Happy Sunday and welcome to InsideAI weekend! I'm Rob May, Founder and CTO at Dianthus. We build AI software for ecommerce marketing but, instead of selling it like a traditional SaaS company, we acquire ecommerce companies and apply it once we own them. It's a different model but really interesting, and we are hiring a ton of people, particularly around what we call "generative marketing." If you want to work for a rapidly growing, profitable, well funded tech company, we need AI people across almost all areas.
I'm taking the rest of the year off so this is the last newsletter of 2021. As a result I want to cover my AI predictions for next year. I'll start by saying that this is probably the hardest year to make predictions since I started doing this in 2016. The primary reason is that there just aren't a lot of really novel breakthroughs on the horizon, and the continued barriers to AI adoption are more workflow and cultural rather than technical. That said, I'll do my best. Here are my top 4 preditctions for AI in 2022.
#1 - Non-Neural Network AI Starts To Rise
I've been predicting this for 3 years, incorrectly, but this year, DeepMind showed that for many language tasks, bigger models aren't working much better, which will make more people realize we need a different approach to solve some AI problems. While there are people working on evolutionary algorithms and symbolic logic approaches to AI, there aren't many, and it isn't popular. Expect that to change in 2022.
#2 - The First Major International AI Incident
I'm increasingly concerned about deploying AI systems we don't fully understand, and I see small errors here and there, some with big consequences. I think this year the first AI issue that somehow straddles international borders will happen. It could be something with autonomous weapons. It could be one major country demonstrates an AI capability that scares the hell out of everyone else... something like that where countries actually think about uniting to stop some kind of AI success (or failure) by one country. (Hat tip to Rob Toews who has influenced my thinking here)
#3 - The Rise of AI Executives
AI, as it is applied at work in the context of traditional business strategy, is still nascent. But books like Competing In The Age of AI and The AI-First Company have laid the foundational understanding of how AI is different. Universities are starting to incoroprate some ideas into their courses, along with certificates and executive programs that teach how AI is different and how it impacts business. I think 2022 will be the year companies start to realize the value of hiring this expertise (as opposed to just AI technical expertise) and we will see it show up in job descriptions.
#4 - The Rise of Creative AI
AI has been tested in some creative domains but we haven't really seen a breakout that has an industry-wide impact. I expect to see that this year. Maybe in Art. Maybe in Intellectual Property creation. Maybe in Music. But creative AI will make waves in 2022.
As you can see, the predictions are more around applied AI than in previous years, because that is where I see things taking hold. What do you think is the biggest prediction I didn't cover? If you have an opinion let me know and if I get several, I'll cover them in the first newsletter of 2022.
Thanks for reading.
@robmay
Comments
Post a Comment