- cross-posted to:
- [email protected]
You must log in or register to comment.
Though the author addresses the flaws of the proposal of algocracy, he conviniently forgets the fault which is the downfall of all the will thought out plans. Human emotions, greed and ambition.
Just as democracy in authors words is nothing more than a beauty pageant. The destruction of democracy are those reasons, and will also be the destruction of algocracy.
Author also side steps philosophical questions regarding AI such as the trolley problem.
Until these problems are solved, algocracy might be even more destructive than what democracy is currently.
You are correct, I did not address these, merely for the sake of brevity. I am acutely aware of human greed, it’s a feature not a bug 😅. In fact, algocracy is the weapon of choice here because of exactly this.
Algocracy aims to reduce the impact of human emotions, greed, and ambition. Calling democracy just a ‘beauty pageant’ underscores the search for a more stable system, one less affected by our natural flaws.
The trolley problem isn’t exclusive to AI. They challenge human judgment just as much, if not more. What algocracy offers is a consistent platform to tackle these dilemmas based on collective societal input. The proposal focuses on AI systems overseen by human direction, reinforced by open-source principles and transparency. Furthermore, its accessibility could be enhanced by allowing individuals to inquire about the system and propose enhancements through LLM interactions.
Arguing that algocracy could be more destructive than current democratic practices is a hasty judgment. The vision is to refine and mold it, leveraging open-source methods and transparency, into the best version of governance we can muster. An iterative, ever-improving model, always under the watchful eyes of society.
Thanks for the discussion.
Trolley problem is not exclusive to AI but consequences of a dissociated program taking a decision is less acceptable as compared to a committee of humans. For all AI it would merely be an equation to solve with no inkling of what that consequence mean. As a human condemning another human to death is much more than a mere dilemma.
The solution as you proposed is human intervention. The question arises however that who would choose those humans. A learning AI might learn to choose only those humans who would not oppose is decision. Any product of human intellect will have the same biases and faults as the humans those who have made them. There are no true creations but rather inspiration from who we are and what we see.
Whereas human overseers chosen by us would have the same problems as it is with democracy.
I am not against using ai but replacing humans is not and should not be the aim.
I appreciate the insights you’ve shared. Here’s a clarification on the core points:
-
Algocracy vs. Human Emotion: The objective isn’t to replace the entirety of human decision-making but to use algocracy in areas where certain human flaws can be systematically minimized. Algorithms, when designed properly, can reduce the impact of biases and inconsistencies.
-
Selection of Human Overseers: The selection would remain transparent and lean towards open-source principles. Individuals would be chosen by peers and those deemed competent within the community. We migh utilize a meritocratic governance model combined with token-based voting for stake, enhanced by a reputation system to ensure decisions are made by competent, committed individuals, all underpinned by open-source principles for transparency and collective oversight (this is oversimplifying it, there is a huge amount of great ways; I hope to have time later to write about them).
-
Purpose of Algocracy: The goal is not an AI or algorithmic “takeover.” It’s about integrating algorithmic governance in specific areas to achieve consistency and fairness, always with human oversight at its core.
The emphasis is on refining our governance systems, not replacing the human touch. Challenges exist, but with collaboration and transparency, we can navigate them effectively.
-
The problem with algocracy is someone has to define the laws in which the algorithm has to base it’s decisions. And if someone is writing those laws, how can we know they’re impervious to corruption.
Quis custodiet ipsos custodes?
No system, made or operated by humans, is impervious to flaws; by design, we have our shortcomings. However, the challenge lies in creating the least flawed system, safeguarding it from our biases. This involves crafting a system that automates away the potential for corruption and greed, yet maintains room for human intervention to prevent a rigid “computer says no” outcome.
By using open-source development methods and absolute transparency; algorithms could be developed openly, visible to all. Furthermore, to make it even more user-friendly — not just for algo developers — we could employ LLMs. This would allow anyone to converse with a bot, querying about the rationale behind laws, their fairness, and so on. If perceived as unfair, one can present their argument to the bot, which then channels the feedback into the system for potential amendments.
The entire procedure would then be under continuous peer review.
I didn’t realise you wrote the article. I just wanted to quickly express how articulate and well written it is. And I completely agree with your points on the free market, government intervention and how unsustainable our current system is.
Though my view with algocracy is more to do with the types of individuals that are running the show. The wealthy, powerful and influential. The types that will stop at nothing to get what they want.
I feel that these people will always find opportunities to bend and manipulate systems in their favour. If it threatened their status, they’d destroy it. If they were somehow pushed into a corner they’d find ways to manipulate it.
It may be pessimistic but I really believe that in a dog eat dog world, there is no such thing as a system that is truly fair. You say an algorithmic system solves that. But I believe we’ll never reach it… Not in a way where these people would lose their influence. They would never allow it.
Thank you! I’ve been able to improve my articulation and punctuation significantly with the assistance of my GPT-4 subscription. Don’t get me wrong – the core ideas and writing are entirely my own. GPT-4 just helps me polish them a bit more.
Never lose hope! They’re counting on our despair. I agree, it’s a fucking brutally insane, dog-eat-dog world. It often feels like we’ll be trampled upon before we can pry anything away from those greedy fuckers hands.
The way you feel is understandable. We are, after all, in what can be called a ‘post-failure’ era (this is by one of the guys who started the occupy wall street movement).
I’ve moved beyond the tinfoil hat metaphor. I now jokingly say that I wear a tinfoil suit. 😄 The powers that be have been establishing their stronghold since before the Industrial Revolution. It’s no wonder we feel overwhelmed, seeing how they’ve shaped our societal functions, thought processes, and actions. Our laws, daily routines, and even our beliefs have been influenced by their interference.
I’ve spent considerable time pondering potential solutions. Here are some broad concepts I believe could lead us out of our current predicament. I’ve written more about the concept here, but here are the broad strokes:
- Establish a Global Network of City-States: These would be independent entities scattered around the world.
- Cryptography & Algocracy: Each city-state would be governed by unbreakable cryptographic systems and algocracy, ensuring that those with ulterior motives can’t manipulate the system. Built-in safeguards would prevent undue influence.
- Resource Sharing: These city-states would cooperate and share resources, within reasonable limits, to benefit the collective. (I intend to write articles about how they could operate their own independent global shipping system, and have one semi-related concept for a quasi-workable communication system).
- Open-Source Ethos: Embracing an “information-wants-to-be-free” mindset would be paramount. This approach could potentially lead these city-states closer to a technological singularity, especially when compared to global states that might be on the path to deglobalization and de-industrialization.
- Independent Economy: These city-states would operate their own economy, rendering the wealth accumulated in our current system irrelevant within their borders.
- Scalability & Incremental Implementation: The design would be modular and scalable, much like open-source architecture with containerized modules. Everything would be freely available, ensuring that individual contributions benefit the whole.
There is a bunch of more concepts that tie into the above such as Degrowth but due to me being a slave to the rat race, I don’t have all the time I wished I had to write about all of these as well.
Don’t lose hope comrade!