Home » Technology » Phind-70B, it closes the code quality gap with GPT-4 Turbo while running 4x faster, available to trial for free

Phind-70B, it closes the code quality gap with GPT-4 Turbo while running 4x faster, available to trial for free

by Alexandra Hartman Editor-in-Chief

2024-02-25 19:31:48
My opinion? I think the term artificial intelligence for what remains a mathematical model is misleading. On a simple test of playing a game of tic-tac-toe (a game that should normally end in a draw), I beat all the best-performing AI models by systematically playing the same sequence of moves. And they were absolutely incapable of adapting their strategy even when giving them the strategy that I applied.

On a second test based on words supposed to be their strong point, they were unable to verify the application of two simple rules of a word sequence game: the given word must have its first letter identical to the last letter. of the previous word and the given word must not have been given before.

So unless you anticipate all possible cases and train them on them as soon as it leaves their dataset, it’s ruined. And given that it is impossible to give all the possible cases of the real world because of technological, energy, infrastructure and climatic limits, these “*AI*” will remain an approximation and a simulation of the intelligence even of a mouse . Unless they specialize in a specific and restricted area.

Perhaps a new technology will appear capable of truly learning, truly understanding the context and reasoning. But it is certainly not a technology that will be based on binary and silicon because the world is analog and not digital.

1708894601
#Phind70B #closes #code #quality #gap #GPT4 #Turbo #running #faster #trial #free

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Adblock Detected

Please support us by disabling your AdBlocker extension from your browsers for our website.