Thank you very much for your comments. If you want to contribute to a
scientific paper these must be accompanied with a literature review and
references to creditable books, journal or conference papers.

On Wed, Mar 2, 2022, 03:02 <[email protected]> wrote:

> The ways to measure intelligence or check if something is AGI is:
> 1) Evaluating AI's prediction score, such as Perplexity, or Lossless
> Compression. Thoughts/daydreams is what allows it to come up with a plan,
> if it did not predict good then it would not be using memories and
> therefore would make random decisions. Humans can score it, but much harder
> and longer.
>
> 2) Games and tasks. Once your AI is good at predicting video in
> evaluation1 above, it can be told a goal. If it can get more points, or
> achieve some mission even against human players, or survive, or ex. get far
> enough in a limited amount of time allowed, and in general is a rich game
> with a lot of complexity and length of time needed to win, then this means
> it is predicting good and able adapt goals and learn around that domain,
> allowing it to solve problems and actually see if its dreams come true. If
> an AI only thinks up a solution, often it can be very fantasy, so if you
> try it out in a world and try to obtain even simple tasks, it can be a
> revelation to see and learn your way to truly recognizing the task
> complete. Humans can score it, but much harder and longer.
>
> 3) Test how long human testers play with or talk to the said AIs after the
> above evaluation2 and 3 are founded. The one that has the most time, is the
> more exciting one, maybe. Humans can score the score counted ex. they fell
> asleep so not such high score after all, but much harder and longer.
>
> 4) Although too long to measure, the longer something lives the more
> intelligent it is. This can be a animal, or a country, or a galaxy. Rocks
> do great but can't evolve or adapt. Also measuring patterns is useful too,
> notice everything in your house is square shaped, lined up, grouped
> together, and even in time. It allows you to predict better if you know all
> homes on the road have the same layout. Humans can score it, but much
> harder and longer.
> *Artificial General Intelligence List <https://agi.topicbox.com/latest>*
> / AGI / see discussions <https://agi.topicbox.com/groups/agi> +
> participants <https://agi.topicbox.com/groups/agi/members> +
> delivery options <https://agi.topicbox.com/groups/agi/subscription>
> Permalink
> <https://agi.topicbox.com/groups/agi/T73f43e502fded7a6-Mfda9b26dcd6d1e6e27cd5a30>
>

------------------------------------------
Artificial General Intelligence List: AGI
Permalink: 
https://agi.topicbox.com/groups/agi/T73f43e502fded7a6-M5c25394ce6cce838f2e59eb5
Delivery options: https://agi.topicbox.com/groups/agi/subscription

Reply via email to