Let's Embrace AI

Over the last 12 months, we have seen significant advances in AI technology across many areas, including AI assistants (ChatGPT, Claude, Gemini), Image Generation (GPT-4, Midjourney), Writing tools (Rytr, Sudowrite), and Marketing (AdCreative, AirOps), among others. It's grabbed the attention of many global leaders.

I've worked in the IT industry for just short of 20 years, and like anyone else, I would probably say things are constantly changing. Technology moves fast. Amongst all the excitement around AI, there is a high level of apprehension too, quite notably around job security. Jobs that we see threatened by AI are nearly identical to those that were threatened by the digital revolution over twenty years ago! The fact is that every industrial revolution has created more jobs and other opportunities, too.

I saw a post on LinkedIn recently, Bill Gates says, "Programming won't be fully automated anytime soon. Why? Because software isn't just syntax and repetition. It's creative judgment, trade-offs, and lots of thinking that still requires a human brain to solve ambiguous problems." He also believes that, despite the progress with AI, we are at least 100 years away from replacing human coders. I'll take this with a pinch of salt, but he did predict the risk of a world pandemic was greater than a war killing millions of people 5 years before COVID-19!

AI indeed has the potential to streamline and automate almost all aspects of software development, but as we are seeing, companies are investing heavily in AI, and the majority of that is employing a vast number of developers and AI experts, which tells you all you need to know about the job opportunities in the IT sector.

We can see day-to-day where AI is being used non-ethically with the use of deep fakes across social media platforms. There are risks that AI can and will be used to manipulate opinions, weakening democracy and the world will be at risk of major cyber-attacks.

I believe we are way off from being able to completely remove the human implications of what we do. AI does not appreciate and will have limited insight into what a system will feel like to an end customer, although we will see continuous improvements in the coming years. We should accept that not every problem needs AI; we risk duplicating efforts with problems that have already been solved with non-AI tools.

There are many constraints with AI in software:

• Yes, AI can create new services very quickly, but we will end up with 1000s of repositories to maintain with limited knowledge of how they work or integrate with other systems

• We want to create a bespoke system that current AI is unable to make a significant contribution to

• We may have limited knowledge of AI tools to use them effectively

• Code created by AI may need adapting to emerging customer needs, but deep human knowledge of how everything works could be lost

• All computers make mistakes, and AI could introduce errors that may take years to surface. Just look at the UK Post Office and the damage it caused

• Could ‘super smart’ AI start to cover its tracks? Recode errors and cover up history? Who knows…

Many of us are using AI every day, and sometimes we don’t realise. Is this a bad thing? No. We need to be pragmatic, think about it logically and prove that there is real value to benefit from. This is what organisations want to see: real value!

I’d love the narrative to change; the questions we should ask should be something like:

• How can we use AI to assist with repetitive day-to-day tasks to free up time for people to take on more creative or strategic tasks?

• How can we use AI to accelerate the development lifecycle, enabling us to add value much earlier?

• How can we use AI to generate a general concept into a detailed mock-up?

• How can we use AI to validate coding standards and ensure consistency?

• How can we use AI to generate system documentation?

• How can we embed AI in the services we build to create AI-native digital services to provide a better user experience?

So, AI won’t become you; it will help you, right?

We absolutely need to embrace change, accept that early adoption of AI is important to stay relevant, but do so with common sense. When designed well, it will make services much easier to build and use. Like anything else, if you don’t know what problem you are trying to solve, then any investment will end up being a waste of money.