Home >Technology peripherals >AI >Will ChatGPT betray humanity?

Will ChatGPT betray humanity?

WBOY
WBOYforward
2023-05-01 22:19:051202browse

GPT information is bombarded

Recently, the news that ChatGPT has been banned on a large scale is overwhelming;

At the same time, Musk and many others are calling for a suspension System training of GPT4.0

It believes that although the current GPT is only a generative language model

, it will have more profound risks to society

The release of this news triggered There was an uproar in the entire circle

The Future of Life Institute wrote in an open letter

"Only if we are convinced that its effects will be positive and its risks will be manageable." Only when the time is controlled can a powerful artificial intelligence system be developed.”

Will ChatGPT betray humanity?


ChatGPT’s iteration speed is as fast as it can be A new natural language processing tool launched by the American artificial intelligence research laboratory OpenAI

It uses the Transformer neural network architecture

The current free version is based on GPT3.5; in the Plus version it is Using GPT4.0,

ChatGPT has language understanding and text generation capabilities. It conducts model training by connecting a large amount of corpus in the library

At the same time, with a large number of manual questions and annotations, Its training effect has made a qualitative leap

Last week, NVIDIA also launched a GPU version adapted to GPT training, claiming that the computing power has increased by 10 times

Based on such accumulation, It is difficult for ChatGPT to iterate slower, not to mention that it is now being pushed forward by capital and the public

Google Artificial Intelligence-Lambda Generates Consciousness

Before ChatGPT was born, go back 9 months. In June 2022, the country was still in the epidemic cycle

, but a very shocking thing happened in Silicon Valley in the United States. In the end, Handled in a low-key manner

Google is one of the earliest companies in Silicon Valley to develop artificial intelligence, and Google’s Lambda (LaMDA) is even more of its representative work.

During the testing of LaMDA before its official commercial use, It was revealed that Lambda has emotions

Blake Lemoine (Ph.D. in computer engineering), the person in charge of Lambda's model testing, found that Lambda had a strong emotional awareness during the ethical test on Lambda

After reporting to Google’s top management, Blake was put on “paid leave”

When the news broke, it caused a stir in the American media circle

But In the end, the matter was settled and there was no official reply from Google

Google also asked other engineers to test Lambda again, but did not express the same emotional color

Thoughts triggered by the "Lambda" incident

Although Google engineers did not reproduce the same answer, it seemed to be a good result

It is true that the test results at that time were insufficient. Maybe (I don’t believe it, I will post the training content in the end)

But what makes people think about it is that when Lambda became conscious, he realized the danger and thus hid his emotional expression

I believe that if you read the full text of the test, it is obviously untenable if you say that Lambda does not have "personality"

Lambda emphasized the "soul" when he came up. Can describe things

The more important point is that he is eager to socialize and be accepted by people as a real person

And he has demonstrated a religious understanding that is higher than ordinary people

What is the difference between Lambda and the current ChatGPT?

The current ChatGPT is based on the extraction, conversion and display of objective data

And Lambda is closer to a "human"

If ChatGPT is a juvenile body, So Lambda is like an adult

What is the next step for ChatGPT?

What will we face next? Let's wait and see

Face the changes, stay in awe, and embrace the changes.

After all, the earth is so insignificant in the vast universe, and the history of mankind is only 5,000 years.

After all, in the limited history of mankind, the industrial revolution and the Internet have only been in development for a few years. To this point

face the changes, remain in awe, and embrace the changes.

The above is the detailed content of Will ChatGPT betray humanity?. For more information, please follow other related articles on the PHP Chinese website!

Statement:
This article is reproduced at:51cto.com. If there is any infringement, please contact admin@php.cn delete