The Moonshot AI Problem - Deepstash

The Moonshot AI Problem

Machine learning of common sense is now considered a moonshot problem, requiring intensive collaborative efforts from different institutions and organizations. 

Understanding natural language and answering simple questions are the first steps towards building common sense in machines.

To learn common sense, machines would have to master sociology, psychology, planning and execution, emotions, threat detection, background knowledge, and many abstract concepts.

38 STASHED

11 LIKES

MORE IDEAS FROM THEARTICLE

  • Computers have mastered many games involving strategy and logic(like GO and Chess), but common sense learning is a dream.
  • AI(artificial intelligence) may be great at algorithmic computations but falls flat when tackling real-world complexities.
  • Modern AI can handle specific problems, but common sense is hard to master as it is vague and does not come with a predefined set of rules.

36 STASHED

10 LIKES

The Superpower Of Common Sense

We take common sense for granted, but this uniquely human trait is difficult to emulate artificially, as it relies on a variety of powerful abilities of our mind.

Common sense is a broad trait, which includes spatial intelligence, reasoning, managing expectations, knowledge of abstract concepts and social abilities.

54 STASHED

23 LIKES

G. K. Chesterton - English Philosopher

"Common sense is a wild thing, savage, and beyond rules."

39 STASHED

14 LIKES

Deepstash helps you become inspired, wiser and productive, through bite-sized ideas from the best articles, books and videos out there.

GET THE APP:

RELATED IDEAS

Although AI researchers can train systems to win at Space Invaders, it couldn’t play games like Montezuma Revenge where rewards could only be collected after completing a series of actions (for example, climb down ladder, get down rope, get down another ladder, jump over skull and climb up a third ladder).

For these types of games, the algorithms can’t learn because they require an understanding of the concept of ladders, ropes and keys. Something us humans have built in to our cognitive model of the world & that can’t be learnt by the reinforcement learning approach of DeepMind.

16 STASHED

1 LIKE

Narrow AI & General AI

Very broadly, AI can be divided into two: narrow AI and general AI.

1. Narrow AI systems handle singular or limited tasks. Also referred to as weak AI sometimes, such systems have applications in email spam filtering, recommendation systems, and autonomous vehicles.

2. On the other hand, general AI or strong AI, refers to a machine’s capability to think and function as a human. It denotes the ability to distinctly recognise other intelligent entities’ needs, emotions, and thoughts.

24 STASHED

2 LIKES

<p dir="ltr">Content marketing...

Content marketing is the most effective marketing strategy . And there is always a competition to write better content, be more creative, and be more engaging. But in this new digital era, can be used in content creation. As you know, evolved into more than simply a futuristic technology. There’s a high possibility you’re already experimenting with , no matter what field you work in. Let’s face it, is everywhere, whether it’s deploying chatbots to collect data about your users’ most urgent concerns or assessing content results with platforms.

4 STASHED

1 LIKE