Yeah, I know… symbolic AI isn’t cool… yet!

Photo by Jason Strull on Unsplash

Symbolic AI (or “hand-coded” AI) has gotten a bad reputation in the age of neural networks and machine learning, where it’s assumed that all knowledge can be extracted (bottom-up) from big labeled datasets. Why bother to hand-code symbols, rules and relationships when these already exist in the data?

However, artificial general intelligence (AGI) is different. If the goal of AGI is to build a computer that can think on its own and solve complex problems, symbolic AI is the way to go. Sure, neural networks still have a role to play in…

All it takes is a common software library across the human species

Photo by Michael Dziedzic on Unsplash

We use language to communicate everyday rules and common sense bits of wisdom to others. “If you keep your key in the ignition with the radio blaring but the engine off for 2 hours, your car battery will probably be dead.” Or “Don’t touch the stovetop if the red light is still on.”

In addition to this simple IF-THEN logic, it’s also possible to communicate more complex LOOPs and variable assignments using words. …

Feelings, emotions and even biases are just algorithms

Photo by Stephen Leonardi on Unsplash

Recently, I heard a loud crash in my backyard. As I rushed outside, I found a strange oblong and metallic craft, damaged and half-buried in a muddy crater, a gaping hole ripped in its side. When I peered cautiously inside, I discovered a shaken but unharmed alien staring back at me, its quivering gelatinous brain illuminated by billions of blinking lights, resembling a miniature data center.

“Are you OK?” I yelled.

The alien said nothing. Instead, a series of words flickered on a monitor attached to its chest:


How to progress toward Artificial General Intelligence (AGI)

Photo by Rob Schreckhise on Unsplash
  1. Basic pattern recognition. Extract features from raw sensory inputs; for example, the ears convert raw sounds to frequency maps; the eyes convert visual inputs to feature maps (used for facial recognition, etc.) Associate new perceptions with existing memories (via dimension reduction). Use algorithms to actively filter out irrelevant stimuli (habituation) and ignore stimuli after repeated exposure (adaptation).
  2. Mental representation. Mentally represent and simulate the state of the environment. Track physical objects across time, space and sensory inputs, and understand that physical objects are solid, can support other objects, and can otherwise affect…

…and how to fix it

Source: Getty images

What’s wrong with AI?

If you want to know what’s wrong with current theories of artificial intelligence (AI) — and avoid the next AI winter —just ask a neuroscientist.

In “How We Learn,” neuroscientist Stanislas Dehaene outlines how babies apply a rich assortment of prior models to understand people, objects, time, space and causality. Babies don’t use anything remotely like current AI techniques, except perhaps in the pattern recognition capacity of the senses.

Today’s AI neural networks are shallow and imperfect. They attempt to learn everything at the same level, gleaning superficial statistical regularities in data and shallow features in…

What could be worth more than that?


Scientists have discovered that an asteroid 230 million miles from Earth called “16 Psyche” appears to be 90% metallic, composed of nickel and iron. NASA has plans to visit the 120-mile-wide asteroid in 2026.

Psyche is large for an asteroid, but smaller than a planet. So small, in fact, that its surface gravity is only 0.015 times that of Earth. If you were to race across the surface of Psyche at 409 miles per hour, you’d achieve escape velocity and lift off into space.

The value of the metal in the asteroid is estimated at $10,000 quadrillion. That’s a lot…

Artificial General Intelligence

And the implications for human intelligence

Source: author

At a minimum, Artificial General Intelligence (AGI) should explain how common human cognitive tasks are accomplished:

  • Planning and Acting
  • Recognizing Objects and Situations
  • Predicting the Future
  • Attributing Sources and Errors
  • Understanding Cause and Effect
  • Reasoning and Advocating
  • Understanding the Motives of others

A graph is a handy way to enable these cognitive abilities and represent the state of the environment (observable world). For example, we might recognize a dog as a specific instance of a prototypical dog. …

Hint: It’s all about feeling, emotion, and common sense.

Duke Lemur Center. Photo by the author.

Elon Musk famously warned that “with artificial intelligence (AI), we are summoning the demon.” He feared that AI might suddenly spin out of control and put society at risk. Ironically, AI must succeed if Musk is to deliver on his promise of self-driving autonomous cars that can navigate through bad weather, poorly marked roads, and novel emergency situations.

In the past two decades, AI has made great strides. Computers can now handily beat human chess masters, as well as world champions at the ancient game of Go. Two AI methods in particular — reinforcement learning (RL) and agent-based modeling (ABM)…

Rob Vermiller

A computer scientist with a passion for AI and Cognitive Science.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store