Sitemap
Pat Inc

A scientific breakthrough in #ConversationalAI. Meaning-based NLU vs. Deep Learning Intent NLU. Sign up for early access: https://pat.ai/

Member-only story

Dogs that know English

--

Animals are much better than today’s AI at key tasks

Dogs seen waiting for the English ‘word’ that allows them to eat. “Oklahoma” doesn’t work, although the ‘o’ sound starts their motion. They know the command for food is “OK”!

Those that follow my work may have seen me highlight how animals demonstrate the goals of robotics. They don’t fall over when catching a ball, despite complex visual backgrounds, uneven footing and complex auditory signals. What is holding back today’s expensive and power hungry AI from replicating an animal’s skills?

It’s the use of the computation paradigm that is unlike a brain.

Patom theory was developed to align with the human brain and therefore it addresses the myriad of known problems going back seventy years that inhibit today’s AI.

In today’s picture, these dogs are dying to eat. They had been waiting with a lack of patience for 15 minutes! I was getting whimpering noises and urgent nudging by both. And yet they have been trained to wait until the sound “OK” is heard before eating. So here, sitting with excitement at the prospect of the meal in front of them, they wait.

Here’s the YouTube video if you want to see their obedience at mealtime (click).

“Chicken!” — no movement. “Aardvark!” — nothing. “Oklahoma!” — an initial move, but they stop as the full sound was not recognized. “OK!” and they move to their bowls and begin eating. They are a little tentative to move when I play…

--

--

Pat Inc
Pat Inc

Published in Pat Inc

A scientific breakthrough in #ConversationalAI. Meaning-based NLU vs. Deep Learning Intent NLU. Sign up for early access: https://pat.ai/

John Ball
John Ball

Written by John Ball

I'm a cognitive scientist working on NLU (Natural Language Understanding) systems based on RRG (Role and Reference Grammar). A mouthful, I know!

Responses (1)