It seems like ChatGPT can do anything these days, from coding a website to answering some of your most pressing questions. One of these pressing questions is where its abilities stop. How far can it go? One of the biggest fears mankind has about technological advancement is technology overtaking humans. We can thank movies such as the Terminator franchise and I, Robot for that.

We won’t go as far as to presume AI self-awareness just yet, but can it do tedious and arduous jobs we don’t want to do like smart contract auditing?

Table of Contents

Chat GPT Explained

There is a lot of speculation going on that Chat GPT and AI, in general, are close to replacing many humans in everyday jobs. The potential industries being hit are creative or cerebral such as engineers and graphic artists.

Chat GPT is a language model that has been created and trained by OpenAI with large amounts of data to understand our language and perform many language-related tasks. Not only that, but ChatGPT can also interact coherently with users due to its learning capabilities.

What is a Large Language Model (LMM)?

What is a Large Language Model (LMM)

Within AI and language processing, Large Language Model, or LMM, describes exactly what ChatGPT is, a language model that is capable of understanding and processing large amounts of natural language data and responding to queries, providing suggestions with coherent language.

What type of language data do LMMs process? It can be anything ranging from books, texts, news – any online content. Due to these capabilities, we see an uptick in LMMs in recent years such as ChatGPT versions and BERT (Bidirectional Encoder Representations from Transformers).

These LMMs have been evolving and adding capabilities to their repertoires such as language translations, and can generate various texts in different tones of voices, which paired with their language learning abilities makes it amazing for developing chatbots and virtual assistants. LMMs can also improve search engines, code websites and programs, and help you make strides in content automation.

Can AI Detect Smart Contract Vulnerabilities?

What can Chat GPT do within the crypto industry? Smart contracts are programs on blockchains that self-execute when terms have been met. The primary purpose of smart contracts is to carry out contract terms for parties involved without intermediaries.

While smart contracts are immutable, they can sometimes have vulnerabilities. Since LMMs like ChatGPT are capable of improving and developing existing programs, can they help detect vulnerabilities within a smart contract?

At this point in time, yes, they can definitely help with manual audits programmers have to carry out to avoid security issues, but they cannot completely replace humans. Will it get to a point where an AI tool can completely replace us? We hope not.

As of now, ChatGPT provides invaluable insights and recommendations that can help our engineers and programmers create foolproof smart contracts for blockchain security.

ChatGPT VS. Humans

ChatGPT VS. Humans

(Image by frimufilms on Freepik)

It’s already possible for humans to have a coherent and maybe even insightful conversation with ChatGPT’s latest version. At the moment, the most dangerous thing that can stem from the language model is misuse. Its content could infringe on copyright laws or be used for impersonation, but that is to no fault of ChatGPT itself.

Academically, ChatGPT can give humans a run for our money in many aspects such as response time and access to a wealth of materials and knowledge we don’t have readily in our minds. However, ChatGPT fails in real-world scenarios where situations require a human touch.


LMMs and AI took great steps in the past few years and have now become incredibly capable. The ability to code, audit and automate content and programming language is just the tip of the iceberg. The evolution of AI and language models have no foreseeable limitations right now, so we can only hope that its future involves working with mankind toward a better world rather than the alternative.

Related Articles:



Leave a comment

All blog comments are checked prior to publishing