September 5, 2026

September 2026: Way of work

Documenting the tools, processes, focus, and thoughts of the design work that I'm doing as of September, 2026.

---

Despite AI being the future, my day to day work has yet to really change. I still am using mostly the same tools. Though, each now has a little text input that helps me (sometimes) do my work a little faster. Daily, that means I'm spending my time across a WYSIWYG design tool (Figma), a code editor (Cursor), and a handful of communication tools (Slack, Docs).

The human processes surrounding the work are familiar, too. The most noticeable process change is the sense of urgency instigating seemingly everything. This urgency is only unfamiliar in the realm of inhouse roles. It is something I felt often while working at agencies.

The current headlines have me believing there should be more change occurring within my own role at this point in time. I have long been anticipating it. These headlines makes me think it will be a swift chop to the back of my knees. I suppose everyone could be sensing this. Maybe, this is driving the urgency I spoke of above?

Organizational thoughts and curiousities

In regards to design and my role, there are a few things occurring that I am particularly interested in seeing play out in the near future:

1. When will AI meaningfully change the processes I work within?
2. How will companies react to the intersection and overlap of roles?
3. Will UI design still be a necessity for most future software?

When will AI change processes?

I think we have only gotten glimpses of AI's efficiency promise. So far, the conversation has focused on task-oriented workflows within pre-existing processes and is measured in terms of personal performance. This results in questions like:

• How has AI made your own task-oriented process faster?
• How has AI enabled you to do more kinds of things?

Though, throughout my career, technical task completion was never the largest time vacuum. Most time tends to be spent upfront, concocting solutions, evaluating paths forward, aligning on decisions, creating timelines, figuring out resourcing and such. For work to feel different while doing it, might this be the part that would need to change?

In current times, where tech work is driven by product managers, it makes sense that LLM functionality has initially focused around providing non-technical contributors more ability to execute. This approach clearly provides the architect (e.g. management) more value.

It is unlikely any role will volunteer themselves as the root cause of inefficiency. We all need to make a living. Disclosing we are the problem will not aid in that. All this to say, it is reasonable and fair for product managers to have oriented the LLM features they are building to handle other role's skills versus replacing their own—we probably would have done the same.

Now, since LLMs are already quite good at making designs and writing code, what are the chances we backtrack to focus on replacing planning work in the near term?

How will companies deal with role overlap?

I ask myself, "If this was my business and this was the context, what would I do?" I think the most likely outcome here is that planners start owning execution. Why pursue planning and management automation if the fruits of design and code generation have yet to be picked? I'd predict this results in a reduction of labor/overhead (to save money)—and assumably saves some time by removing the hassle of resourcing.

As a designer, it is hard to understand exactly where we fit in this. We kind of do both? A design is almost synonymous with a visual plan. Sometimes, at some companies, designers act as product managers. In other situations, designers are treated similar to engineers and are resourced after a solution has been identified. This context probably matters when considering how things may shake out.

To increase my own chances of job security, I'll continue adopting as many product management responsibilities as I can. Besides it being kind of fun to wrangle solutions in this fashion—at the moment, it is my best guess and strongest bet for career longevity.

Today's LLM design capabilities may make this all sound like a stretch. Currently, LLMs are not that good at design or code. Sometimes, they make heinous UI crimes and it takes ages to massage the output. The rate in which LLMs have improved year-over-year has me convinced this will change rather quickly, though.

This leads me to my next question.

Will companies need their own UI?

Once LLMs are able to generate great UI on the fly, do most companies even need their own UI or is it feasible that a user engages with company-specific systems and data from one, unified place?

As a designer focused on the top-of-funnel at a B2B SaaS company, I often think about this in the context of getting a user to sign-up. A much sleeker solution to visiting a marketing or docs site, learning about a product, manually creating an account, and then manually configuring it is to simply ask an LLM to do all of this in less than a sentence.

This holds true for most of the B2B software that powers the things (websites) I help build on a day-to-day basis. For example, evaluating and implementing tools like content management systems, analytics providers, and hosting platforms.

Now, this puts in question whether the things many of us are building are needed anymore in the first place. I have written specifically about B2B SaaS marketing site's relevance in an AI-powered world. I have no real answers, only guesses.

My guess is that if a company doesn't account for mission critical, required infrastructure that is incredibly difficult to build on one's own, it could be in trouble. For example, if an analytics provider's primary customer is other B2B SaaS companies—what happens when users no longer visit the customer's websites or dashboards and solely engage with it via an LLM? Is there still a need for this type of analytics?

I have a harder time seeing this hold true for B2C companies with physical products. While agentic commerce is a thing, I believe people are more concerned with brand and quality. Purchasing behavior, at least in part, has become part of a person's self identity. I cannot imagine most people wanting to defer judgement and exploration here.

When it comes to business software, something like the CMS I use does not factor in to who I am (well, at least I hope it doesn't) 😅.

Reflections

Again, this is all just what I'm currently thinking. I am probably wrong on a lot of it. The timing in which this might occur is unknown. I feel like I've been posting a bunch of doom and gloom content...

This sounds dire for my own employment, but I am not that upset over it. To me, tech gigs have always seemed like a fluke. Like, why am I being paid this much to sit at a desk, within my own home, to do relatively easy work? At times, I have forgetten the underlying goal of technology is almost always automation, advancement, and efficiency. Along the way, I started to enjoy sitting behind screens more than I should. None-the-less, the idea of being out of a great job can be scary.

But, future me probably would enjoy doing something more with my life than sitting behind a desk for 9-10+ hours every day. The pay I'm provided for doing whatever that is may not be comparable—that's ok.