Will Ranjan-Churchill

Some thoughts on AI

ChatGPT has been around for a little while. Lately though, we've had some other big players join the "Generative AI" arena. Perhaps most notable is Microsoft making their Copilot considerably easier to access. For about £25 a month, you get Microsoft's range of AI tools.

So, what does this mean for the current state of play for Generative Artificial Intelligence? Who is using it, and what are they using it for? Perhaps most importantly, is it worth getting?

Let's find out.

What are we using AI for?

In a survey of over 140 IT Leaders, The Channel Company (parent company for, among other publications, Computing.com) found that 76% of respondents use AI for Data Analytics. Additionally, 60% said they used AI Assistants, and 40% used AI for Natural Language Processing.

What is interesting to me is that none of these things are groundbreaking. We've had Machine Learning helping us with data analysis for decades, and AI Assistants have been accessible to the masses in various forms since the first iPhone. Don't get me wrong. The quality of the work from today's AI is streets ahead of where it used to be. The idea that 2011 Siri would do anything useful is laughable. But sci-fi of old promised us a redefined society: a flying-car filled utopia. When are we getting that?

What we're using AI for isn't groundbreaking - but it's reach is about to be.

We're a little way off yet. But what is happening is still hugely beneficial, and we're only just figuring out how the "average" person can use it.

Let's look at Microsoft Copilot again. One of the things it advertises is a better way to attend meetings. You can join a Teams call with your colleagues and have Copilot "sit in" taking notes. At the end of the meeting, you'll almost instantly have a transcript of what was said by whom. But no one's got time to read a transcript to understand what went on, so you'll also get a summary, distilled into 5 or 6 bullet points, as well as any action points from the meeting. That's easily an hour of someone's time post-meeting that they can get back.

What else? Let's say you've got to run through the results of a 300-person survey. You've got to collate all that data, put it into a two-page report for your Senior Management Team, and you also need to produce a 5-slide presentation to give at the next all-hands meeting. How long are you going to block out to do that? A couple of hours? Half a day? A full day? Seems long. And besides, no one has a full day to put aside for this kind of work, so you're going to have to split that time out across three weeks' worth of 2 hours here and 40 minutes there. It's going be draining having to switch back and forth, and it'll likely push back other timelines that you're working on.

That seems like a waste of everyone's time. Now, you can get Copilot to do all that in a matter of minutes, and you only spend half an hour checking it over and making some light edits.

This sort of thing is going to become really popular, really quickly. A lot of people already have a Microsoft 365 license, so adding this feature on for - as all the marketing will remind you - about £1 a day seems like a no-brainer. It's also a much simpler sell for companies compared to, for example, ChatGPT. Copilot is ChatGPT with bells on, but those bells integrate it directly with your email, your calendar, with PowerPoint, with…

Because of the price point, companies will quickly see a return on spending. It's hard to get figures on exactly what this looks like. But even if Microsoft's claim that you can "clear your inbox in minutes, not hours" is the only thing the tool does, and you spend two hours a day on emails, then suddenly £25 buys me nearly 3 days a month back. And that's without the other bells and whistles Microsoft tells me Copilot can do. With that time, I can start to work on more strategic work.

To work that out, I assumed that Copilot could cut down my emailing from 2 hours to 1 hour a day. I assumed a 5-day working week, meaning 5 hours saved over a week. With an average of 4 weeks in a month, that's 20 hours saved. If a work day lasts 8 hours, then overall I get 2.5 days back just by saving 1 hour a day of email time.

But will AI take my job?

Am I worried about jobs? Only a little. As someone who has been a web developer for the last decade and a half, I liken the rise in Generative AI to the popularity of platforms like Squarespace.

Squarespace leveled the playing field when it came to building websites. It's a great piece of kit for anyone who wants to build a small portfolio website, a blog, a modest online store, etc. If you want a site, you want it done, and you want as little fuss as possible - have at it. Spend a couple of hours getting the look right - you don't need to spin up servers, you don't need to configure WordPress or install plugins. £10/month, job done. It's fantastic - and I say that as someone who makes money doing the same thing that you can pay Squarespace to do.

Except I don't - and that's the point. Squarespace isn't going to help you if you need a completely stand-out website. It's not the right tool for you if you want a bespoke CMS for your publication. It's going to struggle if you need a high-traffic, multi-seller, international e-commerce platform. You can't use it if you have specific data security requirements. Squarespace isn't designed to help in those cases. In those scenarios, and hundreds more, you need a web developer. (Hello, by the way! 👋)

Generative AI is like Squarespace - a great tool for a simple and specific need.

I don't believe I've lost significant business because Squarespace exists. I also think that thousands more people have an accessible online voice because of the likes of Squarespace, people who wouldn't have been able to have one without it.

In the same vein, I don't believe this wave of AI will result in millions of 9-5 "office worker" jobs lost. There are things it just isn't designed to do. There was a period where we all got caught up in the hype that AI would solve everything, and we'd be able to kick back in our utopia while the robots did all the work. Maybe that's coming, but not yet. Not soon. And that's OK - technology isn't really supposed to be a broad-brush solution to all problems. Specific interventions need to be specifically deployed to get the best out of it. That's where I believe we're at. We are going to see an exponential rise in the number of people using Generative AI, in a much smaller number of use cases.

I think that's going to be really exciting. The low-hanging fruit, the easy return on investment stuff, will be more widely used. It's will really open up how people can spend their time - not just physical hours, but also the mental capacity to think about more important things.

Final thoughts

It's quite the rosy picture I've painted! Don't misunderstand me, I know there are still hurdles to jump. There are ethical considerations, climate change factors, and myriad data concerns surrounding AI that I've not explored here. In the coming weeks and months, I will break these down. Watch this space.

I also know that there are likely to be cases of some jobs being replaced, despite my sunny optimism. But in the round, I think we're less likely to see entire jobs being taken over by machines, and instead have individual people be "augmented" in specific ways, giving them back time and capacity.

I work predominately in the charity and social sector. It's a space where it's common for any given person to be doing the work of four people. I'm excited for a world where it's accessible - both in terms of price and ease of use - for people in the space to only do the work of three people. Maybe even only two.

I jest (sort of). But it is a real possibility. Small organisations can't afford to hire extra people, and AI isn't able to replace existing ones. But, perhaps for the first time, we have a genuinely viable third option, where for a little extra money we might get a lot more out. It will help people redefine what their work is, and to spend their time doing more meaningful work.

So that's where I think we are. And I am excited.

ai, internet


Want to chat about anything in this post? I'd love to hear from you!

Feel free to drop me an email

Or find me on Mastodon



⬅ Previous post
I've recently changed how I consume content

Next post ➡
RSS: Why I think it's overdue a comeback