What is AI? (no, fr)

You know how YouTube or Netflix know exactly what you might be interested in?
Or how your phone separates out people in your photos app?

That’s AI.

You’ve been using AI for ages, before ChatGPT was shoved down your throat in its name putting it in your fridge or AI-powered toilet paper (lol).

When you give any machine the “smarts” that required humans before, that’s AI (simplified heavily). Or in its fullform Artificial Intelligence.

Recommendation Engines and ChatGPT are both AI?

AI is a huge category, with multiple kinds under its belt.

Artificial Narrow Intelligence (ANI)

The kind of AI we have been using so far, and still are.

It includes various types of AI like:

• Siri/Google Assistant
• Recommendation Engines
• Email Filtering
• Camera systems (AI autofocus, enhancement, etc.)
• Security Surveillance systems

You get the idea.

Artificial General Intelligence (AGI, Sam Altman got wet reading it)

The one we are chasing, with all the advancement in the world of AI Labs like OpenAI, Google, and Anthropic.

This aims to mimic humans in all brain related aspects, while being capable human-like reasoning and learning from its actions. Also with advantage of being correct and optimal most of the times.

Artificial Super Intelligence (ASI)

This theoretical form exceeds human intelligence. But that’s just a theory (A Game Theory…)

AI has existed for ages, but people rarely knew these were AI, because the term caught on when LLMs became a thing in the mainstream.

The algorithms on your social networks, the grammar checkers like Grammarly (not sponsored), word predictors (put em on steroids and you get LLMs) and autocorrect, all are examples of AI which you silently enjoyed.

There are more kinds of AIs that most end users rarely think about, like in medical imaging, drug discovery, predictive maintenance in manufacturing, and more. But that’s a story for another time.