deep blue defeated gary kasparov in may 1997, now almost thirty years ago. deep blue was a program designed specifically to play chess and not considered “artificial intelligence”.
since then it has become taken for granted that humans have no chance against programs, “a i” or otherwise, in playing chess. i remember reading a statement by kasparov himself that “for a few dollars anyone can now purchase a program that can crush a grandmaster”. kasparov’s thoughts on the subject are readily available online. more recently, i read something to the effect that the biggest problem facing anyone trying to design computer chess games today was producing something that a human could beat.
here in the year 2024, it seems that a i , in producing “creative work” of any kind, and superseding human producers, might be at the same tipping point as deep blue was against kasparov in 1997. the time may not be long before a i programs can leave human “creators” as far back in the dust as chess programs have left human chess players.
in the last year or so, kindle direct publishing, amazon’s service for self-publishing, has added a question as to whether a i programing has been used by the author to produce the work being submitted. i do not know if a “yes” answer prevents kdp from accepting the work. probably not - at this point, they may just want the information for statistics,
thrillers, romance novels, fantasy novels, and screenplays (excepting the most arty) would seem the first in the line of fire. it may not be long before - the time may already be here - when an a i program can not only produce smoother, more palatable products of this sort but produce them instantaneously. recently hollywood screenwriters went on strike, and eventually settled with the studios, the studios apparently agreeing - for now - that they “needed” the writers. but for how long? it seemed to me that the writers were rushing into the fiery furnace and provoking the studios into speeding up the process of replacing them with a i.
what about “serious” writing? what about programs duplicating famous authors of the past - shakespeare, jane austen, dickens, whoever? in these cases i think it would depend on the audience’s acceptance. some shakespeare or jane austen buffs might welcome convincing pastiches of these old favorites. books of this sort are already produced, with no subterfuge, of authors both in the public domain and authorized by the authors’s estates. conan doyle is the most obvious example.
but as you get closer to the present it seems less likely that any audience would accept this sort of thing. the readers of “serious” writing want to identify with the author and apparently want to feel they “know“, or at least know something about, them. one evidence of this is the ubiquity of author interviews. another is the concern so many readers now have with “problematic” authors - authors whose opinions on political or social matters, or actions in their real lives, the readers view with horror or disapproval or contempt.
would these readers have any interest in “serious” writing by a i? my guess is that they would not. theoretically, publishers could invent “authors” for these works, but would it be worth their while, given the relatively low sales of such works, and the risk of backlash if they were caught out?
so - a likely foreseeable future is one in which most writing is produced by a i programs, but a small amount of non-genre writing - the kind of writing that wins prizes like booker prizes and national book awards - will be produced by humans - most or all of it by graduates of creative writing programs.
No comments:
Post a Comment