Good day! Hello!
Welcome to another edition of AI versus Marshy.
The newsletter that’s charting the impact of AI from the perspective of a neurodivergent tech-y growth marketer who cares about the impact on the world.
😬 Another reason for saying AI’s a threat - commercial advantage
🔨 Tool of the week: A hands-on demo of how ChatGPT can translate a marketing sprint
🧠 AI can scan your brain and guess what you’re seeing now
Let’s get going because today is a humdinger!
Trust us - and regulate everyone else
Last week we picked on Meta’s AI scientist - who said AI isn’t an existential threat. It’s not the time for grand sweeping assertions as really everything is up for discussion this early in the race.
I’m currently reading Techno Feudalism by Yanis Varoufakis (h/t one our readers) and if history can teach us anything it’s that it repeats itself and humans will find evil ways to use technology - dating at least as far back as iron.
The AFR has uncovered another angle get there - commercial advantage.
Andrew Ng was the founder of Google Brain, and later led AI at Baidu. He told the newspaper when big tech paints AI as a big threat, and poisitions themselves as the force for good to address it, there’s a knock-on effect that will benefit big tech.
Simplified - it goes like this*:
- BT (Racing ahead) AI bad, we’re good, trust us to keep researching how to harness it
- Big G(ovt): We need to regulate this
- BT (racing further ahead) while regulation comes into play: Hehe, that should be enough of a headstart
- Big G: Unleash the regulation!
- Any other competitors: Well sheee–iiiit, we can’t do anything with all of this red tape
- BT: This is quite the moat we’ve built for ourselves, look at all this power!
This is playing out as we speak.
I recently watched a lecture (it’s not recorded sorry) from the Centre for AI and Digital Ethics (from my old university) by Julia Powles - she’s an associate professor of Law at UWA.
The topic was Corporate Culpability of Big Tech and pointed to the fact that tech companies have a history of moving fast and disrupting (including breaching any regulation) to gain advantage before the law can catch-up.
Within tech circles this is known as blitzscaling.
This is what Uber did in new markets and continues to unfold on Planet Earth.
Australia tends to punch above its weight in calling this stuff out, here’s a link to a submission from CAIDE calling out its influence in Australia in response to a Senate enquiry (page 11, the Harms section is on point ✅).
These sorts of arguments are generally on the fringe at best for everyday people - the reason it’s important is because it’s going to set the tone for impacts that will be felt for decades to come.
Personally - I was very convinced by Sam Altman’s belief that the OpenAI approach is going to help the planet.
But me being persuaded isn’t the same as it being true.
Rest assured we’ll be continuing to watch this space! 🍿
Cool tool - Using ChatGPT to translate a marketing sprint
We’re trying something different this week.
Here’s the de-identified output of a 6-week marketing sprint done in Figma from earlier in the year.
This process is what we do when we’re working with a client more hands-on. They need lots of things from marketing and have no idea what’s most important or who can do what.
Post-its (digital or otherwise) are really effective for helping clients see what should follow, what’s in a group, or what should happen before what.
You can now feed this a photo or screenshot into ChatGPT with its image recognition.
I can then ask it to do a time estimate, and suggest what items would be best executed by a Virtual Assistant.
I think this sort of use case, where it’s playing with tasks, contextualising and grouping them, and working off answers that are “good enough” for now is going to continue growing in importance.
The world is going to become more ambiguous and more uncertain, so taking action with incomplete information will outpace any other behaviours if growth is the focus.
I’m a big fan of taking action with what you have now over planning for a future that might never arrive.
We can almost see what you see in your brain now
Meta ARE doing some amazing things with AI. I love the open-source approach and they’re releasing tonnes of research for the betterment of everyone.
They released some research in October that’s pretty wild.
What you’re seeing is a machine translation of what a person is seeing - using brain scanning to “map” and guess what the person is seeing.
Put in even simpler terms:
Machine scans brain and sees what you’re seeing.
It does this with machine learning - matching a scan to an array of images and using a feedback loop to grade/guess more and more precisely.
It’s not like for like - yet - but its getting closer to translating what the brain does with machines 🤯
I told you it was a good edition!
Last week I was in Sydney at short notice to speak on a panel about my lived experience for SANE.
The recording isn’t up yet but I’ll share it when it happens.
I’ve been closing out the year and focusing more on what’s going to happen next year.
It’s been a big one - fatherhood is incredible.
Professionally I’m looking forward to working on more interesting projects and seeing where it all leads to for providing for my family as well as doing some good in the world.