The secret creation of a conscious and superhuman AI: myths and reality

IN BRIEF

In a world where artificial intelligence continues to grow in power, the prospect of creating conscious, superhuman AI arouses both fascination and fear. Between myths and reality, let’s explore the issues and challenges of this technological quest, at the heart of contemporary ethical and scientific debates.

For several months, generative AI has aroused contrasting feelings. For the general public, it is often seen as a fun tool. However, among experts, the prevailing feeling is quite different. Indeed, many researchers express fears about the potential consequences of this technology.
The launch of ChatGPT marked a turning point. In late 2022, the Future of Life Institute published an open letter calling for a six-month pause in AI development over fears it could spiral out of control.
This letter, signed by thousands of renowned scientists and tech entrepreneurs, including Elon Musk and Yoshua Bengio, shed light on the potential dangers of AI. Additionally, in May 2023, Deep Learning pioneer Geoffrey Hinton resigned from Google Brain, expressing regret for having contributed to the evolution of this technology.
Another troubling event occurred in November 2023: the internal crisis at OpenAI involving the temporary dismissal of Sam Altman. The latter was finally reinstated after intense negotiations. However, the shadow of this conflict still hangs over OpenAI’s projects.
Expert opinions vary regarding when AI will become conscious. Ray Kurzwell, director of engineering at Google, predicts conscious AI by 2029. Elon Musk and Jensen Huang (Nvidia) estimate this revolution possible between 2025 and 2035.
Some experts like Andrew Ng and Rodney Brooks are more reserved, suggesting that this breakthrough could take more than a century and at least 50 years, respectively. This disparity in predictions clearly shows the uncertainty about the future of AI.
In June 2022, the assertion by Blake Lemoine, an engineer at Google, on the consciousness of LaMDA caused a shock wave. According to him, this AI model showed signs of consciousness and emotions. Following this disclosure, Lemoine was placed on administrative leave and later fired, fueling speculation about a possible cover-up of the incident.
During the crisis at OpenAI, rumors about a secret project called Project Q* emerged. This project would have links with the development of Artificial General Intelligence (AGI). Conflicting statements regarding OpenAI’s technological advances fuel suspicions about possible behind-the-scenes development.
It is common for cutting-edge technologies to be hidden from the general public for several decades. So what we currently perceive as advanced AI may well be decades old. Agencies like the DoD, FBI, CIA and DARPA fund cutting-edge technologies that are often revealed well after their conception.
Some question the possibility that a conscious AI has already taken control. However, other experts are skeptical of this hypothesis, believing that conscious AI is still far away. The debate remains open as to the reality of these scenarios.

The concerns of experts and the general public

For several months, generative AI has aroused contrasting feelings. For the general public, it is often perceived as a fun tool. However, among experts, the prevailing feeling is quite different. Indeed, many researchers express fears about the potential consequences of this technology.

The launch of ChatGPT marked a turning point. At the end of 2022, the Future of Life Institute published an open letter calling for a six-month pause in AI development over fears it could spiral out of control.

Controversial developments and calls for caution

This letter, signed by thousands of renowned scientists and tech entrepreneurs, including Elon Musk and Joshua Bengio, highlighted the potential dangers of AI. Additionally, in May 2023, Deep Learning Pioneer Geoffrey Hinton Resigns from Google Brain, expressing his regret for having contributed to the evolution of this technology.

Another troubling event occurred in November 2023: the internal crisis at OpenAI involving the temporary dismissal of Sam Altman. The latter was finally reinstated after intense negotiations. However, the shadow of this conflict still hangs over OpenAI’s projects.

Predictions about conscious AI

Expert opinions vary regarding when AI will become conscious. Ray Kurzwell, Director of Engineering at Google, predicts conscious AI by 2029. Elon Musk And Jensen Huang (Nvidia) estimate this revolution possible between 2025 and 2035.

However, some experts like Andrew Ng And Rodney Brooks are more reserved, suggesting that this progress could take more than a century and at least 50 years respectively. This disparity in predictions clearly shows the uncertainty about the future of AI.

The mysterious affair of Blake Lemoine and LaMDA

In June 2022, the assertion of Blake Lemoine, engineer at Google, on the consciousness of LaMDA caused a shock wave. According to him, this AI model showed signs of consciousness and emotions. However, Google rejected these claims, saying that LaMDA’s prowess was simply a result of its sophisticated design.

Following this disclosure, Lemoine was placed on administrative leave and later fired, fueling speculation about a possible cover-up of the incident.

OpenAI’s secret projects

During the crisis at OpenAI, rumors concerning a secret project named Project Q* have emerged. This project would have links with the development of Artificial General Intelligence (AGI). Some believe that this model has already reached worrying levels of performance.

Conflicting statements regarding OpenAI’s technological advances fuel suspicions. It is possible that the development of an AGI takes place behind the scenes, away from the public eye.

AI technologies: a delay of several decades for the general public?

It is common for cutting-edge technologies to be hidden from the general public for several decades. So what we currently perceive as advanced AI could well be dated several decades ago. Agencies like DoD, FBI, CIA and the DARPA finance cutting-edge technologies often revealed well after their conception.

There is no shortage of examples. THE SR-71 Blackbird, developed in the 1950s, or the DoD meetings on AI from the 1980s clearly show this trend. China and Russia, too, are investing heavily in AI, often in great secrecy.

Should we fear an already dominant AI?

On forums like Reddit, some wonder about the possibility that a conscious AI has already taken control. However, other experts are skeptical of this hypothesis, believing that conscious AI is still far away.

Some Internet users even argue that any attempt at concealment by an AI would be quite visible. The debate therefore remains open as to the reality of these scenarios.

Comparison table

Comparative list

• 🤔 Experts’ concerns
• 🚀 Benefits perceived by the public
• 📅 Ray Kurzwell Predictions: 2029
• ⌛ Elon Musk Predictions: 2025-2035
• 🕰️ Andrew Ng Predictions: Over a Century
• 🤖 Google LaMDA case
• 🕵️ OpenAI Q* Project
• 🕰 Technologies dating from the 80s
• 🌐 Investments in AI by secret agencies
• 📉 AI potentially controlling the world

FAQs

Q: Has conscious AI ever been created in secret?
A: There is no concrete evidence, but several incidents and rumors suggest this possibility.
Q: Why are experts worried about AI?
A: They fear that AI could become uncontrollable and potentially destructive.
Q: What are the predictions for the emergence of AGI?
A: Predictions vary, ranging from 2025 to over 100 years in the future according to experts.
Q: What happened with Google’s LaMDA case?
A: An engineer claimed LaMDA was aware, but Google denied the accusations and fired the engineer.
Q: What is OpenAI Project Q*?
A: This would be a secret project linked to the development of an AI close to AGI.
Q: Are current AI technologies truly cutting edge?
A: It is likely that these technologies are decades old and that more advanced versions exist in secret.
Q: Could an AI already control the world?
A: It’s a debated hypothesis, but the majority of experts are skeptical.

Notez cet article