A year ago, AI took the world by storm. Headlines like “workers worry that tech threatens their jobs” and “AI will replace 40% of jobs in 15 years” were everywhere. VC firms were throwing money at AI start-ups and the big tech companies were rushing to jump on the hype train. Now a year later, the dust has settled, and we can see that the AI revolution is not going to happen overnight. Let’s make up the balance.
AI made a big impact on developers but we are still at the beginning
It is impossible to deny that AI increased productivity for developers. The most obvious example is the rise of code completion. Tools like Tabnine, ChatGPT and Github copilot are using AI to help developers write code. This is a huge game changer for us and if I look to myself I am much faster and spend less time on mundane tasks. Actually I found my job more interesting. AI is a great help in the following areas:
- Code completion: AI helps me generating boilerplate code, create code skeletons based on a description and can suggest complete code snippets based on the context.
- Automated testing: I found AI a great help in writing unit tests. It can generate test cases based on the code and even generate test data based on the description of the code. In the past I spent lots of time creating good testdata and now I can generate it with a tap of a button.
- Writing documentation: I saved tons of time writing documentation. AI interprets the context for me and thanks to Natural Language Processing (NLP) it can generate documentation based on the code. I still review it and make changes to make it more readable but it saves me a lot of time.
Besides these good points I also found AI severely lacking in other areas:
- Adapting to new technologies and frameworks: AI can learn from existing data and based on the amount of data it makes its predictions. For new technology it takes a while for more data to become available and to learn from it. I experienced a lot that code completion suggested code snippets based on older versions of the library I was using.
- Personal design choices: when creating a new product besides knowing how to write the code and ship it you need to make sure the product fits the market. Understanding cultural nuances, user preferences and human psychology are hard for a machine to grasp.
- Complex problem solving: I found often that AI is good in solving problems that are defined within a certain (very specific) context. When the context changes or the problem becomes more complex AI is not able to adapt. Creating the architecture of a (part of a) system is still a human tasks due to the complexity of the problem.
How to stay in the game
The areas mentioned above offer us developers a great opportunity to stay in the game but it requires us to learn new skills. To begin with soft skills will become more important and sets us apart from the machines. The ability to understand team dynamics, communicate with management and understand the complex business dynamics will be essential for the future developer. AI lacks the capacity to understand feedback, social nunances and subtle contextual clues.
Next skill is to understand creative design. AI will be much better than us in writing algorithms and analysing data but it will be hard to understand human psychology. We humans are very complex and emotional beings and even us struggle to understand each other. For example selling: AI will always try to find the most efficient solution omitting the human factor why we choose a certain product or service.
Last but not least I think AI will remain having problems understanding ethics and privacy considerations in our business. If we keep up to date with the latest developments in this area we can do a better job at keeping our companies to stay out of trouble.
We can conclude that AI is here to stay and its impact on our work keeps growing. The time of developers that solely write code and solve algorithms is over and I think this will reflect future hiring processes. In the past we needed to solve a complex coding challenges, got mostly judged on our hard skills and work experience. I think their role will be less in the future, instead, I think soft skills, understanding the business and collaboration will be skills companies will be looking for. The future still looks bright for us but we have to adapt.