What Will Happen with AI in 2023

Quantified has been working with AI for 6 years, so we’re grounded in the technology and adoption. Here’s an outlook that will help you anticipate what you might be hearing in the coming year: 

1. You are going to see a dramatic change in leadership’s urgency to leverage AI
This new creative AI era will fuel massive advances in personalized customer service, drive new business models, improve real-time modeling (Southwest Airlines) and pave the way for breakthroughs in healthcare and biopharma from Google‘s DeepMind. Dear CEOs: Pay attention. This will be a competitive issue in the next 12 months. Make sure you get there first!

2. The media will achieve peak AI hype in 2023
A day won’t go by where some journalist isn’t discovering that #gpt3 can write an article for them (I’m already bored of it). Nevertheless, although it will be overhyped, AI is real, substantive, and game-changing (as opposed to the current state of VR, cryptocurrency, and the metaverse).

3. You are going to interact with generative AI applications daily (but you may not know it)
You already might not know whether a human or an artificial intelligence wrote a sentence like this one, with tools like #chatgpt. We’re already in the early days where AI can write with surprisingly human-like flare and produce digital human videos that look almost real. AI will shift into practical applications like Jasper.AI, GitHub‘s copilot, and sales simulations.

4. You will hear a lot about GPT-4 in a few months
Expect GPT-4 to be released early in the new year and to represent a dramatic step-change performance improvement relative to GPT-3 and 3.5. Researchers believe #gpt4 will be able to pass the bar exam. It might also be multimodal (images, text, speech, numerical data). 

5. You won’t see AI legislated yet. But you will see creators opt out of AI data sets
The EU AI Act aspires to create the first regulation for AI, but they can’t even agree on a definition of AI yet, so don’t expect anything this year. I hope they realize that explainability is the path to trustworthiness. Nevertheless, generative AI scraped from public internet data, including copyrighted text and images, will allow creators to opt-out of being included in the training data, and many of them will.