There is a technical mismatch in numerical criteria here with the tweet reading “most” but your description reading “25%” or more of a decrease. Furthermore, the most recent comment between you and EY emphasizes “half” have left the profession and you agree to that.

Please reconcile your resolution criteria and edit your market description accordingly.

If there's a lot of postings offering to pay $10 for a prompter to do the work that an experienced artist once charged $500 for, and half the people who once drew things by hand have left the profession, I think that should count as a win even if there's still lots of "job postings".

@EliezerYudkowsky I’m fine with that.

MIDI didn’t make live music performers obsolete, and I don’t think that our current technologies will make translators and artists obsolete either. With art especially, a strong understanding of the human experience writ large is needed to make truly impactful works, and models trained on the digital slurry that is the internet don’t have that deep understanding.

Of course, the real question is “how much labor will be replaced?” The thing is: even if the capabilities of AI increase, the labor displaced might not increase at the same rate. Besides what I said above, there are several other reasons some people have for not using AI. Those reasons won’t suddenly go away if stable diffusion becomes a thousand times better.

If you're going to base the resolution on what the high profile artists and organizations say, you're likely to conclude that the tweet won't hold up, because the well-known artists are the least likely to be replaced, and the organizations are more likely to talk based on ideology ("the evil machines can never replace our spirit!")

Instead, I suggest this: right now, pick 10 or more little artists. The kind who offer on twitter to draw your character for $50. Contact them now, ask them to take note of how many commissions they are getting now and how much pay, and then compare with how it is in a year, then in two, see if anything changes. Or, if not contact them, at least watch them - are they still around and kicking in two years, or gone belly-up?

@Tasty_Y Good idea, but I think there are lots of non-AI reasons that little artists might go belly up. It will be hard to objectively quantify if AI is the result. Also, I’d imagine the churn rate for newbies is higher than the churn rate for experienced professionals, right?

Professionals on movies and games and such won't be replaced. But "most" numerically... the people that do marketing ads, logos, freelance character design, random drawings of people's art, translators for manuals/guides, etc. I think ad-hoc commission work will not be a good career.

But senior roles and "technical artists" will be very much in demand: A very proficient artist will be able to replace a dozen artists while keeping everything consistent. Localizations will need someone in charge of making the whole translation coherent and with a distinct personality, even if lots of pieces can be machine-translated.

I think it will mostly be hard to get an entry-level job.