Am I an AI Champion?

Recently, a friend gave me a compliment for being an AI champion, and that gave me pause. 

I started my AI learning journey in 2021, before people began conflating AI models with large language models interchangeably, and before anyone had to defend their use of an em dash in a blog post. The product I was working on is now known as Intel Geti, and it still has a great team at Intel supporting it.

That AI product was a computer vision model training solution. It transformed model training from a laborious and time consuming task into a simple one: you marked what you saw on the screen, and the computer learned and fed back what it was learning: simple and effective “human in the loop” model training.

Frankly, it was totally mind-blowing, and helping take that product to market is still one of the high points of my career. But thanks to an early round of Intel budget cuts, I lost my scope and moved on. 

Cut to earlier this year: I’d still been working in software at Intel, but had some free time on my hands. After burning my way through several sitcoms I missed when they were live, I complained to a friend about how I felt like my brain was rotting, and she suggested I see if our local library had a LinkedIn Learning or Coursera option.

I know a good idea when I hear it, and found some solid courses to help me better understand foundation models (both LLMs and the diffusion models powering many image generators) from conception to use.

 That understanding helped me start experiments with foundation models to see how they might fit into my work. I’ve found some places where they help me, and others where they just seem to create confusion or waste my time. 

I started posting about AI because I’m still trying to figure out what exactly I do think about it. And as a writer, doing that “out loud” for me means writing about it. I’ve found a few hooks into it where I think I have something to contribute (like being an AI editor), and I’m exploring those for now to see where they take me.

If I had to describe my position about AI, I’d consider it bifurcated. There are truly amazing use cases being made possible, and then there’s LinkedIn offering to rewrite an already written post with AI. In the category of LLMs, I’m considering myself optimistically skeptical. I’m finding value, but also many limitations.

And if I had to champion anything, I think my position has become, “find the workflow that works for you.”

• You want to opt out? I’m not going to join the chorus of people who tell you your job will be taken by someone who does the same thing but uses AI. I can’t see your future.

• You want to use it for research, brainstorming, or even your first-draft writing? Also no judgment. Just be aware of the limitations and be prepared to still put in some brain power in the process.