Artificial General Intelligence is a term that most of us have heard, a good number of us know how its defined, and some claim to know what it will mean for the average marketer. Here’s what OpenAI’s Sam Altman said “It will mean that 95% of what marketers use agencies, strategists, and creative professionals for today will easily, nearly instantly and at almost no cost be handled by the AI.”
What nobody knows for sure is when it will be here. Some said that GPT5 would herald the dawn of artificial general intelligence.
This episode is airing In mid-2025, and GPT5 has come out…and it is not widely believed to have AGI.
Our guest says AGI is a long way off, and more importantly, that it might not be the sought-for milestone we need for AI to be a revolutionary force in our lifetimes. Today’s guest takes us through what it will take for AGI to truly arrive. We also talk about public vs private models, Mixture of Experts (MoE) models, the Branches of AI like Foundational vs generative, Agents and Agentic Workflows.
Today’s guest graduated from DePaul with an MBA, has headed the AI/Analytics groups at (EY) Ernst & Young, Gartner, CSL Behring and now at the Hackett Group.
He has written several books and is here to talk about his 5th which came out in 2025.
So let’s go to Chicago now to speak about “The Path to AGI” with its author. Let’s welcome back for the 4th time on this show, more times than anyone else, John Thompson.
Hey, Glenn here. It’s the middle of summer when I’m recording this; a time we don a pair of shades, a beach towel and a good book. Funnel Reboot usually shares talks with marketing book authors, but for this show I’m going to share some reads that go a little farther afield.
Come along with me through six books that are all amazing. The subjects range between business, humanities, technology and science fiction.
Sometimes, to reach a solution, we must take unfamiliar paths.
In the early 1940s, a brilliant mathematician named Abraham Wald left his homeland in Hungary fleeing the spectre of war. He moved to the United States, and became part of a team at Columbia University tasked in 1942 with an aspect of the war where the Allies were losing badly to the Nazis. It involved the many Allied planes that would leave from England but never return to their bases, having been shot down somewhere over Europe. These B‑17 and B‑24 bombers had 10-man crews, weighed up to 30-32 tonnes, had wingspans of 100-110 feet, and were defended by machine guns planted along the plane’s entire length. Despite all this, they would lose planes every day, presumably because they’d taken enemy fire and either crashed during their campaign or as they headed back over the English Channel.
Wald’s team had to determine how to minimize bomber losses. They had been poring over aircraft returning from missions, mapping out the distribution of bullet holes across their fuselages. Their plan seemed logical — reinforce the areas with the most damage. But Wald saw what others missed.
Wald realized their sample set of data represented the survivors — the aircraft that had taken hits and still managed to return safely. There were other planes they weren’t examining, ones at the bottom of the channel or in occupied territory, that didn’t make it back. This lack of data could be biasing them to look at the problem backward. The planes they couldn’t sample could have been struck in areas that were more critical. Maybe the fact they were hit in those vulnerable spots was the reason behind them crashing and that the lack of damage in those spots on the surviving bombers simply meant they’d been lucky! the returning planes weren’t the rule, they were the exception.
Having flipped the problem around, the planes received reinforcements where the damage must be catastrophic, and from them on many more B17s and B24s completed their missions, helping the allies to victory in Europe. Some people call what Wald showed intuition, but that’s not what saved the allied bombers. Even though his approach seemed counterintuitive, data guided Wald to the solution.
This is Funnel Reboot, the podcast for analytically-minded marketers. Today’s episode goes outside our comfort zone, showing statistical tools in the hopes we’ll get a bit more comfortable using them.
Our guest today is someone who uses the same kind of critical reasoning – and statistics – to make sense of their product marketing problems. He is both someone who implements analytics tools, having configured over 500 sites, and one who posts prolifically about what he’s learned. He has also taught analytics at several New York colleges, and speaks at regional MeasureCamp events. After earning his MBA from Pennsylvania Western University, he spent about 20 years in corporate analytics. Then in 2017 with the support of his wife and three daughters, he set up his own firm, Albany Analytics. Listen now as he teaches you some tools that might help in your own marketing programs.
We as consumers do a lot of things just because the people around us are doing them. For proof, look no further than some historical examples—from the 17th-century tulip bulb craze in Holland to doomsday cults and prepper movements in the lead-up to Y2K. Buying fads such as pet rocks, fidget spinners, Beanie Babies, and NFTs all show how easily prevailing thoughts influence individual behavior.
The science behind this is well understood. The evolutionary drive to fit in with our peers is very strong. When a group of people’s purchases are plotted as a histogram, we always see the majority of them clumped near the centre – we see it so often we came up with a term for it – the Bell curve.
So even when people think they are expressing themselves, showing individuality by their brand choices, they are only veering slightly away from the norm.
Hey, Glenn here—welcome to Funnel Reboot. Our guest today—who I really do think has positively impacted marketers’ careers—argues that marketers are just as susceptible to conformity as consumers are. We get caught up in prevailing marketing practices when doing our job, while ignoring better marketing options. That’s a recipe for mediocre results.
Our guest is the author of three marketing books and the co-founder of an eight year old digital agency that has attracted clients whose annual spend ranges from thousands to millions of dollars. What does he credit for this marketing success? The time he’s spent on the edges of the Bell curve – doing things that most of us view as too far outside of our comfort zone. And he says to be a better marketer, you too should reject the orthodoxy of conventional marketing.
Unorthodox is the name of his latest book, and I’m glad to welcome back for a second time, Gil Gildner.
Most of the leading AI companies tell us how wonderful their technology will make our lives. In a recent post put out by OpenAI’s head, Sam Altman called The Gentle Singularity, he says “We will figure out new things to do and new things to want…Expectations will go up, but capabilities will go up equally quickly, and we’ll all get better stuff. We will build ever-more-wonderful things for each other.”
Of course, these new things need to be marketed and sold. Sam has good news there too, saying: “Generally speaking, the ability for one person to get much more done in 2030 than they could in 2020 will be a striking change”
This all sounds wonderful; it’s used so heavily by Silicon Valley, it’s been given the title of Effective Accelerationism. It’s essential thesis is that AI will cause progress all by itself. So we should just let it take over? Are we willing to bet our livelihoods on that?
Where we are here in 2025, it’s a challenge to do sales and marketing work using AI. Very few know how to run entire functions with Generative AI, which is why Sam qualified his 2030 prediction by saying that “many people will figure out how to benefit from [AI]” by then. How do we unlock AI’s activation in customer acquisition? How do we get out of the starters blocks?
I had the chance to moderate a panel discussion on “Gen AI Activation in Marketing & Sales” at an amazing event hosted by UC Labs and TCC Canada – please get the links to each of them in the shownotes.
The panel featured myself, Lubabah Bakht, Gary Amaral, Jim Cain, Peter MacKinnon, and Brett Serjeantson, zig zagging through everything from day-to-day challenges to legal and privacy concerns to the lack of skills barring our progress.
I count myself fortunate to not only share a panel with these experts, but for being able to call them friends.
And now, please listen to these experts on Activating generative AI in marketing and sales.
