20th January 2023
Predictions for the development of assistive AI technologies in 2023
With everyone talking about the exciting (and slightly scary) new chatbot, ChatGPT, we take a look at what this could mean for the future of our industry.
Every year, it seems that people make the same two predictions: fusion power will soon be only a decade away, and this year is going to be the year that AI breaks into the mainstream.
Well, we haven't really had that much news on fusion power recently. We’ve had some promising developments in the last year but sustained fusion power is just as far away. However, it feels like AI is actually making it into the mainstream. And the main reason for that is the natural language chatbot, ChatGPT.
The meteoric rise of ChatGPT
Within five days of its release in November 2022, ChatGPT had amassed over a million users (for comparison, Facebook took ten months to reach the same milestone). Everyone's talking about it. It's free to use, it's a fascinating tool, and there's loads of resources available online to provide further insights.
The actual technology behind ChatGPT isn't very new. It's just that some of the big players, like Meta and Google, haven't really dared to release something to the public before, because they've got too much to lose. And when they've released something, the consequences have been really bad press. We’ve had reports of chatbots promoting anti-semitism or eating glass, and one Google engineer was fired after claiming the LaMDA chatbot was sentient.
As a relatively small player, Open AI (with big support from Microsoft and its Azure Cloud) has been able to sidestep the PR storms that have plagued other chatbots. Their technology may not be breaking new scientific ground, but for them the PR storm is one of amazement rather than anger.
It isn’t just big tech companies that are using assistive AI now. We’ve seen many news organisations revealing that they've been using AI to help them write their stories. Politicians have used AI to help them fine-tune their speeches. This technology is here and it's here to stay, so let us consider the interesting implications of that.
The consequences of AI
For big players like Google or Meta, whose income comes from advertising, a freely available and seemingly intelligent chatbot presents a big problem. What if people stop using search platforms full of adverts because they can tap something into a chat, or perhaps soon send a text message or WhatsApp, and get the answers they need?
Although ChatGPT isn't currently able to search the Internet, in the coming months it may be attached to Bing - and then what’s next? Speech and home devices are bound to be next. If someone would hook up ChatGPT to a text-to-voice system that sounds like the WOPR from WarGames then I would be using it all the time. Sorry Alexa and Siri, you're just not fun to talk to anymore.
Also, it's cheap. In fact, at the moment it’s free, but it's a struggle to reliably get access. However, it's been confirmed that ChatGPT is going to have a pro version that will be $42 a month, which is nothing. So if you use a computer in your job and that job has anything to do with text, whether it be marketing, sales, customer service or even programming, then ChatGPT can easily save you $42 worth of time, probably every day. It just makes your job more efficient.
So it's kind of a no-brainer - as soon as the pro version of ChatGPT comes out, you really need to get a licence and get on board because it's going to save you a lot of time. Sure, you’ll have to invest some time to work out which areas of your job will benefit, adapt some processes and learn how to best use this new tool, but you better start using AI to improve your productivity - because if you're not doing it, your competitors will be.
Is AI coming for your job?
I think one area where we are misunderstanding the impact of AI is around impact to employment.
ChatGPT isn't going to replace your job - it's something that augments a job - but if your company doesn’t recognise this, then jobs could be at risk as work goes to more productive competitors.
A lot of people thought the ‘Rise of the Machines’, if I can use that analogy, would affect manual jobs first. And sure, we have seen this in many industries already. However, to replace humans with robotic systems you need clearly defined processes (like we see in manufacturing, distribution and retail) and a lot of investment for some industries. The numbers still favour humans.
It has been said that “the Romans could have invented steam engines, but they didn't bother because they had too many slaves.” So with robotic automation, there is an economic limit to how quickly and extensively it can be used. Sometimes robotics is just not worth the investment required when labour is (comparatively) cheap.
On the other hand, chatbot technology uses cloud infrastructure which, although it requires expensive computing power, is already being deployed at massive scale. End users don’t have to build new data centres, Microsoft is already doing this, you just need to pop in your card details and away you go.
However, when it comes to something like chatbots, instead of it being manual jobs which are affected, it's actually professional office jobs. White collar workers generate higher value per person/hour but then demand higher salaries in return. The opportunity value for professional office workers to use assistive AI and chatbots to increase their output is huge. Ignore this technology at your peril.
Just as mainstream access to the Internet and mobile phones created opportunities for some businesses and abrupt endings for others, I believe that mainstream access to assistive AI technologies will do the same. Watch this space - whatever happens, the world will never be the same.