in a possible downturn. Holding 20–30% still keeps me in the game if the sector pumps further.

Q: How much downside are we talking about here?

A: If a full bear market materializes and I’m right, a lot of tokens are gonna go down -70-90%.

Q: When do you expect Wave 2 to start?

A: Can’t time markets but I’d say somewhere around second half of February or March. Don’t hold me to this, no one can time markets.

Q: Are you going to keep posting?

A: Of course we are (and I will too). AI Agents are not the only thing in crypto and they will also come back bigger, better and stronger. This post is part what we promised you, we said we’ll do posts on big innovative ideas and warning of wild market swings too.

Q: What about the new post dated 25 January 2025?

A: Date will depend on valuations but we’re already seeing great agent ideas start to take shape, we’re also helping a team build one. We talked about The 100x Club subscribers being rewarded and given access before others. We’re still in the “talks” stage. No more details (don’t ask, we won’t answer at this time).

That’s my honest take on where we are and where we’re headed. If Wave 1 taught us anything, it’s that catching the right sector early can produce life-changing gains. But no bull run lasts forever. A reality check is healthy—and it’s in that reset that tomorrow’s biggest opportunities often emerge.

‍