Generative Systems (ChatGPT, DALL-E, etc.): Home

There is a great deal of information (and a great deal of confusion) about what ChatGPT, DALL-E, and other generative systems are, including their capabilities and potential dangers. Here is a quick guide:

In hearing about ChatGPT and the like, people will often use the term “artificial intelligence” to describe them. Something very important to note is that ChatGPT is not a true artificial intelligence, or at least cannot be called such with absolute certainty. The Stanford Encyclopedia of Philosophy defines artificial intelligence as “the field devoted to building…artificial persons (or at least artificial creatures that – in suitable contexts – appear to be persons).”1 Ascribing motives, personality, or “wants” to ChatGPT or any other generative language model is incorrect, and misses the point of what it is.

So what is ChatGPT, then? It’s a very sophisticated chatbot, using what is called a “generative pre-trained transformer,” or “GPT” type of language model to amass huge amounts of text and information from the internet, which is then used as the basis for its responses to text prompts. It has enormous flexibility, able to suggest recipes, write poetry, have conversations, or even speculate on hypothetical forms of methane-based life existing elsewhere in the universe. The possibilities for its usage in academia, business, or entertainment are incredibly vast.

However, it is important to note that ChatGPT, and any similar generative system, in spite of the incredible amounts of information and tools at its disposal, is not always right. ChatGPT will make up answers when it has none, is potentially prone to repeating racist and harmful language it has acquired over the internet, and has difficulty with abstract or ethical concepts. It is capable of generating speedy answers, but not necessarily accurate ones. If you intend to use ChatGPT or similar programs as a source of information, carefully evaluate the answers it gives you, and check the information with other sources already known to be legitimate on the subject. In that sense, it is not different than any other resource; think of it as a sort of “man on the street” interview, an individual who can provide an answer on virtually any topic, but one which must be weighed carefully, since it has been provided by a random person with a bias all of their own (yes, even ChatGPT can have a bias).

With all that in mind, let’s get started!

1: Bringsjord, Selmer; Govindarajulu, Naveen Sundar, "Artificial Intelligence", The Stanford Encyclopedia of Philosophy (Summer 2020 Edition),  https://plato.stanford.edu/archives/sum2020/entries/artificial-intelligence. Accessed 10 Mar 2023.