What we talk about when we talk about AI

For this week's case study, I took a look at how different groups of people are talking about AI. I compared public AI-related posts from (1) people who make decisions about technology in their workplaces (operating executives) and (2) people who professionally invest in technology (VCs).

Spoiler alert: They're different! And it's interesting!

Who I included

I hand-selected 100 tech influencers who are regularly posting about AI, a mix of operating executives and investors. This is how I picked:

• I had to pick a platform. At one time, the choice would have been obvious (RIP, Twitter). Now it isn't. I went with LinkedIn because I use the platform enough to understand it and because people mostly use it to talk about work.
• I focused on people who are current operating executives and/or professional investors. I did not include any other kinds of AI commenters (consultants, politicians, media, etc). Maybe I'll look at these groups later on.
• I picked people with at least 10,000 followers who are posting about AI at least once a week.
• Everyone I picked is based in the US. I may look at other regions later on. There were clear regional patterns within the US data, which I might share in a future case study.
• Within the constraints of the above, I attempted to get some diversity of voices. Interestingly, this was easier for personal demographics like gender and race than it was for geography.

Using these guidelines, I picked 100 AI influencers and analyzed their posts. I could have easily picked a different 100. However, the patterns I found are striking enough that I suspect I'd have found similar insights.

What I looked at

After I picked my people, I analyzed their posts from the last three months. First, I classified their AI-related posts into topic areas. The set of topics in AI-related posts is rather limited, with 93% of all posts fitting nearly into five main areas:

• Personal experience using AI
• Business impact of AI
• Ethics / social impact of AI
• Deep tech observations
• Privacy /security

Broadly speaking, I wanted to see whether the operators and the investors are talking about the same stuff when they talk about AI. I also wanted to compare their overall sentiment towards AI: Are the posters excited and enthusiastic? Are they anxious and stressed?

Privacy and security keep the executives up at night; the investors, not so much

Unsurprisingly, the business impact of AI is the most common topic for both groups, followed by people sharing their own personal experience using AI apps.

But check it out: The executives are talking way, way more (like 5.5x more!) about privacy and security than the investors are. The pattern among executives matches what I expected anecdotally. I don't know many business leaders who aren't worried about what the rise of AI means for the security of their data.

Given the struggle that most organizations currently have selecting and vetting AI solutions for data privacy, I expected more investors to be talking about this. Their B2B portfolio companies are definitely feeling the struggle. I found this gap between executives and investors striking.

What are investors talking about in place of data privacy and security? The long tail of AI topics discussed by investors (that "Other" category) is much larger than it is for executives. Long-tail posts are about things like AI policy, IP rights, AI and government, predictions about the future of AI, and that tried-and-true classic: updates about one's own AI investment portfolio.

No single one of these "Other" topics was prevalent enough to form a whole category on its own. But all together, the long tail of topics made up 12% of the investors' total AI posts, compared to just 2% of the executives' posts.

Investors are excited! Executives are excited too, but also kinda worried

The easiest way to visualize the sentiment differences between executives and investors is with simple word clouds. The same words showed up over and over again among both groups, but in very different proportions. First up, the executives:

The word worried shows up in the executives' posts more than any other single sentiment, followed by nervous and excited. Exec posts contain a range of sentiments, and they often include apparently conflicting sentiments within the same post.

Now let's look at the same visualization for the investor group:

Across the board, sentiment in investor posts is much more clearly and universally positive.

Given that the execs are much more likely than the investors to be talking about privacy and security, their sentiments make sense. The data privacy considerations with AI are legitimately worrisome.

True story: A few months back I talked to a business exec at a major hospital who shared that their doctors had been pasting real patient notes into ChatGPT to write case summaries faster. If you were the CISO at that hospital, you'd be freaking out too.

What do you think? Are you surprised by these insights?

Why do you think the investor group is so much less focused on privacy and security than the operating executives are? I'd love to hear your theories.

I'm going to do a post like this every week, so tell me what you thought of this one. What else should we look at? Should we break down the blue state/red state AI divide? Leave me a note on nerd processor to let me know.

And if you liked this, tell all your friends to subscribe too!​

Thanks for reading!

Kieran

kieran@nerdprocessor.com
​Unsubscribe · Preferences​