ChatGPT — the AI chatbot powered by OpenAI — finds use in numerous fields.
It has further revolutionized the place of Artificial Intelligence in the technology space, so much that a crypto exchange executive has tested its capacity to detect flaws in Ethereum (ETH) smart contract.
In simple terms, ChatGPT is an AI model that engages in discussions. It is a chatbot similar to the automated chat services on some companies' customer service platforms.
Developed by OpenAI — a tech research company — ChatGPT is an upgrade on other chatbots. This is because it was generated with the help of high-level human trainers and reward models.
To that end, it can dispute false inferences, admit errors and respond to follow-up questions.
More than just a question-and-answer bot, ChatGPT can be used to create content for multiple sources, translate text from one language to another, write computer codes, etc.
Now it has a new version, GPT-4, capable of handling pictures alongside words and much more. But how 'much more' can it handle exactly?
Conor Grogan, director of crypto exchange at Coinbase, inserted a live Ethereum smart contract into ChatGPT-4, and the effect was astonishing. He said,
"It highlighted a number of security vulnerabilities and pointed out surface areas where the contract could be exploited. It then verified a specific way I could exploit the contract."
He further posted screenshots of GPT-4's analysis which proved that it could accurately point out pertinent issues that may exist.
It deduced that the smart contract "should not be used, as it contains critical vulnerabilities and is based on an illegal scheme."
All of these did not come without controversy, though, as some people offered dissenting views.
In the comment section of Grogan's post, a bit of a disagreement brewed. Some users claimed that GPT-4 did not uncover these flaws and that they were already attainable online.
Interestingly, Grogan had earlier affirmed that the ETH contract was hacked in 2018, with the same vulnerabilities that GPT-4 pointed out being mentioned. This made them say GPT-4 re-outlined them, as they were already public knowledge.
In addition, some smart contract audits on ChatGPT previously found that the first version could also spot any irregularities in codes.
In 2022, Stephen Tong of auditing firm Zellic asked ChatGPT to uncover flaws in a smart contract code which was done successfully; a similar thing was also done in May.
Irrespective of the fact that there was prior information online or if ChatGPT v4 figured out the vulnerabilities, there's no gain in saying that its functionality is impressive. It is improving the world one sector at a time while being relatively accessible.
Conversely, some tech insiders and critics have indicated concerns that ChatGPT might be biased.
One such is Elon Musk, who thinks ChatGPT is 'woke' and has joked about the need for an alternative, "TruthGPT." Others feel the chatbot leans more to the political left, which does not set a good precedent.
Disclaimer: Voice of Crypto aims to deliver accurate and up-to-date information, but it will not be responsible for any missing facts or inaccurate information. Cryptocurrencies are highly volatile financial assets, so research and make your own financial decisions.