With the dawn of ai, I can't help but think about the similarities with other monumental leaps forward in technology.

Specificity

The internet: Right now, we broadly say "ai". I image it was the same at the beginning of the internet. "Did you send that through the internet?" "Did you make that with the internet?" Sounds weird. Because now, virtually everything is connecting or using the internet is some way. So instead, we refer to the specific tools we use. I imagine it'll be the same with ai. It's not going away. It's also going to be less isolated. Soon, like the internet, it'll be integrated into nearly every digital product we use, we won't even really be able to tell what's technically 'ai' and what's not.

I don't feel as though it's far away when we won't manually interact much with computers.

We'll just ask our local, built-in ai to do something for us.

The Risk

When I think about the possible data security and redundancy risk, which is clearly present, I think we'll keep doing what we've done since the first computer: trade off some risk for convenience:

First, we stored physical, printed files in file folders.

Then, we were able to store files digitally, but local on a single computer. The risk? The hard drive could be corrupted, computer could break, files could be deleted. If internet-connected, could be hacked and deleted or stolen. Clear security and data redundancy risks, but eventually we accepted those in exchange for convenience in technological advancement.

Next, we could use proprietary online tools like Notion, Asana, Google Docs, etc. We didn't own the files at all, we paid for rented access to a better (and again, more convenient) way to access our data through another computer's method. The tradeoff? No different. Even greater data security and loss risk in exchange for efficiency and technological advancement. But eventually, we accepted the risks for the rewards. Some tools exist (Obsidian, NAS systems like Synology to own your own cloud...) that still fundamentally believe in in the older way and have a loyal user base who believes the same.

Now, we arrive at the cusp of ai. The possibilities are growing by the day. Some people are psyched, some terrified. Many a mix of both. Some people swear it's the future, some swear the bubble will burst. I think we'll do what we've always done at a key innovation point: Eventually accept the risks for the reward.

Apple

What did Apple start doing years ago? Building their own highly performative chips, with built-in ai processing.

Today I downloaded LM Studio, with Gemma 4, a local ai trained similarly as Google's Gemini, only taking up about 16gb of disk space. The capabilities of this are astounding for something that can run on-device, locally, without an internet connection.

Once something like this even has slight agentic capabilities, features likes Spotlight will quickly be able to perform pretty much any task on a computer by typing in plain language as if you're talking to another person.

Exactly what ai did for Google search, an already highly used product. Previously functional, but now can summarize all the results for your specific query and ask follow-ups.

Probably what will start to become very common is people who can train ai to do tasks. I could likely pretty easily train an ai to create entire social media sermon clips. It will soon be no different than writing out a step-by-step process for a course, or internal documentation. Because it's repeatable.

Could ai develop a concept and a creative video opener piece for a church service? Probably eventually. That's less clear. But having an ai run locally and know what buttons to click that are very repeatable each time? Seems pretty easy to me. I can already drop the transcript into ai and it'll choose ideal sermon clips based on timestamps easily.

Maybe there's a market for people who can come in and implement ai to automate tasks.