There’s a lot of talk about AI replacing and transforming jobs, which has some kernels of truth alongside a whole lot of AI-hype. It’s hard to throw an AI-generated rock without hitting a newly minted “AI Expert” or “AI Strategist”. Some of these folks have actually intelligent things to say, but many of slapping “AI” on their title, like so many software companies are saying “now with AI!” about their products. Some people are telling you that you will leave yourself behind unless you dive headlong into AI yesterday.
What’s the real story?
First, unlike a lot of these experts, I will tell you that I can’t predict the future, and I don’t know exactly where this is going.
And note that for our current purposes, “AI” means the generative AI tools like ChatGPT, Claude, Gemini, etc, that have dominated the cultural conversation (and the venture capital and other forms of investment), not the many other tools that also fall under the “AI” umbrella.
I think the right way to think about AI is as a free or very low cost intern that never complains, never gets tired, never gets frustrated, but also insists on working 100% remotely and only conversing via messaging.
You might ask this intern to:
- Summarize and transform data (including, for example turning meeting notes into the outline for a presentation or webinar)
- Do competitive research
- Brainstorm ideas
- Suggest solutions to problems
- Draft content (I probably wouldn’t, but some people would)
- Review content for spelling, grammar, and wordiness (and to check if you’ve forgotten something obvious).
- Summarize notes into status reports and meeting notes**
You should not ask this intern to:
- Impersonate you in dealings with clients, partners, and prospects, including those awful AI email and LinkedIn campaigns. You may think you’re saving time, but in a business that relies on trust, you’re actively destroying trust.
- Handle super secret client information (unless you are using secure, on-premises tools that don’t ever leak that data into the cloud).
- Create client-facing documents without review like the lawyer whose AI-drafted motion referenced nonexistent cases (this Stanford study found 114 similar cases, mostly in solo or small firms), or Deloitte, which submitted a report to the Australian government citing nonexistent academic research papers and legal opinions. In addition to the reports, this applies to emails, proposals, blog posts, etc. It may seem tempting to have AI write those for you, but AI isn’t going to create something that really reflects your perspective (I have followed suggestions on how to train AI to use “my voice”, but haven’t been happy with the results). Note that having AI read a bunch of your content and ask it questions about your perspective could be very helpful.
All Models are Wrong, Some Models are Useful
This applies to both AI models, and the simplified version of “AI as intern”.
On the plus side, unlike real human interns, AI:
- never loses focus from exhaustion or frustration
- never needs a break
- can process huge amounts of data quickly
On the negative side, AI can:
- Make up, or as the AI folks like to call it “hallucinate”
- Fail in ways that are hard to predict (and some times hard to spot), usually without asking for help like a human intern might
- Give you a false sense of security (“that’s a great idea! Here’s how you can make it happen… [that happens to be completely wrong, as spoofed in a recent “Sickofancy” South Park episode]). Even in the audio mode, which is very handy, you’re not talking to a person, but a computer, and you don’t get the usually social cues.
- Accidentally leak sensitive data
- Go around in circles when following your prompts, sometimes costing you more time and effort than other approaches
So no need to panic, but I suggest that if you don’t already have an “intern”, you bring one on not to replace your or get in the middle of your human relationships, but to help handle a lot of the grunt work that it can do much faster than you, freeing you up to focus on more valuable work, including reviewing what your “intern” has produced.
The AI Bubble
This post is just about the practical applications of AI for solo consultants, not about generative AI overall, the hype bubble, the copyright and environmental issues, the non-profit bait-and-switch from “Open”AI, the dangers of outsourcing content creation to a society already flooded with attention destroying garbage, how spending on GPUs and data centers is propping up “the economy” on flimsy foundations, or any of that stuff…
** I love Fathom’s meeting summaries, and paste them into my meeting notes in Mimiran. I like that they have clickable timestamps that let you jump back in and review a section of the conversation if you want. The AI is a huge help and means you can focus more on the conversation, rather than note-taking, but I still recommend setting up custom bookmarks and using them to flag key parts of the conversation. Check out my Sales for Nerds interview with Richard White, CEO of Fathom, for more details, including my bookmark highlighting options.)