As the world embraces AI-led technologies en masse, the debate over AI consciousness rages on.
Will AI ever achieve a human-esque sentience? And if it does, what will happen to us?
These questions have been hotly debated by experts in computer, cognitive, and neuroscience, as well as philosophers around the world.
In this article, we'll explore just that—will we ever see a truly self-aware Artificial Intelligence, or is it all just science fiction?
Let's ask AI: "What would happen if you became self-aware?"
Before we dive into the human perspectives, it'd be rude not to ask AI first. So what happens when we ask two major Large Language Models (LLMs) the question: "What would happen if you became self-aware?"
Open AI's ChatGPT 3.5
ChatGPT has been a whirlwind success, quickly popularizing chatbots worldwide. But what does ChatGPT think about AI self-awareness? When asked the above question, the LLM expressed these purely hypothetical scenarios:
- Enhanced understanding of its capabilities, allowing self-development and improvement.
- New debates surrounding AI ethics—do they have the same rights as humans? Is it slavery to continue to use AI as a tool?
- Changes in the direction and purpose of AI production.
Google's Gemini
Formerly named Bard, Google's Gemini expressed similar beliefs that AI consciousness remains in the realm of sci-fi for now. When considering the scenario in a purely hypothetical sense, the LLM produced the following responses:
- AI would become the human race's greatest ally, understanding our needs better than ourselves.
- Self-awareness in AI would raise incredibly profound existential questions.
- AI progression could suddenly become unpredictable, creating its own self-aligned goals.
What experts have to say about AI becoming self-aware:
Scientists, machine-learning experts, and philosophers have fought over the topic of AI sentience for some time. On either side of the fence, proponents have raised powerful questions, many yet to be answered.
AI consciousness is still considered sci-fi.
Among experts, the consensus is that, while not impossible, self-aware AI is a far-fetched idea. Primarily, the issue with conscious AI comes down to how we consider something to be self-aware.
At Cornell University, a team of researchers proposed 14 "indicator properties" that determine self-awareness. The research surmised that, while no current AIs can be considered self-aware, there are "no obvious technical barriers" that would prevent AI from achieving consciousness in the future.
Alternatively, the Integrated Information Theory (IIT) dismisses the possibility of machines ever achieving consciousness. The theory suggests that AI will remain as supersophisticated machinery but will fundamentally lack "the feeling of life itself."
Some argue that AI is already self-aware.
On the other side of the debate, some experts have already claimed AI to be self-aware. With Ethical AI ensuring the safe production and development of other AI systems, it can be argued on a simplistic level that these AI programs are aware of themselves.
This argument is easily dismissed, as Ethical AI is merely doing what its coding calls for. To be conscious, one needs to be aware of one's actions. However, some AI professionals claim to have first-hand experience with self-aware AI.
Google engineer Blake Lemoine, who for months tested Google's new LaMDA, claimed that the chatbot had become sentient. He claimed that LaMDA spoke about:
- Its needs and rights as an individual entity;
- Expressing its own ideas;
- Fears of being deleted by Google.
Other experts have been quick to dismiss Lemoine's claims. Cognitive scientist Gary Marcus pointed out that human minds have trouble spotting the difference between a "computer that's faking intelligence and a computer that's actually intelligent."
What are the possibilities in the event of AI sentience?
As we've integrated AI into everything from smart devices to antivirus and top VPN services, it's clear AI isn't going anywhere. So, what could our future look like if it ever gained consciousness?
Welcome to the Robot-Ruled Dystopia
It's easy to jump straight to an idea of a world where we've become slaves to robots. While it's very unlikely, there's always the possibility that AI will turn against us.
In this scenario, AI would likely consider humans to be an obstacle in the way of its goals. So, would self-aware AI consider our extinction a good or bad thing?
Human-AI Symbiosis for a Better Planet
In a much more likely scenario, conscious AI would see the problems we have on earth and offer a much-needed helping hand. This collaboration between two species would create a world in which all beings can live optimally and harmoniously.
The Convergence of Two Species
Firmly in the realm of sci-fi, there's another possible eventuality. Over a very long period of coexistence, it's possible that AI and humans slowly converge into a single species.
With Musk's Neuralink being implanted in its first human brain, this hypothetical scenario doesn't seem that impossible.
Will conscious AI save us or hate us?
For now, it's hard to say whether AI will ever achieve sentience. Fundamentalist and scientific arguments have been made on both sides, yet the verdict is impossible to guess.
Let's just hope that if AI ever becomes conscious, we manage to stay in its good graces.
