Put Generative A.I. In A Daycare
Rather than a pause on Generative A.I. we should put them in a digital daycare with some adult supervision. Let's have some fun.
They’re probably one of the funnest, most valuable experiences in our lives; daycares. They’re also very germy with lots of sniffles, drool, hugs, sneezes, wheezes and ouchies. Toddlers don’t have gender or racial biases either. They are where we play and learn about one another. We move fast and break things and it’s okay.
Before I roll into this short essay, let me be sure to say that I am a proponent of AI. I think it can benefit humanity. Done right. It will be messy. I’ve been deeply embedded in the world of technology and humans for over 25 years. I’ve seen some really good things. I’ve seen some silly things. I’m a pragmatist and an anthropologist focused on being human in the digital age.
It is also important to recognize that there are some very good AI uses underway. Some are being used for drug discoveries, others for combating climate change and others for medical advancements.
While the AI tools such as Generative AI are presenting some exciting opportunities and of course risks, how these tools have been thrust into the public sphere has left a lot to be desired in terms of stewardship. Lot’s of altruistic statements, very little adult behaviour. These are of course amplified by the OpenAI debacle of governance and the antics of X and Grok, who’s claim to advancement is the ability to use foul language.
Many thought leaders in industry and academia called for a 6 month moratorium to consider how we’re advancing AI. This turned out to be more of a play by industry leaders to build moats around their AI for their Lego fortresses. Nobody likes anyone to smash their Lego bricks fortresses.
So perhaps, rather than a pause, we could put all these companies in a digital daycare. Maybe in a neutral country like Switzerland. So China, Russia and others can all get in on the playing. And when they don’t get along, they can all have nap time. And enjoy some nice fruit snacks afterwards. Not too much sugar though, they might break too many things.
The social media giants, lead by then business toddler Zuckerberg, exemplified the mantra of the daycare; move fast and break things. Knocking down your buddy’s Lego fort is a lot more fun than building it. The other kids in the then social media daycare liked this idea too, so off they went breaking things. And when they got caught, they all pointed the finger at the other guy. And sobbed a bit for good measure.
Trouble is, the playground was the public sphere. And they played hard and the harder they played, the more things they broke. We shouldn’t really blame them though. The adults in the room were too busy doing other things like seeing if chewing gum might stop the smoke stacks from polluting.
Then one of the kids in the social media daycare made a lot of other, real kids cry. Now the adults are suing the social media companies who grew up to be hormonal adolescents doling out dopamine in the corner of the schoolyard.
That crop of social media kids has been replaced by a new generation. The A.I. kids. And they too like the move fast and break things game. They learned from their older siblings. Well, not really, but it sure did look like a lot of fun. They seem to have missed the adults in the room holding the social media kids accountable. But hey, every new generation always thinks it’s much more special than the last.
In some ways it’s almost as if the Venture Capitalists and the now wealthy, spoiled brats of the social media age just mainlined the A.I. kids with a sugar and caffeine drip then let them run rampant in the backyard of society….be sure to come home for dinner kids and don’t hurt anybody!
The AI kids are so hopped up on the sugar rush that some are overloaded and screaming that the sky is falling in and the other bunch is yelling at the adults telling them their world is over and time to get with the amazing changes. Everyone else is watching TikTok videos, the dopamine must flow!
This is why the AI industry needs a daycare. A friendly place they can go with a bit of adult supervision while still being able to build, and break, things. Except they don’t break things that hurt people when they’re in a private sandbox. We should’ve learnt this from the social media thing. There’s a reason human trials for drugs are supervised and food production is too. Things we stick in our minds can be as dangerous as things we stick in our bodies.
Since AI is very much a Cloud thing, we certainly could build a Digital Daycare for AI. Where they could all bring their Lego sets and Play-Doh, drink some caffeine beverages, eat some candy bars and have at it in the digital sandbox. Need some test subjects? How about synthetic data? Or some tech bros?
This however, is unlikely to happen. AI is a geopolitical technology and it can (already has been) be weaponized. Let’s face it, if I can knock your Lego fort down with a steely gaze from a hundred metres away, I will. While that’s a bit Hobbesian, well, daycares tend to be that way at times.
Humans have long played fast and loose with new, revolutionary technologies. While sometimes that’s been bad, it’s more often been good. We need technologies to survive as a species, it’s how we advance and learn.
When it comes to Artificial Intelligence, many of the AI tools that have been around for decades, from Natural Language Processing (NLP), Machine Learning (ML) and so on, they’ve largely been governed reasonably well. With the exception of a lot of gender and race bias, which has cost people jobs and rights. But we’re learning.
But so far, the roll out of Generative AI tools like Grok, Claude and ChatGPT, seems to be run by petulance and ego. Their behaviours seem to run opposite to their pledges of altruism and social responsibility. We need some adult supervision in the room.