Reverse Alphabet of the AI Bullshit
Zero knowledge: That’s what it takes if you want to become an Artificial Intelligence guru. Want to become a member of the AI armchair philosophy club? Take a tour below…
“You are already using it everyday”: Your best shot for an introduction. Mention Siri, your spam filter, and explain it’s already everywhere. Everywhere.
Work: Allegedly annihilated soon by the AI rise. Although nobody has ever seen any job destruction due to AI, you have to look deeply worried about this upcoming tsunami (“If only you knew as much as I know…”).
Visual recognition: R&D field entirely dedicated to solving a crucial problem: identifying a dog in a picture.
Unsupervised (learning): Useful spell when arguing with a data scientist invoking the hard work of datasets labelling. Helps avoiding stupid questions like “What is it you want exactly?”. Your move: “Just apply unsupervised learning, and let me know what comes out”. Leave the room.
Technology: Too important to be left in the hands of IT Departments. Should instead be handed over to the Business. Side note: it’s always ok to add up to the corporate clivage between IT (= conservative retards) and the Business (= boundless visionaries).
Startup: Any internal project. Gets funded more easily when AI-powered.
Robots: Use pictures of humanoid robots all over your AI powerpoint presentation. It helps. Plus it’s classy and original.
Robotics Process Automation (RPA): AKA “Expert Systems” in the 80s and 90s, in reference to blackbox systems packed with thousands of business rules, that nobody since was ever able to maintain. In its current version, defines a blackbox system packed with thousands of business rules, that Jim said he was ok to maintain over the next decades.
Rise (of AI): Unlike the rest of the world, you saw it coming, and rushed to finish that nuclear shelter you started building when Terminator was released.
Regression: Keep hard forgetting we all learnt linear regression in 11th grade (we all did). Side note: whatever the problem, it’s quite safe to suggest a regression model as a first approach.
Operational Research: The Art of Mathematical Optimisation. Used for decades, especially in Supply Chain (= not cool), suddenly considered a field of Artificial Intelligence (= cool).
Neural Networks: Always talk about them like you are a brain surgeon.
Machine Learning: It’s learning faster than us, it never sleeps, it never complains, and by the time we realise, it will have surpassed us. Like Japan in the 80s.
Intelligence: Inessential when talking about its artificial counterpart.
Import: Your rabbit hole to Data Science Land. Datasets not supplied, bring your own.
from sklearn import svm
Gradient Descent: That thing you struggled learning, before realising you would never have to use it ever. Like learning to code in Assembler in 2018. An ace up your sleeves in tech troll contests though (like Assembler).
General AI: You know, AlphaGo, and… and (…)
Ecosystem: A club, sort of. You have a platinum membership there. It’s packed with AI startup founders. You know them all by their first name, and you know the bouncer personally.
Deep Learning: Digs the answers far below your subconscious. Like really deep, Bro.
Data Scientist: A developer with a beard, working on a Mac, who turned to Python and once typed the above import command. Always have one around.
Data: The boring side of AI. You need some.
Chatbots: That thing you suggest for replacing every callcenters in the world, although you wouldn’t trust one yourself for ordering a pizza.
Blockchain: Often mentioned by AI visionaries, even though based on a distant set of technologies. The term follows the same pattern as “social networks” 10 years ago, as randomly associated with any business application whatsoever: “blockchain for maintenance”, “blockchain for food”, “blockchain for chains of blocks”, …
Big Data: Summon its power if you deal with 1 million rows or more. If anyone objects that it could fit in a regular database, he or she is from the IT Department, and can therefore be ignored safely (see “Technology”).
Autonomous Cars: <irony-off> The next big thing. A large scale application of Machine Learning, among other techs. Never associated with AI in its press coverage. </irony-off>
Artificial Artificial Intelligence: Humans, secretly mimicking machines, supposedly mimicking humans.
Artificial Intelligence: Just a (not so) new technology toolbox, with lots of cool business applications. Like relational databases in the 80s, or object-oriented programming in the 90s. Better left in the hands of technologists. Like relational databases and object-oriented programming.
Yup, irony was off in that last one too.