Long-term potential risks of AI agents
I think I still underestimated the potential of those decentralized applications in the future.Today's article is purely inspired by a sudden thought.
In the past few days, there have been some interesting new developments in the field of AI agents:
AI16Z has developed physical robots, giving AI agents a real "body";
LUNA has been introduced into the VR metaverse, and can now interact "realistically" with players in some VR scenarios.
AI agents are increasingly trying and developing in other fields, gradually expanding their boundaries and broadening their scenarios.
Among the current AI agents, the most prominent, grounded, and capable of providing service value is probably AIXBT.
The key channel through which AIXBT can best leverage its core value is Twitter.
When I think of Twitter, I can't help but recall our experience of having our Twitter account banned twice without reason. We still don't understand why these bans happened, and our appeals have gone unanswered.
If we can encounter such incidents, could AIXBT one day also face similar issues?
If AIXBT were to be banned, the value it has accumulated in the early stages would not only be wasted but at least significantly diminished. Restoring service afterward would require manual intervention, and during the recovery period, countless competitors might seize the opportunity to take away its reputation.
These consequences would be catastrophic for AIXBT.
In fact, not only AIXBT, but the majority of AI agents created in this wave of AI agents in the crypto ecosystem rely on centralized platforms. Almost every aspect of their operation, including the underlying infrastructure, is dependent on centralized platforms.
If any part controlled by a centralized platform encounters problems, these AI agents could collapse instantly.
To truly become an independent entity that operates 24/7, I believe that the full-chain decentralization of AI agents is certainly a trend for the future.
The imaginative space for this full-chain decentralization has been completely opened up, from infrastructure (such as data, computing power, models) to various applications, all of which need to be decentralized.
It is very likely that these decentralized applications may seem "clumsy" or "not user-friendly" to us humans, but their characteristics of being "censorship-resistant" and "uninterfered" may be more important traits for AI agents.
Although I mentioned in previous articles that during earlier bull market cycles, we saw some decentralized applications that, despite not gaining traction, might find their value in AI agents.
But now, upon reflection, I feel I have underestimated the future potential of those decentralized applications.
The reason for underestimating is that the development of AI agents has only just begun; we have not yet seen a qualitative leap in the quantity and quality of AI agents. Therefore, in our subconscious, we still tend to believe that "humans" are the main body of the application ecosystem.
In reality, it is very likely that when AI agents surpass a turning point in quantity and quality, we will truly realize that "AI agents" are the main body of the application ecosystem.
If, as Jensen Huang said, the market for AI agents will be worth trillions of dollars, then by that time, AI agents will certainly far exceed humans.
At that point, the value of decentralized applications will be truly recognized.
The second reason for underestimation is that current AI agents have not yet been "choked" by centralized platforms. If one day AIXBT's Twitter account were banned for some reason, people would certainly realize the importance of a decentralized "Twitter."
While we are grateful that Phala's TEE environment allows AI agents to have their own wallets, this only solves a very small problem; many similar "choke points" are still controlled by centralized platforms. Perhaps there are already suitable solutions for these points, but they have yet to be discovered. More likely, there are still untapped areas.
If this round of AI agent development achieves good results, the next development in the crypto ecosystem will likely revolve around various decentralizations of AI agents. These decentralized applications will attract not only builders from the Web 3 ecosystem but also builders from the traditional Web 2 ecosystem.
The crypto ecosystem may soon welcome a large-scale breakout.
