Video Meetings as seen by AI
Meme goldmine Dall-E mini is now Craiyon. It is an online AI picture generator. Despite a backlash forming (over wannabe Big Tech IP thievery), first contact pulls you in.
You tell it what you want. Cogs whirr for a minute or two. Then serving up nine images it has created based on your creative command.
You can literally conjure anything.
Most requests of those losing an hour of their life to discovering this toy/vital-tech [delete as preferred] fall into broad categories.
A pop culture icon in a crazy setting. Mashups of two things never before seen together. The deliberately abstract, trying no doubt to 'fool' the artbot. And if not 'photorealistic', then all in the style of a famous artist or artistic style.
The obvious point to state, is that for a concept you want to get across in a slidedeck, showing such output may well make for a memorable slide.
Along the lines of, 'what would AI make of _____?' 'Well, let's see...'
Add, enhance, shuffle to your hearts content. Well, maybe with one follow-on slide perhaps.
Given my current day-job, I couldn't help but provide the assuredly tame instruction as seen up top; video meeting set up in pop art style.
Yet what I was presented gave me as much the shudders as AI running amok in sci-fi dystopias.
Is this really how semi-sentient zeroes-and-ones see it?
And of so, what does it say?
By this, I mean that the flat, static, homogenised nature has seemingly set in.
There is no vibrancy at all. Let alone flare or imagination.
AI seems to think staged, limp stock photos are real-life.
Each pic - whether for hybrid or remote - shows inactive, passive attendees.
If you tot up say twenty-six screenscapes on show, maybe only two or three such zoom backgrounds show anything other than a single coloured wall backdrop.
That's a lot of hostages.
Not a plant in site.
And certainly - inexplicably - a shared screen graphic nowhere to be seen.
Even more galling, is the motionless of the participants when judged in context that not a single hand gesture is on display.
Finally, you clocked the, ahem, hereditary of each face?
Maybe craiyon does all its calls around the Baltic...
But here we are focused on anyone's selling video meeting performance for better outcomes.
Imagine if someone showed you a grid of nine pics. Each taken 60 seconds on from the previous, for a recent ten-minute meeting over video of yours.
How would that look?
Without resorting to snap-in apps, you can and should deploy a chart, diagram or image. One which can easily stay on view for a minute.
Not forgetting the judiciously deployed prop. The physical pen-and-paper, model-cum-artefact example, or printout for annotation.
And naturally, the expressionful participant actually moving and not looking gun-barrel straight down the line.
It's a start, hey. Take that AI, and your future attendees will thank you for it.