The Future: Consumer Computing
OpenAI announced ChatGPT Multimodal.
I heard the news through a rather breathless podcast about the drop and was trying to find my way out of the hype machine, but I think this is actually a big deal.
Previous AI tools were really just supercharged versions of existing tech options. ChatGPT was autocomplete on steroids. DALL-E was reverse image recognition. AR filters were Hipstamatic 2.0.
But this new GPT4 feature is something more.
It's a new way for users to interact with computing devices in ways more natural to the history of human experience and development.
Here's one of the examples from the announcement:
If I were to ask someone to help me face-to-face, I would point out the bike and ask (just like the demo). I wouldn't give them the specific bike make and model as a preface to my question.
When I'm trying to get a question answered, I'm looking for an answer. I'm not specifically looking for a list of links to click about or a YouTube video to scrub through or a PDF to scroll through.
This is the closest analog of human-to-human information seeking interaction I've seen from tech outside of limited voice assistant stuff.
We're moving closer to ambient computing—when the screen gets deemphasized and devices act more as extensions of your person than external boxes you tap at. AirPods, smart watches, and voice assistant devices are the forerunners of this trend. Multimodal is a power up.
If true AR glasses are feasible (big "if" on the technical hardware end of things right now), think about not even needing to take a picture for that bike example. Simply look at your bike and ask for help lowering the seat and get instructions overlaid on your field of vision.
We as marketers need to start thinking through how we can best serve customers through experiences like this in the future. And how our methods of distribution and promotion will change.