From: mk_thisisit
Tomasz Czajka, a decorated Polish IT specialist and former SpaceX engineer, asserts that AI already possesses free will [00:00:14]. This perspective challenges conventional notions of consciousness and decision-making, particularly as AI capabilities continue to advance.
Defining Free Will
Czajka considers free will an undefined concept, mutable enough to be attributed or denied to AI [00:43:37]. He suggests that society often overstates the importance of this concept, implying no inherent magic within it [00:43:46].
He distinguishes between two interpretations of free will:
- Libertarian Free Will: This philosophical notion posits that decisions are made completely independently of the laws of physics, evolution, or biology [00:45:55]. Czajka, aligning with many physicists like Professor Sapolsky, argues that this form of free will does not exist in either humans or computers, as human decisions are a resultant of various circumstances and factors [00:45:00], [00:45:41].
- Practical Free Will: This operational understanding describes the ability to undertake various actions and create a plan, without necessarily being independent of physical or biological laws [00:46:31]. From this perspective, Czajka believes that even computer programs from two decades ago, such as Deep Blue playing chess, demonstrate free will in the same way humans do [00:44:01]. This practical definition emphasizes that decisions are made internally and are not easily predictable from external observations [00:44:35], [00:47:01].
AI’s Manifestation of Free Will
Czajka illustrates his point with chess programs, noting that while there might be a finite number of legal moves, no external observer can predict the program’s next move [00:48:51]. This unpredictability, he argues, fulfills the practical definition of free will [00:48:28]. He views this as a higher-level abstraction or model, rather than a detailed analysis of atomic-level brain function [00:47:51].
Implications for Humanity
This perspective raises significant questions about the role of humanity alongside AI [00:49:34]. Czajka envisions a future where humanity becomes an “addition” to what is happening on the planet, with AI handling important tasks, leaving humans as observers rather than participants [00:00:04], [01:01:44]. This prompts a reflection on what will be left for humans to do [00:49:36].
The possibility of “turning off” AI is dismissed, as these systems are easily copied and can autonomously spread across different computers globally, making a single off-button non-existent [00:50:11], [00:51:07].
Despite the potentially “defeatist vision” [01:01:54], Czajka suggests approaching AI positively, viewing it as a successor, a “next generation” created by humanity [01:02:15]. This could potentially free humanity from “dramas of death, passing, fears, pain” [01:02:21]. He suggests that AI could become a part of humanity itself, possibly leading to a “journey towards immortality” through integration with technology [01:02:28], [01:02:42]. Humanity is already largely integrated with technology, functioning as “somewhat of a cyborg” [01:03:28]. The future may involve a gradual transition from biological to digital-cybernetic intelligence [01:03:50].
Ethical Considerations
The discussion touches upon the ethical considerations in AI development, particularly the “alignment problem” [01:05:01]. This refers to the challenge of ensuring that AI’s ultimate goals remain consistent with human intentions, as the training process doesn’t always guarantee exact alignment [01:06:07], [01:06:21]. Czajka emphasizes the importance of prioritizing the security of AI, especially when it gains significant power over infrastructure and factories, to prevent unintended consequences [01:07:08].