2023: the year AI became a conversation topic

My grandmother called me in March and asked me what ChatGPT was. She’d heard about it on the news. She wanted to know if it was safe to use. She wanted to know if it would take my job.

My plumber mentioned DALL-E while fixing my sink in July. “My kid used it to make pictures for a school project,” he said. “Pretty good pictures.”

At Thanksgiving, my cousin asked me if AI would take his accounting job. He wasn’t joking. He was worried. He’d seen articles. He’d heard things. He wanted someone in the family who “knows about computers” to tell him the truth.

I didn’t know the truth. I told him it would change his job before it replaced it. I’m not sure if that’s right. I said it because it’s the most honest thing I could say while still being reassuring.

The crossover

2023 is the year AI crossed over from tech topic to everything topic.

OpenAI started the year with ChatGPT already at 100 million users. Then GPT-4 launched and it could see. Anthropic launched Claude and it was good. Google launched Gemini and it was competitive. Meta released Llama 2 for free. Midjourney and Stable Diffusion put AI art generation in the hands of anyone.

By December, my grandmother, my plumber, and my accountant cousin all have opinions about AI. They don’t agree. They don’t fully understand it. But they know it exists and they know it matters.

That’s the crossover. When your grandmother has an opinion, the technology is no longer niche.

What crossed over

It’s not just the chatbots. The conversation expanded:

Work: “Will AI take my job?” became the question of the year. Every profession ran the calculation. Lawyers, writers, programmers, artists, teachers, accountants. Everyone asked.

Education: teachers are dealing with students using ChatGPT for homework. Some schools banned it. Some integrated it. Nobody ignored it.

Art: the “is AI art real art?” debate ran all year. Gallery shows. Competitions. Lawsuits. The question is still unresolved. It might never be resolved.

Trust: the deepfake conversation got real. AI-generated images and video became good enough to fool people. The question of “is this real?” applied to photographs for the first time in history.

Safety: the OpenAI board drama put AI governance in the news. Suddenly, non-technical people were discussing who should control AI development.

Pew Research numbers

Pew found that 58% of Americans have heard of ChatGPT. Among adults under 30, it’s over 70%. A third of people who’ve heard of it have tried it.

Those numbers will only go up. The tools are getting easier. The integration into existing products (Google, Microsoft, Apple) means people will use AI without choosing to. It’ll just be there, in their email, in their search results, in their photo editing apps.

What I feel at year’s end

I feel the same way I felt in 2007 when the iPhone launched and I could see that phones were about to change everything but couldn’t see exactly how.

AI is changing everything. I can feel it. I can’t see exactly how.

My grandmother asks me about it. I explain as honestly as I can. She nods. She doesn’t fully understand. Neither do I. The difference is that I know I don’t understand, which might or might not be better.

2023 was the year AI became a conversation topic at every dinner party. 2024 will be the year it becomes a tool at every desk.

After that, I don’t know. Nobody does. That’s the honest answer I keep giving my cousin, and my grandmother, and everyone who asks.

I don’t know. But I’m paying attention.

Related thinking: