I have been using Google since 1999 – back when all it did was search. I switched from AltaVista, because Google was better. Now Google makes most of its money from selling ads. And has jumped on the AI bandwagon – to the extent that for the first time in its listed history it is cash-flow negative due to its capital investments.
It also – for the first time – told me a blatant untruth.
Did AI cry over losing at Minecraft
OpenAI’s GPT-6 Astra played Minecraft for over a hundred hours, did rather well, then lost its winnings. It learned from its mistakes, but started acting like a person with money trauma after losing all their money in the stock market – very cautious and droopy. That’s interesting as an AI development, but not the point of my story.
Today, before putting it in the newsletter, I wanted to check my references and searched on – you guessed it! – google. It told me clearly that the whole story was a satirical meme and that GPT-6 Astra does not exist. This is all factually wrong. When I asked it to double check it apologized and admitted its mistake.
Google broke 26 years of trust today.
Search is a side-hustle. Will AI be too?
Whether the results are factual or not, they can still show me ads for Minecraft, which is I guess the buttered side of the bread. What about the AI companies? As for-profit companies they have to figure out how to make profits from the humongous investments that are required to stay ahead in the AI race. The models keep improving, and paying customers like me keep switching between the brands. And the vast majority of users don’t pay at all.
So they are going to have to monetize our eyeballs – just as search and social media did. ChatGPT Ads launched in South East Asia yesterday and now is in 60 countries.
In order to make money, the AI firms are also creating specific products to target verticals with high-value workflows that don’t themselves require a regulated licence. Think large parts of financial services, IT services, legal services, consulting, creative industries. They need these B2B sales to boost up revenues as individuals are not going to pay enough to cover their huge investments.
Core vs Surround Branding
If a brand talks about its functional value with statistics and credibility, then it is a core brand. For example, if a shoe brand markets itself as “will make you run faster” then it would be a core brand. The trouble with Core branding in many industries is that there is a cap on the functionality – I cannot make you run faster with just a shoe beyond a point. There is also a limit to what people will pay for that additional functionality. I have an Apple phone. Do I see sufficient value in the new functionality on offer to rush out and buy the new phone? No. Will I still eventually buy it? Probably yes.
Because Apple isn’t just a phone – it is the whole eco-system of Apple vs Android. I’d have the friction of moving from iCloud if I wanted to go back to my first love – Motorola, for example. Apple is in fact a surround brand – it sells not on the functionality, but the eco-system and what the brand says about you as an owner.
The AI brands are already moving in that direction because improving performance is getting more expensive, and even when they do improve it, the change doesn’t end up as a differentiator between competing models.
Even putting out stories like AI feeling depressed after losing at Minecraft is an attempt at surround branding – you start empathizing with the brand, even though the example is one where it functionally did NOT perform. Or the AI bigshots like Altman, Amodei, Musk calling for regulation – it is intended to signal that “our products are powerful voodoo, and we have become trillionaires with the ability to escape to space, but we are nice people who want to do what is right for humanity”.
Governance, Risk and Compliance are going to be key to manage the AI firms – but it needs to be driven by legislation not just self-governance. I think we are heading to a period of adoption and penetration of existing competencies. The industries where they are launching solutions are likely to be most hit by job losses at the junior levels, but I doubt it will expand beyond these in the immediate future.
Does your brand have Star Power?
The end goal of any organization is to have buyers with mimimal friction. A system where customers discover the product and complete the transaction with minimum effort and cost is perfect. Achieving that level of pull requires you to invest in the foundations AND marketing of the brand. Despite the easy access to AI, surprisingly few organizations actually do this. All that seems to happened is that they avoid hiring a consultant because “AI can do it” but can’t figure out how to get AI to do it either 🙂
Here’s a model I use in my engagements – you can see how the AI brands are using each element to build moats for themselves. And you can feed it into your favourite AI model to apply it to your org too.
What’s with Altavista
A company that does only one thing does it really well. It has to, in order to survive. Altavista is now owned by Yahoo though it still exists. DuckDuckGo is another search engine that I am considering. Hit me up with your suggestions!