Thanksgiving Reflection: One Year of ChatGPT Changed Everything

One year ago today, OpenAI released ChatGPT. Within a week, it had a million users. Within two months, a hundred million. It became the fastest-adopted software application in history.

And professional services has responded... cautiously. Pilots are running. People are experimenting. But compared to the hype, adoption is still spotty.

I want to reflect on what's actually happened in the last year, and what I expect in the next one.

What Did Happen

The Enterprise Tools Got Better — Claude got released and quickly became better than ChatGPT for professional services work. ChatGPT Enterprise launched with real compliance features. Google Gemini is coming. The level of AI capability now available to firms is genuinely impressive.

Adoption Started at the Edges — First it was hobbyists and early adopters. Then consultants who used it on the side. Then firms started endorsing it for specific use cases. Most adoption is still bottom-up, not mandated from leadership.

Pilots Worked Better Than Expected — The firms that actually ran pilots (not just tried ChatGPT on their own) found real value. Email triage works. Document drafting works. Meeting summaries work. The technology delivers.

Regulation Started Moving — The EU AI Act progressed. The US Executive Order was issued. Regulatory uncertainty went from "we don't know what's coming" to "we know what's coming, we just don't know exactly when."

Vendor Risk Became Real — The OpenAI chaos in November showed that "pick the best AI platform and integrate it deeply" is risky. Firms started thinking about vendor diversity and portability.

What Didn't Happen

Mass Replacement of Human Workers — The doomsday scenario of AI eliminating professional services jobs hasn't happened. If anything, professional services firms are still hiring.

Autonomous AI Agents Doing Legal Analysis or Audits — Despite the hype, we're not at the point where you can ask an AI to "analyze this acquisition for tax risk" and get back a comprehensive analysis. AI assists human judgment; it doesn't replace it.

Dramatic Productivity Improvements — Some firms have seen 10-15% productivity gains in specific workflows. But the "AI will double your productivity" crowd? That hasn't materialized. Real improvement comes from thoughtful implementation, not from having AI.

Complete Rethinking of Professional Services Economics — Some firms have adjusted their business models around AI. Most haven't. Billing by the hour is still the norm. Utilization targets haven't changed. The profession is adapting slowly.

What I Expect in the Next Year

Consolidation of AI Tools — Firms that have been experimenting with 6-8 different AI tools will consolidate to 2-3. The fragmentation will decrease. Enterprise tools (ChatGPT Enterprise, Claude, Google Workspace AI) will dominate.

Regulation Will Get Real — Not comprehensive US legislation yet, but agencies will publish guidance. Industry regulators (SEC for finance, state bars for law) will clarify what's allowed. "Gray area" will shrink.

Skills Shift — The ability to use AI effectively will become table-stakes. Firms will start valuing "AI fluency" in hiring and promotion. Training will shift from "here's what AI is" to "here's how we use it in our workflows."

Custom AI Will Proliferate — Custom GPTs and Assistants-based applications will become common. More firms will build AI applications tailored to their specific workflows instead of trying to fit their workflows to generic tools.

Some Ambitious Firms Will Get Meaningful Advantage — While most firms are still experimenting, a few will have deployed AI across multiple workflows, measured real ROI, and built competitive advantages. These firms will be noticeably more efficient. By 2025, this will be visible in their economics.

The Realistic View

AI isn't a magic bullet. It's also not overrated. It's a tool that, when used correctly for the right problems, delivers real value.

One year in, we've seen that the tool works. We've seen that adoption is slower than the hype suggested. We've seen that firms that approach it thoughtfully get better results than those that approach it as a fad.

If I were a managing partner right now, I wouldn't panic that I'm behind. But I also wouldn't assume I have time to wait and see. The next 12 months are when ambitious firms will create advantages that will take others years to catch up with.

That's the real story of this year of ChatGPT: it proved AI works for professional services. The question now is who will actually use it.