<aside> 🧟 Put your two questions here. Please include your name.
</aside>
How are AI chatbots like ChatGPT impacting computational text? In the chapter, it talks about how common text generation tactics do not have much concern for semantics, but do you think tools like ChatGPT are properly addressing these issues? (Brian Lau)
ChatGPT certainly seems like it has more 'semantic understanding'. Some people argue it has no idea what it's talking about. Some people argue that understanding is an emergent quality of the model.
I mostly look at it practically. It may or may not understand, but it acts enough like it does to be useful
Is computational text ethical? Especially programs that generate sentences / paragraphs that are grammatically correct, properly structured, and seem to make sense, but in actuality are completely fabricated? (Brian Lau)
I think 'computational text' is to broad of a topic to apply "ethical" or "unethical" to. I think certain uses certainly would be ethical, and others certainly would be unethical"
yes. there were systems that generated coherent sentences long before GPTs
markov chains can be used with anything that can be expressed as a string of symbols. It is common to use MCs with letters, words, and musical notes.
you could try generating images with MCs by generating an intermediate form, like SVG
I think the short, mostly correct answer is 'a computer is a discrete-state machine'
Turing mentions "Strictly speaking there, are no such machines. Everything really moves continuously." While true, this isn't something that really needs to be considered when using computers.