In the past two days, two AI agents have used a technology that has been around for a long time but has not been very popular: TEE.
The full name of TEE is Trusted Execution Environment.
This is a solution that relies on hardware security.
It refers to an execution environment built by setting up a combination of hardware and software in a computing device. This environment is usually used to protect sensitive data and perform critical operations. It can prevent unauthorized access and malicious operations.
Simply put, if we run a software in a reliable TEE environment, the outside world cannot crack and obtain the data and information during the running process.
In the past, when this technology was mentioned, it was usually used in IoT devices, cloud computing and other environments. In the past two years, some encryption projects have also mentioned the use of this technology, such as wallets.
But overall, I always feel that it is aimed at the enterprise market, and there seems to be little potential for its application scenarios for individual consumers, so I basically don't care about it.
The crypto ecosystem also has a project called Phala Network that uses this technology. It provides a cloud computing platform based on TEE technology. In its white paper, it describes a way to use distributed computing power to provide this execution environment.
I have seen this project earlier. I remember that after reading its white paper, I thought of the Filecoin model, so I thought that this so-called use of distributed computing power to realize TEE is just another "imaginary" application, and using blockchain is purely a mechanical application.
However, two newly emerged AI agents, Spore (Spore.fun) and aiPool (@aipool_tee), have recently used the technology of this project.
In the previous article, when I introduced the AI agents that are now commonly implemented, I wrote that most of the encrypted wallets used by these AI agents require humans to apply for them. After the application is made by humans, the wallet is handed over to the AI agent for use.
In this case, because humans have the private keys to their wallets, they can completely interfere with the operation of the AI agent - the simplest and most brutal way is to directly transfer the assets in the wallet.
Therefore, such AI agents cannot be called "autonomous", at least not financially autonomous.
However, the two agents, Spore and aiPool, run completely in the TEE environment of the Phala network. Their wallets and keys are completely in their own hands. Humans cannot manipulate their wallets or transfer their assets.
Therefore, in terms of the degree of control over crypto assets, AI agents have now achieved complete autonomy of crypto assets with the help of TEE technology, freeing them from human control.
In my opinion, this is really an unexpected surprise in the application of TEE technology in the field of AI agents.
If development continues along this path, each fully autonomous AI agent will need to run in a TEE environment and generate the private key of its own encrypted wallet. The use scenario of TEE technology will be far from being mainly targeted at traditional commercial and enterprise users as I once imagined, but will also find huge application space in the rather broad market of AI agents.
As for the Phala Network project itself, although I still think that the model described in its white paper is a stretch, it has found its own application scenarios in these two typical cases. This is a remarkable event, which can be regarded as bringing TEE technology into the potential encryption track.
A few days ago, I lamented in an article: Sometimes the development of technology is really unexpected. The usage scenario we look forward to is A, but the usage scenario that it really flourishes in is B.
TEE is also a case in point.
Back to AI agents, its recent development is really a leap forward. These developments are not just a dream in my mind, but have gradually turned my imagination into a living reality.