AI

I have to confess that I’ve been cynical about AI. Which is a bit weird, given my tech background and inquisitiveness.

But I’ve seen some demos recently which are pretty interesting. Here’s what I learned: these tools don’t replace human intelligence. But they can simplify drudgery.

There are a couple of places which are particularly useful: Pulling together tons of data, and creating summaries.

On occasion, I’ll spend hours doing Google searches and diving into trusted websites to look into a new unfamiliar area. Just investigating the general landscape to find out important trends and how far I need to explore.

AI can be surprisingly good at this, because it has stitched together millions of websites and can determine what’s been said about the topic. Then convey that in a simple English summary using approachable language.

When creating a summary, like of meeting notes or something, it can quickly capture the essential points – topics, decisions, action items, and so on. If you do it with a Zoom call, it immediately knows who said what and can attribute its points to particular people.

I can totally see how this is useful, especially having done this work myself in various settings.

So what’s the downside? Well, AI (like humans) can make mistakes and misinterpret things. It may well not know which websites are trustable and which are driven by emotions or malicious intent.

I’ve had this happen when I was taking notes from a meeting. I thought I was capturing what someone said, but later found out that it’s not what they thought they intended. This happens all the time in human conversation, but AI usually doesn’t have the opportunity to go back and ask what was behind a statement.

That might come, I suppose.

AI was not used to create this post!