Previously on the AI Academy Diaries…
In AI Academy Diaries #1: Confessions of a Content Marketer Learning AI, I tackled the foundations, embraced JOMO (the Joy of Missing Out), and started replacing that nagging feeling of being behind with something much more useful: confidence.
In AI Academy Diaries #2: Busting the Biggest AI Myths, I busted some of the biggest myths about AI, learnt the difference between AI and Generative AI, looked at algorithms from an AI perspective rather than a “the Instagram algorithm hates me” perspective, and discovered why we should probably stop battling the machine and learn to work with it instead.
Following the last instalment of the AI Academy Diaries, I’ve definitely come away with a much better understanding of the context around AI.
It feels considerably less like an ominous presence lurking in the corner when you understand that it isn’t actually a sentient being plotting our downfall, but just a tool predicting patterns. It still very much needs a human keeping an eye on it, though, because AI can be susceptible to the same biases found in the human-created data it learns from.
If the last time was about dismantling my walls of “AHHH! AI! TERMINATOR! SKYNET!”, these latest sessions have helped me build them back up again, except this time they’re much more sensible and useful walls, built around understanding what AI can get wrong, why it gets things wrong and why we humans still very much need to be involved.
So I suppose I’ve moved from “DON’T TRUST THE MACHINES” to “trust the machines, but keep an eye on them in the same way you wouldn’t leave the oven on while you pop out of the house.” You’re not terrified of the oven, you use it all the time and it makes your life considerably easier, but that doesn’t mean you abandon all common sense and leave it unsupervised at 220°C while you pop to Tesco.
So, let’s dive into the ethics of AI and the importance of critical thinking.
🎶 Let’s get critical, critical. I wanna get critical. Let me hear your mind talk. 🎶
The Importance of Ethics and AI
The ethics surrounding AI was probably one of the biggest things that made me hesitant to explore it properly in the first place. I’m naturally quite cautious, and the idea of putting information into a machine with no real understanding of where it could go, who could potentially access it or what happened to it afterwards made me nervous.
So, before I really started using AI, I did a bit of research, worked out some boundaries that I felt comfortable with and carried on from there, mostly hoping that I was doing it ‘right’.
What I really appreciated about this part of the Learning Journey was that it helped me build on those initial ethical markers I’d created for myself and actually understand the reasoning behind them. I’m so glad the course gets into the nitty-gritty here, because it would be very easy for AI training to focus entirely on all the brilliant things these tools can do while sweeping the security risks and mistakes under the rug.
Instead, we actually looked at where things have gone wrong, why they’ve gone wrong and, crucially, what humans can do differently to help avoid them happening in the first place.
There’s also really clear guidance around the kinds of information you should avoid feeding into AI systems, which I think is something organisations would find immensely valuable. I know I did, because my cobbled-together personal rulebook now had some actual foundations underneath it.
And once again, educator Nicky has this knack for taking a subject that could quite easily send you spiralling into “OH GOD, HAVE I BEEN DOING EVERYTHING WRONG?” territory and guiding you through it without making it frightening.
She’s basically a lighthouse in the dark, muddy waters of AI ethics, calmly guiding you around the rocks while you’re in a tiny boat clutching your laptop. Thankfully, by the time you reach the shore, you understand that responsible AI use doesn’t mean being terrified of the water, you just have to know where the rocks are.
Keeping the Human in the Loop
The Human in the Loop Learning Journey is all about what I hope will eventually become the normal way we use AI without even really thinking about it: not cultivating an overreliance on it, but matching its greatest abilities with ours to create something better than either of us could produce alone.
A big part of that is learning not to take whatever the machine spits out and immediately run into the world with it like it’s been delivered on a stone tablet.
Nicky uses some really useful examples to show how seemingly small AI mistakes can have very real consequences for a business. Say you ask AI to help you write a client proposal and somewhere along the way it adds a discount and promises the project can start on a particular date. Lovely! What an efficient little robot pal you’ve got yourself there.
Except nobody with the authority to make those decisions has actually approved either of those things.
If that proposal goes straight to the client without a human properly checking it, you’ve potentially made promises your business can’t keep, and telling the customer afterwards that “the AI said it” really won’t land well.
