Below the waterline: The "AI iceberg"
Some things we may see, and often don't see, where diverse applications of AI and machine learning models are affecting our daily lives (or will soon)
The AI Iceberg
Most of the chatter about “AI” is about a few small subsets of AI. But there’s so much more to AI than those visible applications. AI applications are like an iceberg. A few types of AI are just the parts that are ‘above the waterline’. The rest of the types of AI and machine learning tend to be hidden ‘below the waterline’, invisible to most people. That doesn’t mean they can’t sink our ships, though!
Here’s a quick look at what’s above and below the AI iceberg waterline.
What’s above the waterline of the AI Iceberg?
Generative AI and large language models (LLMs) are taking up a lot of the oxygen in recent months.
In movies and some of the more futuristic speculation, the topic is “AGI” - artificial general intelligence. AGI talk predicts and promises us computers that will (someday) have human-like capabilities - or better.
Demos of robots that can seem to carry on a conversation also get a lot of visibility. Examples: “Sophia”, robot dogs, CES demonstrations, the robot conferences in Asia showcasing humanoid “helper” robots
“Self-driving” autonomous vehicles have already been trialed on public roadways in the US and other countries. They draw extra media attention when they have accidents, or when their collision detection features cause car alarms to go off all night in parking lots.
What’s below the waterline of the AI Iceberg?
Recommenders (many, such as movies, videos, songs, books, e-commerce)
Speech To Text (captioning and transcription tools available in meeting tools, word processors, podcasts, etc.)
Text To Speech (reading aloud, audiobooks, generating podcasts)
Image and Face Recognition (car safety features, security purposes, and more)
Computer Vision (intrusion detection, industrial/military, and other purposes)
Optimizers and Navigation (industrial processes, vehicle routing, fleet logistics, and more)
Sentiment Analysis (movie ratings, public opinion, and many more purposes)
Anomaly Detection (credit card fraud, security, equipment health, and other purposes)
Social Media Algorithms (what content and ads you see, and much more)
Predictions (many, such as stock market, fuel prices, sports outcomes)
Language Translation (converting live or recorded speech or music to a new language)
Dynamic Pricing (ride-sharing, concert and event tickets, e-commerce)
OCR (optical character recognition; tools that scan a document or photo and pull out the text from it)
and more.
How many of these areas of AI / ML are above your waterline?
How do these ‘below the waterline’ uses show up in our lives?
For examples of ‘below the waterline’ AI applications, see:
Coming soon: karensmiley.substack.com/p/but-i-dont-use-ai-8-more-sets-of-examples-of-everyday-ai-everywhere (this link will be converted to an embed once it’s published; subscribe to be automatically notified)