The Humans Behind the Digital Curtain
Turns out there's a lot more humans behind AI than we've been lead to believe. This is actually good for AI and it's role in our society.
There’s been a fair bit of kerfuffle over the past year or so that Artificial Intelligence (AI) in various forms, is coming for all our jobs. That most of will live off of a Universal Basic Income (UBI) or we’ll all move out into the woods and go off grid growing veggies and things. Perhaps some day. But that day may be a little ways off.
Turns out, there’s more humans behind the AI curtain than we’ve been lead to believe. So what does this mean for the future of AI and it’s role in our sociocultural systems? Let’s explore a little.
There are already some excellent uses of AI tools in industry, healthcare and other aspects of our society. Many of these will get even better in the coming years. One AI tool that’s proving a bit of a sticky wicket however, is Generative AI, like ChatGPT, Claude, Gemini and Midjourney. Multi-Modal. While GAI can be of value in various ways, it’s not quite meeting the hype.
That’s actually par for the course with many revolutionary technologies. Our cultures react in different ways and often, the technologies that are hyped take a little longer than we think to become truly effective for human use.
In recent months, we have come to see that not all is as it seems with AI tools across several well promoted scenarios. Amazon’s Just Walk Out stores it was revealed needed a lot of humans watching humans in real-time to do what the AI was supposed to do; check out customer purchases.
GM shut down it’s Cruze automated car service because humans had to interfere with the autonomous vehicles every .05 miles. A wee bit inefficient. Drive-thru automation company that purportedly used AI for order taking at fast-food restaurants revealed through SEC filings that humans had to help with over 70% of orders with workers in the Philippines.
Then there was Expensify, which claimed that AI could do all your expense reports automatically. A wonderful idea. Turns out that they used a team of technicians behind the scenes to do all that report filing. There’s a great list of ten of these cases here.
In the delightful story of the Wizard of Oz, near the end we discover that the all-seeing mighty wizard is in fact a benevolent old guy hiding behind a curtain. It is a great metaphor for where we ar with many AI tools today, not just GAI.
Things Are More Complicated Than We Realise
Thinking in complex systems is easy for some people, but they’ve usually spent many years learning how to do this. For many of us, we struggle with second and third order thinking. Then there’s the whole thing of thinking fast and slow as put forward by Dr. Kahneman.
We reference our world through the framework of the cultures and societies within which we are raised. If we remain in the sociocultural system in which we were raised, it’s hard to fathom other parts of the world. Or just how complex our world has become.
It is one of the reasons that when a new revolutionary technology comes along we can get all out of sorts. Some adopt the new technology and start evolving it in very interesting ways. Others are fearful and reject a new technology. It’s always a bit messy. It’s how we humans roll.
All Artificial Intelligence tools, and there are several, are very good at doing one thing. They are all what’s termed as Narrow AI. When done well, after lots of training and coding and mucking about with algorithms, they can become exceptionally good at doing a particular job.
Humans on the other hand, while some do train to be highly specialized at a particular job, like a heart surgeon or an architect, tend to be able to quite easily move across different skills. We are also highly social and communicate in various ways to collaborate.
In knowledge work, GAI is proving to be useful, to varying degrees with some jobs but so far, isn’t replacing jobs. Because very few jobs outside of manufacturing, can be fully automated. Why? because humans tend to do a lot of different things in their jobs. Read most any job description on a recruiting website and you can see why.
We humans love to simplify just about everything. Simplifying things gives us comfort and a degree of security. While we know far more today than we did a century ago about human behaviours and psychology, we are still learning, re-shaping and turning over age old assumptions. And our societies are growing more complex than ever before in our history.
Where AI Will Actually Be Useful
We are realizing today that we’re a really complex species. While computer scientists and engineers take the approach that everything humans do can be distilled down to algorithms, this not quite turning out as planned. Which is actually quite wonderful and interesting.
AI tools will improve. Some jobs will be replaced. It is, as has been before with revolutionary technologies, more likely that new jobs will be created. This will however, take a bit longer than some have predicted.
We’re in more of an experimentation phase with most every AI tool right now, especially Generative AIs like LLMs and RAGs. If we look beyond the hype, we can see that not much has actually changed over the past year. That many companies are finding implementing AI into their workforce is a lot harder and more expensive than initially thought.
As reality sets in, industry will start to take a more critical look at AI and will be more demanding that it prove its value. Wizards behind the curtains will not be acceptable when shareholders demand to know what’s really going on.
Slowly, we will, as we’ve done before, figure out where AI fits and how to make it work. For now, jobs that are highly automated, very narrow in focus and require minimal complex thinking are the ones ripe for automation. But that’s actually very few jobs in the bigger scheme of things.
Think In Complexity and Systems
If you’re worried about AI, step back and start to think about complex systems and systems as a whole. Then remember that we’re a rather quirky species and that we actually know very little about ourselves, our world and our universe.
When you hear proclamations from the industrialists in the AI world, take them with a grain of salt and understand why they make such grandiose and defining statements. When a company says it’s doing amazing things with AI, pull back the curtain, look for the wizards doing the actual work.
The really interesting and exciting advances in AI are more often being done behind the scenes, quietly and with patience. Often in boring industry sectors but that will have the most benefits for human society.
Apply critical thinking, use mental models, think slow, not fast. Understand how systems work and constantly evolve and remember that the ultimate arbiter of all technologies we humans love to create, is culture. That we are complex, messy and always do unexpected things is what makes us interesting and the future unpredictable. For now, we are still humans behind the digital curtain.