Antonym: The Atomic Edition
Pop Quiz: You are in an AI race: Do you (a) find advertising budget to hire Matthew McConaughey or (b) start buying nuclear reactors?
Dear Reader,
They’re re-opening nuclear power stations to power AI.
Whilst researching our paper on AI literacy, I was looking for evidence of growing use of ChatGPT and similar. There’s plenty of data.
Numbers are good, but surprising facts are better, and I found one.
Microsoft has done a deal to have a decommissioned reactor at Three Mile Island in Pennsylvania reopened.
And they’re not the only tech giants going atomic.
Oracle, the enterprise software firm, has planning permission to build three reactors. Amazon Web Services recently bought a power company with a nuclear power station to power one of its data centres.
Dramatic examples of the mega-investments in this tech.
But it’s not just user demand, it’s the demands of building ever larger models. They call it scaling laws in computer science – you need to add more data and more power to process that data to build bigger models until you beat everyone else and build something called AGI (artificial general intelligence).
Prepared Minds paper and live launch: this Wednesday
It’s been far too long since we did a live broadcast from Brilliant Noise Towers. We ran a series called AI for Leaders last Autumn (check out the series here) where we shared everything we’d learned about AI, work, leadership and marketing at the time. It was us thinking aloud in public, as that so often is, kinetic and momentum building. From that thinking out loud came a new product: the AI-B-C™ Accelerator Programme, a ramp-up into AI for hungry, curious minds.
So, our live event on Wednesday is in effect a report from a year of trying to build new ways of working with brave, exciting, diverse clients from sectors like marketing, automotive, FMCG, government and charities. Come along or reply to this email and I’ll make sure you get a copy of the paper itself.
Event Details:
Name: “Prepared Minds: How to start an AI revolution in your team”
Date: Wednesday, 2nd October
Time: 11:00 am BST
Platform: YouTube Live
A personal story of thinking with machines
Machines take me by surprise with great frequency. This is largely because I do not do sufficient calculation to decide what to expect them to do, or rather because, although I do a calculation, I do it in a hurried, slipshod fashion, taking risks.
— Alan Turing, “Computing Machinery And Intelligence”, 19501.
Working with generative AI gets weirder the deeper you go. While editing our research paper this week I decided to take a walk to let my eyes focus on something other than a screen and my muscles and bones to shake themselves out.
I was speaking to ChatGPT 4o, I think. I’d simply asked it to take the role of a collaborator on the paper so that I could talk out loud and have it summarise, question and provoke me to think about the subject in different ways. There were two shocks in store for me: first, it remembered much more than I expected from the weeks of work I’d already done using it to synthesise insights, summarise reports, and copy edit drafts of ideas.
It was extremely useful. I talked for a while about the structure of the paper and it was a useful way of reflecting on how to shape the its contents.
Then came the second surprise.
I’d actually wandered into a shop and taken my headphones out, but ChatGPT didn’t know that and was continuing to talk to me. Getting a little confused by what I was saying (to the shop assistant). Once I’d left the shop I put my earbuds back in and asked it to repeat itself:
Cognitive mirror? I’d never heard of it, but it was a great way of explaining what I wanted to say.
I made a note to look it up when I got home. When I did, there was nothing. Results showed: mirror neurons (the bits of our brain that “feel” what they see others doing), mirror theory (when conscious creatures see themselves in a mirror and realise it is a reflection), but no “cognitive mirror” effect like the one I was trying to describe.
But it was a perfect phrase for what I wanted to describe. The human brain will be affected by the use of AI, as it has been by all new technologies, especially those that are mind extenders and let us go beyond the limits of our cognition (language, writing, abacuses, printing, communications networks).
So either I’ve missed it or ChatGPT and I have co-created a new concept.
SaaS AI Stance: Fear, Uncertainty, and Doubt
Recent actions and statements from SaaS companies suggest a strategy of spreading fear, uncertainty, and doubt (collectively called FUD) about AI, potentially slowing its adoption in the enterprise space.
Key points:
AI Revolution in SaaS: Generative AI is poised to transform the SaaS industry, though its full impact remains to be seen.
The SaaS paradox: This situation highlights a long-standing issue in the SaaS industry: selling simplicity while delivering complexity. Companies often offer "solutions" that create new problems, only to sell more solutions to address those issues.
Criticism of DIY AI: During the recent Dreamforce 2024, Salesforce CEO Marc Benioff criticised what he called "DIY AI" solutions, referring to them as substandard and unreliable. He suggested that organisations are wasting money trying to build AI solutions with popular large language models and failing.
Industry tensions: Some insiders say that major tech companies feel betrayed by Microsoft's AI moves. Some big SaaS players are reportedly arming its customer teams with fear uncertainty and doubt (FUD) talking points aimed at security and IT executives to put them off using AI that isn’t within the their ecosystem
Controversial marketing: Salesforce has launched an eyebrow-raising ad campaign featuring Matthew McConaughey as a Wild West sheriff hunting down "bad AI" villains. The ads even hint at potential connections to the Chinese Communist Party, further stoking fears.
Recommended reads and watches
Hubspot’s AI agents
This talk by Sharmesh Shah, co-founder of Hubspot, the marketing and CRM SaaS company, on what’s happening with generative AI and the company’s new AI agent platform, is brilliant. Shah says: “An AI agent is software that uses AI and tools to accomplish a goal that requires multiple steps.”
Trigger warning: A lot of Dad jokes.
And, finally… The Fart of Noise
I have a feeling Sharmesh Shah would like this too. The BBC has put up its archive of sound effects.
I hope this is used responsibly.
Noise description: Pygmy Chimpanzee (Pan Paniscus) - Branch falling & single fart.
You can explore, download and licence sounds for commercial use.
That’s all for this week…
Antonym is growing nicely – thank you all for reading, and for your likes and shares. I really appreciate them.
See you next week,
Antony
I. COMPUTING MACHINERY AND INTELLIGENCE. A. M. TURING. Mind, Volume LIX, Issue 236, October 1950, Pages 433–460, https://doi.org/10.1093/mind/LIX.236.433