Where A.I. is heading in 2024
I rewrote that headline a few times, switching between “My AI predictions,” “Where AI is going,” and “Where AI is headed.” I went with headed because it felt satisfactorily ironic. AI doesn’t have a head. Or a heart. No matter how human-like it sounds or behaves.
Who am I to make predictions?
I’m no soothsayer, insider, nor expert in the field. My information sources are dependable while limited. All I’ve got is that I’m anxious enough about the existential threat posed by unfettered AI development that I do a lot of reading, try to keep up with what’s happening, and think about this a lot, maybe more than you do if you’re well-rounded and/or have friends and/or are not prone to generalized, unproductive anxiety.
I’m paying most attention to AI Large Language Models (LLMs) and chatbots because that affects writing and writers the most. I haven’t spent a lot of time on DALL-E or Lumen or other visual art apps. As far as my predictions go, I feel like they’re easily transferred to the hopes and concerns of artists in every medium.
Here’s what I think we can expect to see in 2024, when it comes to the general public’s use of chatbots.
We will have evermore LLM choices.
Since ChatGPT shot out of the gate in November of 2022 it has been joined by more and more versions and providers of LLMs, both paid and free. I started on ChatGPT and quickly subscribed to Jasper, willing to pay for a service curated specifically to help produce marketing material, content and copywriting. Then I quickly unsubscribed because it was terrible (My AI is being an A-hole). Right now I have Poe.com bookmarked, which lets me choose and use evermore free versions of ChatGPT, Claude from Anthropic, Llama from Meta, Mistral, plus image-making AI and more. Chatbots are now allowing users to upload (or cut & paste) longer and longer text into their prompts. So far, I still find it labour-intensive to use AI to generate useful copy. The responses are lightning fast but obviously bot-generated. We can discuss chatbot quality and usefulness over dinner; what I know is, they’re gonna keep coming.
Monetizing, rather than helping, will motivate AI development.
From what I’ve read, this was what drove the whole Sam Altman kerfuffle. Altman was CEO of OpenAI which originated in 2016 as a non-profit. In November ‘23, the board at OpenAI called him up and kicked him out. From this and other NYT articles, my takeaway is that he made some big money deals to help fund development, including with Microsoft, and maybe turned a blind eye to some of the non-profit’s guiding principals. 5 days later, OpenAI welcomed Altman, and Microsoft, back, with a restructured board and a focus on “improving governance.” Nobody knows what really happened inside that boardroom, but I believe this article’s headline: AI Belongs to the Capitalists Now. (Sigh. Doesn’t everything?)
The doomsdayer vs techno-optimist divide will widen.
Techno-optimists will keep raving about the good AI can unlock, while doomsdayers rant about the existential threat to humanity. Believe it or not, I’m really trying to stand in the middle. I’m protective of our creativity as a species, and wonder what will happen if we offload too many of our reflective, hard-thinking tasks to AI so we can avoid the discomfort. Meanwhile, I love hearing about how AI is helping people brainstorm, take notes, transcribe, and work out their ideas. It’s the same way I feel about GPS. Did I love reading maps? Yes (and I relished the puzzle of refolding them perfectly). Is GPS easier? Yes. Is it foolproof? No. Do I rely on it too much? Probably. Does it keep me from arguing about the best way to get across town? Never.
There will continue to be bad actors.
The potential for AI-enhanced crime is overwhelming: super-powered scams, spam, deep fakes, plagiarism, phishing, impersonation, baiting, political manipulation… there’s nothing AI can’t do at lightning speed. The speed thing is dangerous because we can’t keep up. It’s one thing for our institutions to stay a step ahead of bad people. It’s another to try to manage the hyper-replication, proliferation and evolution of crime by a generative super-intelligence. But then again, maybe the robots will end up scamming and phishing all the fake accounts they’ve created to produce a whole eco-system of meta-scamming, and just leave us humans the hell alone.
There will continue to be human creators.
For every Create 6 months of content in an hour with AI ad I see, I can name a writer, creator, marketer, communicator, journalist, screenwriter or author who is encouraging their people to nurture their own creativity and voice. Like me! There is still something to be said for the slow process of working out your own words, sitting through the discomfort, overcoming the resistance, and writing / painting / knitting / doodling / dancing / playing out your ideas. There is still only one species and one kind of intelligence that can share in the human experience. Let’s make sure we keep talking about that in 2024.
Every article is written by a human which is me.