Artificial Intelligence Acronyms Decoded

Artificial intelligence has started to transform how we process and consume information, and it’s also given rise to a new class of buzzy acronyms that now flood our Twitter feeds. The three big ones include: AI for artificial intelligence, ML for machine learning, and NLP for natural language processing. This can seem like a game of alphabet soup or like listening to a robot say the alphabet, but I think I can help.

At Astro, we’re using AI, ML, and NLP to make a new class of workplace tools, starting with intelligent email apps. As the first marketer at Astro, I’ve had to get up to speed quickly on the meanings behind these acronyms to be able to explain how AI can improve email and communications. So, I thought I’d shared what I’ve learned and break them down in plain English.

AI: Artificial Intelligence

  1.  A branch of computer science dealing with the simulation of intelligent behavior in computers (Merriam-Webster)
  2. The capability of a machine to imitate intelligent human behavior (Merriam-Webster)
  3. Any device that perceives its environment and takes actions that maximize its chance of success at some goal (Wikipedia)

When to say “AI”

AI is a branch of computer science (CS) that involves machines taking actions like a human. If you aren’t sure if something is powered by AI, ML, or NLP, play it safe and say AI, since NLP And MLP are fields of AI. In the analogy of all squares are rectangles, but not all rectangles are squares, ML is the square and AI is the rectangle. Sorry to explain a CS topic using math analogies, it won’t happen again.

Artificial Intelligence is a branch of Computer Science, and Machine Learning is a subset of AI

The relationship between machine learning, artificial intelligence, and computer science. If it wasn’t obvious, this is a simplified, non-technical diagram.

Is AI a good thing?

We think so. Artificial intelligence can improve your efficiency, increase convenience, and add delight by highlighting important content, automating repetitive tasks, personalizing content and recommendations, etc. AI doesn’t mean machines are taking over you life or that the singularity has arrived. AI is currently powering apps, websites, and devices that you currently use, and these AI-powered tools save you time and effort.

AI often works by collecting and analyzing large data sets to learn and make predictions. While you should always be careful when sharing information on the internet, AI doesn’t mean your privacy is more at risk. Data that AI learns from is usually anonymized.

ML: Machine Learning

  1. A branch of artificial intelligence in which a computer generates rules underlying or based on raw data that has been fed into it (Dictionary.com)

When to say “ML”

When we talk about software being powered by AI, we are usually talking about machine learning. Put simply, machine learning is a technique used to allow machines to learn from experience. Without machine learning, we’d have to teach or program a computer to do every action we want it to do. With machine learning, computers are able to learn based on previous data and actions.

If the software you’re using gets smarter based on how you use it, there’s probably machine learning happening behind the scenes. Movie and tv recommendations on Netflix or Hulu and shopping suggestions when ordering on Amazon are great examples. If you always order  frozen pizza, you’d probably not only buy frozen pizza again, but also (based on what others frozen-pizza-lovers buy) you may also be interested in buying a box of macaroni and cheese. Similarly, ML can be applied to any type of software, like email. At Astro, we’re creating email apps that will get smarter about classifying and prioritizing your emails the more you use them.

Computers can learn now?!

How machine learning works is pretty similar to how humans learn. We learn a set of facts, and we use these to make predictions, which then affect our behavior and decisions. For example, if it’s windy outside and the sky is gray and cloudy, we use previous experience to predict that it’s going to rain. Based on that, we bring an umbrella to work. We don’t need to memorize every set of possibilities and the appropriate response.

With machine learning, computers can do more than predict the weather, and tell you what food you should buy. Machine learning enabled a computer to beat a master player of the game Go. The computer observed tons of games of Go, identified patterns, and then used it to determine its next move. It continued to learn and get better just like a human playing a board game (well even better than a human, no matter how many times I play Go, I don’t think i’ll ever be able to beat the reigning European champion). A more life-changing application: machine learning is beginning to change healthcare, it can be used to recommend personal care based on a patient’s unique symptoms. Machine learning still requires code written by a human and inputs of data sets to learn, so like I said, the singularity isn’t here quite yet.

NLP: Natural Language Processing

  1. Natural language processing (NLP) is the ability of a computer program to understand human speech as it is spoken. NLP is a component of artificial intelligence (AI). (TechTarget)

When to say “NLP”

NLP is the reason that a computer can respond to you when you type or say something in sentence form. Chatbots are powered by NLP, as is the ability to search in question form rather than just keywords. TLDRs (too-long-didn’t-reads, yet another fun internet acronym) that are generated without a human writing them are also made possible through NLP. Sentiment analysis on social posts is powered by NLP, and so is speech recognition via your Amazon Echo.

Natural language processing is at the intersection of Artificial Intelligence and linguistics

NLP, at the intersection of artificial intelligence and linguistics. If it wasn’t obvious, this is a simplified, non-technical diagram.

Linguistics + AI = NLP

NLP combines linguistics and AI. NLP must interpret language and understand things like humor, sentence phrasing, slang, etc, and since we can’t possibly teach a computer everything that there is to know about language, it must get smarter over time (that’s where the AI, specifically ML comes in).

How can these acronyms help me?

The technology these acronyms represent will continue to change how we experience, use, and interact with software. NLP can help you interact with computers without having to modify your language so that you can type as you would to a friend and get a response back like you would from a very smart, computer friend. AI, and specifically ML, means computers will advance in their abilities to help you faster than ever before.

At Astro, we’re excited about how this affects everything we do, from self-driving cars to movie recommendations. Even more exciting to our team is how AI, ML and NLP can tackle information overload, fix email and help you work more efficiently. Astro is currently available in Public Beta on Mac, iPhone, and iPad.

Additional Reading

Here are some articles I read to get up to speed:


Also published on Medium.

Thanks for downloading Astro for Mac

 

If your macOS download didn’t start, please click here.
Astro requires macOS versions 10.11 or later.
By downloading our app, you agree to Astro’s Terms of Use.