AI is about to revolutionise websites and marketing strategies.
A/B Testing is a relatively common technique used by web designers now. It is the trial and error of the marketing realm, experimenting with change and rolling out the successes, incrementally improving the effectiveness of your site.
A/B Testing, and the incremental improvement of business strategy, takes it’s inspiration from Science, but it can now also be seen in sports, particularly cycling with their marginal gains strategy.
Incremental improvement is more subtle than breakthrough improvement, think evolution rather than revolution. However, as revolutions are higher risk, their potential impact (good and bad) could also be high.
Predictive-code and Autocomplete by AI
One breakthrough in recent times has been Artificial Intelligence (AI). AI is a branch of computer science which aims to create systems capable of machine learning and performing tasks that previously required human input.
This is a pretty broad definition and can include tasks such as:
- Content generation
- Way-finding (self-driving cars)
- Manufacturing robots
- Chat support / bot
- Web Design and Development
As web developers, we are extremely excited about content generation, chat bots, and coding.
We have been closely following the work of the team at OpenAI, a non-profit AI research and development firm based in the US.
Their most recent development (at time of writing) is OpenAI Codex, which powers a system called copilot – a joint venture between OpenAI and GitHub.
Essentially, think of it as autocomplete, but for code. Simply detail what it is you want to achieve IN PLAIN ENGLISH(!), and the AI will automatically generate lines of code to achieve what it is that you wrote, or at least how it interpreted what you wrote.
This is game changing. Not only will it speed up the time it takes for developers to create a site, but I don’t see us being too far away from members of the public uploading briefs and having AI do all the work.
At present, our jobs our safe, as you must still be able to understand the code that Copilot suggests, picking from its numerous suggestions, and approving the snippet(s) of code you would like.
AI recommending improvements is not new. Google implemented this a while back within their PPC Adverts. Recommendations such as “Increase your budget by £x and see y% increase in clicks, or “add these keywords to see an z% increase in conversions”.
It’s not all bad news for the fate of web developers, as when the concept of incremental improvements made by AI is applied to web design, the often monotonous task of A/B testing can be fully automated.
All aspects that would usually be done by hand, such as website layout, use of colour, and even content, can be changed automatically with the successes rolled out.
More than this though, AI systems will be able to generate its own best practice from the collective results of numerous websites, meaning there will be fewer errors in its ‘trial and error’.
The advancements in AI over the last few years is unfathomable. It is progressing at a rate faster than any other area of computer science. If we (or AI) are capable of the above now, it’s extremely exciting to see what impact AI will have in another 5 years.
An interesting dilemma worth thinking about, though, is this: who takes the credit (or payment) for work commissioned by a client, but generated by AI? Is it the web developer who utilises the latest AI technology? Or is it the computer scientist who created the AI?
How will copyright work in an industry where ideas are not generated by people, but are merely posed by people and generated by AI?
We thought it was apt to conclude this article with content generated by AI.
When prompted with “The future…”, AI suggested that the “The future is unpredictable.”
We couldn’t agree more.