November 15, 2023

Voices for the AI Summit – prepare don’t drift

By Tristan Mobbs

This post continues our series in response to the global AI Summit hosted by the UK government to tackle our need to prepare for an AI future. By stepping back & considering our progress and behaviour to date we can help avoid drifting into foreseeable pitfalls.

Our author this time is guest blogger Tristan Mobbs. Regular readers will recall that Tristan is the “Data Translator” who has built a great reputation on LinkedIn for sharing practical advice. He has also shared some popular past posts with us on topics including developing your thinking skills and making the most of your existing tech.

In response to my challenge to all our guest bloggers, Tristan shares below what he would have contributed if invited to that famous AI Summit. Over to Tristan to help us step back, think about both the opportunity of AI and what we should have learnt already about ethical considerations.

Recognising the case for AI

Like it or not, AI is here to stay. AI is another tool in the toolbox, a supercharged tech version of the Swiss army knife. How we utilize it to make everyone’s lives better, more efficient, and not just full of fake mediocrity, will be up to us.

Why would we use AI? Even the basics of ChatGPT responses deliver great efficiency gains in answering questions compared to searching on everyone’s favourite search engine. This is just the beginning we are told. Soon we will be able to summarise documents, touch up and edit photos just by using words, and even do data analysis. The UK has seen a declining rate of productivity growth in recent years. Is AI the catalyst to drive the next surge of productivity?

AI can and will deliver here, we will be able to do more with less. We will be able to get started on things in less time and be able to deliver a level of output that is at least acceptable. If you are rubbish at writing, well AI can do that. If you struggle with expanding on your one idea, well AI can give you a plethora of ideas to expand on it. What about risks? AI can tell you about those too.

So here we are. At the start of a journey of having a new tool at our fingertips. We have the chance to increase productivity hugely. We have the chance to work less. Do more with our time. And do it at a level that is acceptable. We won’t be the best, but sometimes we don’t want to be. People with exceptional skills will still be able to shine and will stand out even more for their work, they will likely be rewarded more handsomely too. Yet is this likely?

Our track record with using (or not) new technology

Manufacturing automation was meant to enable us to work less. But we just work the same and produce more stuff. AI will get added into everything else we already do, to do even more things. 

We have had Excel since 1985. We have had databases and SQL for longer than that. Thirty-seven years later we are still manually copying and pasting numbers around, making mistakes in formulas, and making work for ourselves in places where we add no value. We create mind-numbing tasks to keep ourselves occupied and ‘useful’ to an organization to earn a wage. What difference will adding another tool into the mix make?

We still have banks making you print out documents to sign them and scan them back in to prove your identity. We have the DVLA operating by post and paper forms. Yet we can get approved for a credit card in under one minute. Clearly, we still live in a mixed reality, despite all that technological progress.

The ethics of how we use AI in future

Where do you think AI will be most used? It won’t be for the people doing the mind-numbing jobs. They have no incentive to change. It won’t be used to improve those inefficient processes we have to do under the sub-heading of life admin. It will be used to profile. To identify where we can make the most money. It will be to remove those entry-level jobs where graduates could add a bit of value while getting a few things wrong along the way. Why pay an entry-level salary for a year of blog posts? ChatGPT can write them for you in seconds for a token amount each month.

Where else will AI dumb down society? We can be spoon-fed so much information so quickly, AI makes this even easier. What if that information is wrong? What if it is manipulated? You only need to look at the case of Brexit and Cambridge Analytica to see the potential there. AI could increase the opportunity and efficiency of bad actors by some scale.

As AI becomes more embedded into our lives, you ask the question – who is accountable? If someone makes a mistake at work, then they can get fired. When AI gets something wrong, who is at fault? If we come to rely on AI so much that we cannot validate the outputs, then how can we ever know it is wrong until something catastrophic happens? With all our excitement and willingness to chase the next big thing we often go too far too quickly that we can’t back out. How do we plan for when the inevitable happens? How can we ensure that the impact is contained? The 2010 flash crash caused by High-Frequency trading could be made to look insignificant compared to the potential of AI to compound misadventure.

The verdict: it’s up to us

The opportunity AI presents is massive. How do we ensure it is used in ways that benefit society and ensure that people don’t get left behind? We need to educate and show what areas can be trusted and where we require more critical thinking. We need to change the way schools teach. Information retention isn’t an issue anymore. Validating information, critically thinking through additional knowledge, and being able to find the gaps or crossover between two disciplines; here is where the opportunity lies. 

Humanity evolves because we strive for more. Ego and ambition have shown we can achieve great things. It has also delivered some horrors along the way. AI is going to do similar. There are going to be some great leaps, people will deliver more. Yet there will be manipulation and suboptimal outcomes for society too. We can choose to work less as we increase our ability to complete tasks in less time. Or we can work the same amount, produce more, and waste more. We can use AI to create even more detached forms of reality, keeping us stimulated but with no time to think.  At some point, we might consider which path we want to choose.