Code Reader: Etherscan’s new crypto AI tool for investigating smart contracts

Blockchain exploration platform Etherscan has officially launched the beta version of a crypto AI tool called “code Reader” capable of providing in-depth information on smart contract codes.

Through this function, the amount of complex information contained in various contracts in the crypto world can now be simplified.

Full details below.

Etherscan launches the beta of the crypto AI tool

Yesterday, Etherscan, a popular block exploring platform for the Ethereum blockchain, announced in a blog post the launch of the crypto AI tool “Code Reader.”

Initially available in beta version, Code Reader helps users get additional and detailed information about a specific smart contract code by going deeper into a specific section or function through the use of artificial intelligence.

Etherscan itself stated in an interview that:

“This feature is particularly valuable when interacting with a smart contract for the first time, as it provides a thorough understanding of its operations and functionality.”

To use the tool, it is necessary first to go to the official Etherscan website and, by opening the drop-down menu under “developers,” select “Code Reader.”

Once the tool is open you will simply need to paste in the code of the contract you wish to investigate further and specify the prompt, i.e. the precise scope of interest.

It is important to note that to use this feature it is mandatory to provide a valid OpenAI API key and sufficient usage limits.

After providing all this data, the artificial intelligence application will go to type specific information related to the source code of the selected contract.

Common use cases include capturing precise information and explaining it, obtaining complete functions, and explaining how the smart contract works within decentralized applications.

Etherscan, in a tweet in which it introduced “Code Reader” to the public, included a dutiful disclaimer saying not to assume that the answers are correct, as they are generated by OpenAI’s AI and not by the block explorer itself.

It was also clarified that the platform does not store its own API keys.

Despite the fact that technologies working in the field of artificial intelligence are making great strides, it is worth mentioning that current prototypes still need to be improved.

In particular, these need to address some problems with data synchronization, network optimization, not to mention dangers related to user privacy and security.

By now it is a foregone conclusion that AI in the next decade will be crucial in a multitude of different areas, yet it must be used with awareness and in moderation.

These kinds of tools are useful as a supplement to human ingenuity and intelligence: if replaced, they risk becoming extremely dangerous.

Smart contracts and AI: detecting scams

Etherscan’s AI crypto tool helps the Web3 community more easily explore different types of smart contracts by providing detailed information based on certain prompts.

There is no doubt about what functions a contract can operate or what applications it can run on, but the problem of detecting fraudulent smart contracts remains.

It is not yet clear whether Code Reader will be able to warn users that they are investigating a scam contract that contains dangerous functions that may endanger the funds of those who come into contact with it.

Juggling firsthand the preliminary study of contracts is not easy and subtends several hours of studying how they work.

On the other hand, it is not certain that AI platforms will be able to unearth all the red flags hidden in different smart contract cases, and especially whether they will be able to distinguish legitimate from fraudulent cases.

For example, there are mint functions that allow the contract developer to create additional tokens within their wallet.

Although this feature could be seen as a move to increase supply exponentially to the detriment of the community, it does not always conceal cheating.

In fact, many yield platforms, such as PancakeSwap, include this possibility, but it is dictated by the fact that new CAKE tokens that are “created out of thin air” are given to liquidity providers as an incentive for their own transactions.

Hence, how to distinguish a scam contract from a regular one?

Moreover, we do not know whether the most egregious cases of scam contracts will be flagged as such through notices to the user, as in the case of honeypots, which are particular tokens that can only be bought and not sold, except for the crypto deployer.

Unfortunately, at the moment the only weapon with which one can defend oneself is the study and knowledge of how the world of smart contracts works.

As mentioned in the previous paragraph, awareness of the key concepts in this context, combined with the conveniences provided by AI tools, are truly a game changer.

For less experienced users, if you want to verify that a token does not hide honeypot risks, you can consult this website.

In the meantime, study!