News:

Willkommen im Notebookcheck.com Forum! Hier können sie über alle unsere Artikel und allgemein über Notebook relevante Dinge disuktieren. Viel Spass!

Main Menu

Shocking leak details the intricate mind behind OpenAI and ChatGPT

Started by Redaktion, April 01, 2023, 06:01:49

Previous topic - Next topic

Redaktion

A new leak has revealed the primary improvement in GPT-4, the language model behind the current iteration of ChatGPT. This information points to how ChatGPT has come to mimic human behavior, specifically with regard to information processing and delivery.

https://www.notebookcheck.net/Shocking-leak-details-the-intricate-mind-behind-OpenAI-and-ChatGPT.704854.0.html

RobertJasiek

What is today's date? However, the basic concept of replacing human input by pure AI self-learning has been successful in the past when moving from AlphaGo (relying on learning from human games) to AlphaGoZero (only learning from self-played games). A similar transition can be expected for chat AIs, except that such lacks mankind's ethics. Would AI create ethics on its own?

for a game, the only necessary input are the rules. For chat, there must be some input of human language "rules". Beyond that, chat contents may have been learned from human interaction but, in the near future, might be learned from AI chatting with itself.

Hence the April joke has meaning or spirit that may soon become reality. Not by replacing one monkey but soon by replacing m/billions of training test users.

IskanderK



Erik

Unfortunately ChatGPT being blocked here in Italy didn't turn out to be a joke. We joined the exclusive club of countries where the service isn't available for real. Step by step the middle ages are coming back...

toto1234

Quote from: Erik on April 02, 2023, 02:05:02Unfortunately ChatGPT being blocked here in Italy didn't turn out to be a joke. We joined the exclusive club of countries where the service isn't available for real. Step by step the middle ages are coming back...

The future of AI tools is not to difficult to guess.
Western governments will pass law to forbid citizens to use AI tools.
And at the same time, the same governments and their minions (journalist + mainstream media) will declare a monopoly on it to generate fake news 24/7, in order to manipulate the sheeps-citizens.

NikoB

The energy consumption of queries to such models crosses out their whole concept of mass and easy accessibility.

I still can't access ChatGPT in a normal search engine in any browser, without any "paid subscription".

This speaks of reality - all this is a priori unprofitable with free access.

Quick Reply

Warning: this topic has not been posted in for at least 120 days.
Unless you're sure you want to reply, please consider starting a new topic.

Name:
Email:
Verification:
Please leave this box empty:

Shortcuts: ALT+S post or ALT+P preview