It feels like GenAI is taking over the world. It is everywhere on LinkedIn—in automated comments, poor-quality AI-generated content, and incredibly high-value content like this.

I’m hearing from a lot of people that they’re worried they’re falling behind because they’re not using it day-to-day. My general reassurance is that given two weeks of regular use, you can overtake most people in GenAI usage.

Here are three practical things you can start doing today:

Put it where your eyes and thumbs go

People do convenient things, so make GenAI convenient for you. Place it where you spend most of your time so you see it and it’s easy to select.

For most people, that means popping it on your phone’s Home Screen, placing it in your dock if you’re really keen, and pinning it as a tab in your browser on your laptop.

These simple things mean you’ll notice your preferred AI tool more often. Your usage will increase.

Prioritise the task, not the tool

The point of the general AI assistants is to help boost your productivity. If you have a task that will take 30 minutes and spend an hour using AI to produce a poor first draft, you’ve prioritised the tool instead of the task.

If you’ve decided to build a custom GPT or workflow for this task, by all means go deep. Invest the time. The effort will pay off over time.

If you were trying to save time right now, something’s gone wrong.

Next time, fire up the GenAI and get your task started. If you don’t make good progress working five minutes, cut your losses and get back on task.

The search for quick wins includes the discovery of many quick losses.

Try including what you don’t want

As a rule of thumb, people are terrible at knowing what they want and even worse at expressing it. This is demonstrated by that classic piece of creative feedback: “I’ll know it when I see it.”

This human trait is a challenge when using GenAI. The less clear you are about the expected outcome, the less reliably sound the output will be. So, how can you overcome this challenge.

Try being clear about what you don’t want. Write a prompt that says, “I want this, and I do not want this.” This narrows down the range of possible good outputs without you having to go into granular detail about the exact output you’re looking for.

Try these three things over the next couple of weeks and I’m confident you’ll feel less behind, and realise you’re not behind at all.

We’re all busy integrating AI into our lives and the products we build. One thing that is clear is that AI adoption by users is not guaranteed; there’s emerging data that far more people use general models as a novelty rather than incorporating them into their day-to-day activities. This is not surprising and to be expected. The challenge is: how do you get the tools you’re building to be locked into users’ lives over the tools others are building?

Optimization vs. Adoption

For years, outside of those building new products, most UX professionals have been focused on optimizing mature systems. We’ve been laser-focused on refining workflows, polishing interfaces, and enhancing usability to deliver smoother, more intuitive user experiences. Optimization has undoubtedly played a crucial role in improving user satisfaction, engagement, and performance.

However, with the advent of AI, we are now facing fundamentally new workflows and outputs. AI doesn’t just enhance existing processes; it transforms them, often creating entirely new modalities of interaction. This requires a pivot of mindset from optimization to growing adoption—ensuring users embrace these new tools.

Delivering Success Through User-Centric Design

To successfully embed AI into our tools, I think we should prioritize user adoption. This involves designing experiences that are not just functional but deeply intuitive and engaging. We need to create environments where users feel confident and excited to explore these new technologies.

I’d argue that with AI, excitement is there. Confidence, however, is a composite of factors, including consistency, predictability, and successful outcomes. Of these, the latter is, I think, the most important for encouraging repeated use.

So, AI-powered tools need to deliver successful outcomes to encourage adoption. What success looks like will be different for every product, but ultimately, it matches the phrase, “I easily got the output I was expecting.”

Continuous feedback loops remain critical

Many of the adoption workflows are similar to optimization workflows. They are both iterative processes. Establishing continuous feedback loops ensures that we are attuned to user needs and challenges. Regularly soliciting and acting on user feedback allows us to iteratively improve the AI experiences we design, fostering greater trust and satisfaction.

A new priority but the same priority

Focusing on adoption is a new priority for many, but the true focus remains on those using the products we build. It’s just that the best thing we can do to support users as AI rolls through is to create systems that guarantee success in order to drive the adoption of this exciting new wave of technological advancement.

Here’s a list of interesting things I read in March.

1. Unleashing the power of data visualization in music debates

Imagine walking into a bar and sparking a conversation by contemplating, “What defines the greatest album of all time?” This exercise in subjectivity resonates with many, yet often dissolves into the mumblings of agreeing to disagree. Enter The Pudding, an online journal that doesn’t just start these debates — it arms you with the sharpest stats to win them.

Have a gander: What makes an album the greatest of all time?

2. Exploring Negative Capability

Developed by John Keats, ‘Negative Capability’ is the ability to exist in a state of intellectual uncertainty and mystery. Lack of resolution does not disturb the mind in pursuit of the truth. In the backdrop of tech’s relentless push for definitive solutions, this essay by Ness Labs was a welcome respite for those swimming in the sea of innovation.

Read it: Negative capability: how to embrace intellectual uncertainty

3. Google, Gen AI, and publishers

A recent Adweek piece revealed Google is paying some in the digital publishing world to experiment with an unreleased Gen AI platform. Google is investing so much in so many areas, it’s hard to keep up!

Check it out: Google is paying publishers to test an unreleased Gen AI platform

4. The AI Learning Labyrinth: Selective Forgetting

The success of artificial intelligence often becomes a rabbit hole of datasets and forgotten inputs. Quanta Magazine’s piece on selective forgetting in AI learning really stood out. The crux of the essay pivots on the hypothesis that pruning excess information can actually clarify model learning.

Read the essay: How selective forgetting can help AI learn better

5. BBC AI usage guidance

Trust is the lifeblood of any media organization, and the BBC, with its responsibility to inform the public, wields it with the gravity it deserves. With Artificial Intelligence seeping into the editorial process, the corporation’s Guidelines on AI Usage aren’t just another rulebook; they’re a benchmark against which to reflect on our own

Scan the guidelines: BBC guidance: The use of Artificial Intelligence

6. Balancing the bias in Generative AI

Adobe shares its thinking about how to reduce biased and harmful outcomes from generative AI. The piece went beyond the token acknowledgment of AI biases, offering a blueprint that advocates for a combination of cultural competence, transparent development, and ongoing monitoring.

Understand Adobe’s approach: Reducing biased and harmful outcomes in generative AI