Well, 2025 was a big year for AI, and as we enter 2026, we may be at the start of a shift as monumental as the industrial revolution, or perhaps in one of the most enormous speculative technology bubbles we’ve ever seen. As a web designer, I’m inundated with endless AI products that promise to make my work effortless and instantaneous. I remain a little bit skeptical to this point.
A Paradigm Shift
There is no denying that Artificial Intelligence has shaken things up over the past few years. Since the launch of ChatGPT in November 2022, the technology has continued to progress at breakneck speed.. There are dozens of major LLMs available now, and every major software company is introducing new AI functionality. It seems that AI is now used in tasks as varied as creating blog content and taking your lunch order.
In some ways, our society is really putting AI through its paces in a “throw it to the wall and see what sticks” approach. Some of it is very useful, and some of it is not. We’re still, collectively, trying to get our heads around what AI can do, and more importantly, what it can do better and faster than we can. Do we need to start thinking about jobs that AI can’t do, like tending goats?
What AI Does Really Well
Brainstorming
AI is excellent for generating ideas. If you want to come up with an outline for an essay or even a book, AI can suggest a structure that makes sense and come up with ideas, not all of them great, which you may never have thought of on your own. Like the human process of brainstorming, it can quickly generate many possibilities, which can then be whittled down to a more manageable set later in the process.
Repetitive and Tedious tasks
If you have ever had to do something as menial as reformatting large amounts of text, you can understand why this is extremely helpful. I have, on more than one occasion, received a lot of content from clients that, when pasted into a website, caused all kinds of problems, because the Word Processor imparted all sorts of embedded styles. Getting AI to reformat this text is a snap, saving both time and frustration.
Basic Coding
I recently had a client who was using a third-party service to schedule meetings. The scheduler redirected the user back to the client’s site after booking, passing along the booking data in the URL. The problem was that the data was not in a format that was at all useful. The use of AI to write a code snippet that could read that extra data and format it correctly took a few minutes, saving me quite a bit of trial and error. Writing similar small programs to perform a basic function is a very valuable feature.
Image Generation
While AI image generation is far from perfect, being able to create images from a prompt can be extremely helpful when you cannot find the ideal image in a stock library and do not have the budget to get custom images shot. The ability to edit existing images with AI is perhaps more impressive when the image is “almost perfect” and needs a tweak or two. Indeed, this can create results so much faster than manually retouching and compositing images in Photoshop.
What AI Doesn’t do So Well
Writing Long Form Content
In theory, AI can do a great job of creating content for things like blogs, but the content has a particular feeling; it often says a lot without saying anything at all. Loads of fluff, strange syntax, and an overabundance of em-dashes make AI content pretty easy to see. High-quality content still needs to be at least retouched by a human. A human who can inject a bit of their own personal experiences and opinions into the piece, so that other humans can better connect with it.
Designing or Coding Entire Projects
While AI can augment a designer or developer’s workflow, making it more efficient, the temptation can be to generate entire projects with AI. Currently, this typically yields poor results and can be challenging to iterate on. If the person using the AI doesn’t understand the fundamentals of what they are asking the AI to do, it will be challenging to ensure the work is done correctly and even more difficult to re-prompt the AI to fix bugs or clarify points. If you don’t understand the fine details of what you want to achieve, your prompt will probably not contain the details needed to allow an AI to generate what you’re after.
Image Generation
Wait, wasn’t that in the list of things AI does well? Yes, it was. AI image and video generation is useful, but it can also produce inconsistent results. Earlier image-generation models struggled to create something as ubiquitous as hands. Text is still often junk, and random, nonsensical objects frequently appear in AI-generated content. In videos, objects may spawn out of nowhere or morph, and characters and objects are often inconsistent from scene to scene.
So, are you looking for a job tending goats?
I say that in jest. Some individuals and small businesses will use AI tools to handle much of the heavy lifting in web design projects. That said, I think the lion’s share of such companies would be the same ones that employ DIY website builders like Squarespace, Wix, or GoDaddy Website Builder. Organizations that rely on their website to generate sales and leads will probably find the promise of AI-built sites a pipe dream. The AI will enable new efficiencies, but that will lead to increased competition. We’ll still need individuals who can think critically about how to solve problems, and not just produce something that looks and sounds like everything else there.
The AI Hangover – A Sober Reminder
Since AI content has become very easy to create, the internet has been flooded with it. When a deluge of low-effort, low-quality AI slop begins to make us despair, genuine human creativity may begin to shine, as we realize that a noisier world isn’t necessarily a better one. Even now, people have grown tired of the meaningless drivel pumped endlessly into their consciousness, and an anti-AI sentiment has arisen. That alone may bring us back to a more sustainable balance in using AI to enhance our creativity, rather than replace it.