The road to hell is paved with AI flyers

The world lacks the technical and governance tools to deliberately pace frontier-wide progress –1,386 employees of frontier AI companies –  www.pacingthefrontier.com

• We are in a Jurassic Park moment. No guardrails

If you are uneasy about the pace of AI, you are not alone. And then there are AI flyers. A contagious rash polluting our feeds. 

We are at a critical juncture. More than a thousand employees of AI companies—that include Google, Anthropic, OpenAI and others—have taken a stand and called for a slowdown. Why? Because, to quote one of them, it is “just like a runaway nuclear chain reaction.” We are running out of time, they say, because they realise that the world “lacks the technical and governance tools to deliberately pace frontier-wide progress.” The organisation to which they are signatories is called ‘Pacing the Frontier.’ (See https://www.pacingthefrontier.com/)

AI flyers, which I will get to in a moment may seem trivial by comparison. But I believe it reveals the blind obsession with this shiny new technology. Between scylla and charybdis of content creation and instant design, these apps are hard to resist. But there’s a price to pay downstream. David Rubenstein, billionaire and co-founder of the Carlyle group recently observed that the age-old human skills such as reading and writing are what will never be overtaken by learning to prompt and summarise information using chatbot. Those meta-cognititive skills for critical thinking, and being able to communicate effectively are at risk if we outsource them to ‘smart’ AI tools. 

Cosmic AI flyer pollution

Which brings me to the quick and easy, down and dirty rash of AI flyers. I am sure you have noticed them. How they all look the same. Or at least how they use similar bland elements. Designers are horrified. They point out how these flyers use similar formulaic graphics: (a) Icons—a plethora of them. (b) Subtitles for any image or icon. (c) Very similar fonts—some kind of brush stroke for a heading. (c)  Overpowering backgrounds. The companies touting how to create an AI flyer aren’t even embarrassed to showcase their samples. The Church flyer shown here is just one example. See what I mean by the icons and the brush-stroke for one part of the headline? 

Unfortunately, even companies like Adobe have joined the bandwagon. Yes, the company we once considered the last word in design. The company who owned the two letters ‘AI’ which, at that time, meant Adobe Illustrator. 

We are in a Jurassic Park moment with AI running amok

Meet Tilly Norwood. A cartoon not a clone

British TV personality Piers Morgan is often controversial. In a recent show on ‘Piers Morgan Uncensored,’ his guest was a synthetic actor–a woman named Tilly Norwood. Created using AI. She He set it up with a ‘professor’ Jiang, (not a real professor) who was also asking questions of Tilly. She responded to questions but quickly got off topic. When Piers asked her about the brouhaha over whether AI characters may put Hollywood actors out of business she was candid, admitting that “I am closer to a cartoon than a clone.” But then she went off the rails fast, repeating herself and suddenly switching to Chinese. This seems the equivalent of your smart TV glitching mid movie on Netflix and suddenly playing a clip from an adult movie, when the children are watching. (That never happens, thankfully.) Is this simply funny? Or is it a commentary on our times when we not just get to watch a ‘meltdown’ of an AI creation, but have the media try to interview a non-person? And my point? Given the lure of AI, we are going from the ridiculous (AI flyers) to the sublime (embarrassing AI actors), and everything in between. 

Let me be clear. I do not doubt that there is a time and place for us to use AI, just like there is a time and place for us to use a notepad and pen, or a computer and mouse. But that doesn’t mean we could turn over everything we do in education, work, and entertainment to an algorithm and hope for the best. We have a huge asset with human meta-cognitive skills that Rubenstein referred to. Skills such as problem-solving, critical thinking, and ‘thinking about thinking.’ The synthetic variety may seem smart but possess none of these, despite what a handful of tech companies tell us. They, of course, want us to drink the Kool-aid.

In an upcoming school newspaper I publish I asked a math teacher to write a column on how AI might be used in mathematics education. She made a good point that there are many “low tech, high tech, or no tech approaches used in the classroom,” but homework does require students to master skills, and uses those (no-tech) skills for problem-solving before looking for the answers through a calculator (low tech) or an app (high tech). We can do this. Without capitulating to the AI bandwagon.

Sneaking in

The other day I noticed that Microsoft 365 was loaded on our school’s computers, and when my students opened the application I noticed how Microsoft practically hid PowerPoint, Excel, and Word. That was because the screen real estate was taken up by Copilot–Microsoft’s AI play. Our high school does not allow students to use AI to generate essays or presentations. (Even if this was not a school policy I would require students to wrestle with sentence structure, spelling and punctuation; layout and design etc.) Very quickly our IT department was able to give me a workaround for students. They loaded the three apps on the portal, so students would not be hijacked by the AI.

I am not alone on this. On a podcast, On The Media, about how chatbots are infiltrating the classroom, staff writer of The New Yorker Jessica Winter spoke of her personal experience. How her 10-year old child (using a school-provided Chromebook) was being enticed to use Gemini to ‘upgrade’ a presentation done for homework. The whole point of a child researching and putting together a document is to learn from the process, not to showcase how tech savvy he or she is using some swanky app. As teachers we expect the dog-eared essay, the one with a few missing capitalisations and punctuations. These become teachable moments. Google, Microsoft, Anthropic, and Meta don’t seem to get it. So will we pushback? You bet!

If we don’t, who will? This week the White House came out with an ‘Accord on Super Intelligence.’ This was a response to public outcry against the above four horsemen. It contained the ‘regulation’ jargon but was not even close. Basically, Sundar Pichai (Google/Gemini), Dario Amodei (Anthropic/Claude), Mark Zuckerberg (Meta/MetaAI), Greg Brockman (OpenAI/ChatGPT), and Elon Musk (XAI) agreed to the following: Trust us, we know what we are doing; We will audit ourselves, thank you very much; Over time, we may tiptoe toward regulation. If you read between the lines this was smoke and mirrors by the billionaires, for the billionaires. 

Not eggs, but algorithms

I will end with Jeff Goldblum’s warning which was prescient–a line in Jurassic Park: “Don’t you see the danger, inherent. You wield it like a kid who found his dad’s gun.” We are in a Jurassic Park moment with AI running amok. Substitute the Ricard Attenborough character with Sam Altman, and… they are playing with science. Goldman continues: “It didn’t take any discipline to attend. You read what others have done and you took the next step. You didn’t earn the knowledge for yourselves so you don’t take any responsibility for it.” The camera cuts to a close up of Richard Attenborough shaking his head at this sceptic in his team. Are you shaking your head at what I have just said? Or are you nodding with Jeff Goldblum?