AI, Explained: A Human Guide
Where AI came from, what the words actually mean, and where you have been using it all along. From a practitioner who has been working with it since 2010.
Updated as AI changes. Last reviewed October 1, 2026.
Ryan Vet is quoted in
You have been using AI since 1980
Long before the chat box, the quieter kind was running inside things you already owned. Some of it is real machine learning. Some of it is clever rules. Knowing which is which is the whole skill.
-
1980
Pac-Man ghosts
Each ghost follows its own chase rule.
-
1992
Credit card fraud alert
A neural network scores every swipe against your normal pattern.
Machine learning fico.com : FICO Falcon, 1992 (opens in a new tab) -
1993
AutoCorrect
Swaps common typos for the right word from a list.
-
1996
MapQuest directions
Calculates the shortest route across a road map.
-
1998
Google Search
Ranks pages by the links pointing to them.
-
1998
"Customers who bought this…"
Amazon recommends items often bought together.
-
1999
T9 texting
Guesses your word from keypresses using a dictionary ranked by how common words are.
-
2000
Netflix recommendations
Predicts your rating from people with similar taste.
-
2002
Roomba
Bumps, turns, and spirals by fixed rules.
-
2003
Spam filter
Learned what spam looks like from millions of labeled messages.
-
2005
Camera face detection
Finds faces in the frame and focuses on them.
Pattern recognition dpreview.com : Nikon face-priority AF, 2005 (DPReview) (opens in a new tab) -
2006
Google Translate
Learned translation from already-translated text.
-
2007
Live traffic in Google Maps
Pools phone speed data to predict traffic and arrival times.
Statistical, later ML blog.google : Google Maps traffic and ETAs (opens in a new tab) -
2008
Search autocomplete
Predicts the rest of your search from what others search.
-
2010
Facebook tag suggestions
Face recognition suggests which friend is in the photo.
November 30, 2022
Then it got a box
ChatGPT put all of this behind one plain text box, the same box for a teenager and a grandparent. That box is why AI suddenly felt new, and it is why every generation now meets it the same way.
When people say “AI,” they usually mean one small part of it
ChatGPT showed up on November 30, 2022. The field is centuries older than that, the name goes back to 1955, and the quieter kinds have been running inside your email and your credit card for decades. Ryan puts it this way:
“People call ChatGPT AI, which is true, but it’s not all AI. It’s kind of like calling a paper airplane a rocket ship.”
- Artificial intelligence the whole field, named in 1955
- Any machine doing work that normally takes human intelligence: recognizing, predicting, deciding, and now creating. Pac-Man’s ghosts count. So does Face ID.
- Machine learning it improves itself
- Software that finds patterns in data and gets better without anyone rewriting the rules. Your spam filter learned from millions of messages people marked as junk.
- Generative AI it makes new things
- The kind that writes, draws, speaks and codes from a prompt. This is the slice that arrived all at once, and the slice people mean when they say the word.
When someone says “we use AI,” your next question is which ring they mean. All ten terms, in plain language
Millennials use it most. The youngest adults trust it least.
Who actually uses it
Weekly AI use, by generation
Barna Group, “Hesitant & Hopeful,” published Jan. 2024 (surveyed July and Aug. 2023). The survey is a few years old, which is itself worth knowing: nobody is measuring this as fast as it is moving.
Millennials lead. They are the generation that taught itself Photoshop and Myspace. If you want to know who can help your team with this, look for who is building and experimenting, whatever their age.
Who is souring on it
Different studies, different age brackets, kept separate on purpose
- 55% of U.S. adults under 30 are more concerned than excited about AI, up from 31% in 2021. Pew Research Center, Aug. 18, 2026
- 22% Gen Z excitement about AI, down 14 points, while anger rose to 31% and weekly-or-daily use held at about half. Gallup with Walton Family Foundation and GSV Ventures, 2026
The souring is concentrated in the young, and they are heavy users too. Using a tool every week and trusting it are two separate things. Everyone gets the same box now, so what still divides the generations is trust.
Original research, Fall 2026
We read every AI speaker the bureaus list
Everyone says AI speakers are everywhere. Nobody had checked what is behind the listings. So Ryan did, and published the whole thing: the rubric, the aggregate data, and the parts that did not come out the way anyone expected.
-
33%
Practitioner or Practitioner-Researcher
showed publicly verifiable hands-on AI work, either on its own or alongside disciplined study.
139 of 423 people
-
61%
Continuity
had publicly verifiable AI work dated before ChatGPT went public on November 30, 2022.
258 of 423 people
-
25%
Buyer evidence
showed none of the three things a buyer can check: stated scope limits, public client references, a described update process.
104 of 423 people
Every AI speaker listed by 6 major North American speaker bureaus was read, 423 people in all, and checked against public, sourced evidence of hands-on AI work, of disciplined study, and of AI work predating ChatGPT. Not a sample: the whole list.
Who is telling you this
Ryan Vet has been working with machine learning since 2010, twelve years before ChatGPT. He designed the logic behind matching and prediction systems at four companies, two of which were acquired, and he now teaches a university course on humans and machines. He is not a researcher claiming peer-reviewed AI papers, and this hub says so plainly on the page where he runs himself through his own rubric.
- 2010 First machine-learning pitch deck Twelve years before ChatGPT
- 2019 Boon’s matching algorithm Covered by Kingscrowd and PR Newswire
- 126 Published audience reviews Four specifically about the AI talks
- 2027 Humans and Machines A course he designed, William Peace University
Ryan Vet is quoted in
Pick a chapter
- 02 The AI Timeline From a 1642 calculating machine to ChatGPT, with the dates checked.
- 03 You’ve Been Using AI All Along Spam filters, fraud alerts, T9. Can you tell what is really AI?
- 04 Say It Right AI, generative AI, LLM, chatbot, agent. What each one actually is.
- 05 Questions, Answered The questions leaders ask Ryan most, answered straight.
- 06 Every Generation, Same Chat Box One interface for everyone. The divide that is left is trust.
- 07 Real or Fake? Voices and faces clone in minutes. How to verify on purpose.
- 08 My AI Story From a 2010 pitch deck to a university course on humans and machines.
Outside this guide: The Velocity Gap, Ryan’s framework for why this time the speed itself is the problem, attributed to him by name in Newsweek.