We can’t take AI at face value. It’s technology created by humans and trained on human-created data, so the idea that it would somehow emerge completely immune to mistakes, biases and inaccuracies was a tad naive of us.
That was probably one of my biggest takeaways from this particular Learning Journey: we can’t just blame AI when something goes wrong. If the final output has our name next to it, the responsibility and risk are still ours, so just give it a good read and fact-check before firing it into someone’s inbox.
Let’s get critical, critical. I wanna get critical… thinking.
The final Learning Journey I looked at this month covers what I think is one of the most important things we’ve discussed so far: how to apply critical thinking to AI.
You know when your mum sends you a video of a cow juggling and you have to gently ask whether she really thinks that cow has spent years perfecting its circus skills, or whether there’s perhaps a small chance it might be AI-generated? That’s the kind of wariness we need to build.
We can’t automatically accept everything we see at face value anymore, which is personally devastating because I would very much like to live in a world where every video of an animal doing something ridiculous is completely real. Maybe one day cows will learn to juggle. I can dream.
In this session, we explored how AI outputs can contain facts, opinions and truth claims which, when you think about it, sounds suspiciously like… you guessed it… a human! And just as we wouldn’t automatically believe everything another person tells us, we shouldn’t suddenly abandon that instinct because the information has arrived in a very authoritative-looking little box on our screen.
One thing Nicky says very early on really stuck with me: you cannot use AI to “outsource your thinking.” We still need to evaluate and assess what an AI tool gives us rather than assuming that because it sounds convincing, it must be correct. Nicky introduces a really useful three-point checklist for evaluating AI outputs, which gives me something practical to come back to whenever I’m deciding whether an answer deserves my trust.
We need to automatically assess AI output and ask what assumptions might be hiding in there, whose perspective could be missing and whether bias might have crept into the response.
Basically, we need to treat AI like a politician backed into a corner on Question Time: listen to what it’s telling you, but keep your cynicism tucked in your pocket at all times.
And that’s probably my biggest takeaway from this month’s Learning Journey. Nicky helped me realise that one of the greatest assets we can have in the AI era is a critical eye.
We need to train our critical thinking like a muscle because, in a world filled with scarily convincing AI-generated content, being able to stop, question something and decide whether it actually makes sense is becoming a bit of a superpower. At least until cows actually learn how to juggle.
Next Time on the AI Academy Diaries…
Next month, we’re at the halfway point and exploring why AI can give different answers to the same questions, why it sometimes gets things spectacularly wrong and why it can be just a little too agreeable for its own good.
Basically, we’re going to find out why AI is sometimes less of an all-knowing supercomputer and more of a mate who doesn’t want to upset you.
I’m excited for this one. See you next month! 👋
Solvaa: Strategy, Solutions, Support, Education
UK-based Solvaa® is a systems, process and automation consultancy specialising in improving business workflows. Solvaa® works with SMEs worldwide to implement digital tools, better processes, use business process automation and generative AI to boost productivity, enhance lead management, and remove unnecessary time, costs and hassles from stretched business owners’ days.
We help you simplify and streamline existing processes, integrate your business tools to make data transfer more efficient, and automate manual tasks. Our business process automation specialists have extensive experience using multiple cloud-based business tools, including Zapier, Pipedrive CRM, SmartSuite, Xero, Google Suite, ClickUp, Asana, Airtable, Hubspot, Active Campaign, Salesforce, Mailchimp, Quickbooks and many more. We are Zapier Certified Experts, SmartSuite Certified Partners, Pipedrive Certified Partners and Xero Certified Advisors. We also educate teams about using AI in their daily roles, mindset, governance and strategy in our Solvaa AI Academy.
Solvaa® was founded in 2016 by Kelly Goss, a no-code solutions architect, business process automation specialist and author of Automate It with Zapier (first edition) and Automate It with Zapier and Generative AI (second edition).
A thought leader on business process automation and digital adoption, Kelly attended an invite-only Roundtable on Artificial Intelligence (AI) with the UK government at No.10 Downing Street, providing insights and recommendations on how SMEs can benefit from productivity gains by using digital technologies, automation and AI. She has also been invited to speak at the House of Lords and at many other events.
Contact us today to find out how Solvaa can help your business make big savings on operational costs and time, and significantly increase productivity.
Related Posts
