As an aside- I'm also an academic and resent this sudden nonsense push to 'embrace AI' from people that have no clue what it even is. Some of my research is close enough to be getting renewed interest and funding from this... but I find it offensive the my same research ideas weren't interesting on their own merits a few years ago but only suddenly are now because they involve 'AI.' It suggests a frivolous trend following unbecoming of scientists.
I don't know what this author researches, but she deeply understands why technology needs to be aligned with human interests, and how to make sure it is. That seems to me something academic departments interested in AI research need now more than ever.
I've also always found this poem darkly captivating, because it imagines that positive humanizing technology will be possible but itself seems like a vision that would be infantilizing, and the author leaves it unclear on if they notice this or not, and if it is serious or dark satire.
I hope her inbox is filled with sympathy, and promises to destroy the dehumanizing parts of women's health.
I would posit that it is the people using the technology that also need to be empathetic and ensure the use is not disturbing.
I want to prevent everyone else from getting distracted by the shininess of it all, because the world really needs all these scientists and organizations to keep progressing the substantial work they are best at.
The recent Nobel in physics was the perfect example, and I think HN echoed the general feeling that we mightily respect the field, probably above what we are doing, and that it is an outrage for it to get overshadowed like that.
It's also weird to see how not being focused on AI is disrupting so many people, but being focused on AI doesn't help you that much either. It is definitely not the trump-card people think it is with investors, and even less with the market. Sometimes it is better not to emphasize it because it cheapens the whole thing! Among other things, people (understandably) do not understand the spectrum of possibilities for using LLMs, in their minds it all boils down to asking ChatGPT, which is obviously not very substantive. At the end of the day, it's all just technical details, people want results and value.
What are you talking about?
> the Dean of my college told me...that I should look for long-term academic employment elsewhere. My research and practice was not centered enough on “AI” and “emerging technology” to fit within the institution...
Follow the money.
If you're used to "follow the money" and have done it a lot, you may find this money trail doesn't lead to the usual places. But look at the size of the AI bubble. There's a lot money available to fund PR and push "influencers" to run around telling people that they need to get on board or get left behind... because if people don't get on board that AI bubble is going to pop. And I think people underestimate the amount of control a donation to an educational institution gets the donor, unless they are very cynical.
I do not claim or believe that "AI" is useless. I do firmly believe that on top of the foundation of useful stuff that it can do, some of it if anything underutilized and understudied (I agree with an article in HN a few days back that embeddings are really underutilized and studied versus all this generative stuff which is a lot less interesting than meets the eye), there is a staggeringly enormous bubble. And with that bubble, a lot of well-resourced people motivated to keep it inflated, as there always is.
You'd think by 2024, the techno-fetishist "If it's NEW technology it's GOOD technology!" would be dying down in the general consciousness. Tech is not automatically good and helpful just by virtue of being good.
Echoes of Facebook showing people “memories” (photos) of their deceased children on their bday. This is one of the saddest stories I’ve read in a bit.
An LLM probably could have warned them before they sent the email, if only it had been integrated into their email client.
https://en.m.wikipedia.org/wiki/All_Watched_Over_by_Machines...
here is the direct link to the poem.
(sorry to be That Person, but I'm really hoping Hacker News doesn't eventually become like Reddit where the comments are almost completely divorced from the content of the article).
And I think the poem is satire, although some have read it as being serious.
I like that poem, and I hope we eventually get there, with our artificial descendants as our partners.
Spinning like a ghost
on the bottom of a
top,
I'm haunted by all
the space that I
will live without
you.
From Wikipedia: Curtis argues that computers have failed to liberate humanity, and instead have "distorted and simplified our view of the world around us".
https://en.wikipedia.org/wiki/All_Watched_Over_by_Machines_o...
/s
With that in mind, I still find his opinions interesting, original and provocative. Also, his films have some of the coolest imagery ever — if I recall correctly he used to work at the BBC, dealing with their archives.
In a community that has no explicit rules, implicit rules emerge. Power accumulates quietly to the people who know the rules and can bend them often. In those situations the powerless are even more vulnerable to the powerful because of the lack of an explicit power structure with rules that would usually require checks and balances in order for people to opt into the system.
There's rarely anything more in people's political prescriptions other than, "I do not have enough power, and I want more" -- and there's nothing inherently wrong with this demand. Only that when it is disguised by the false promise of power for everyone, it means power for almost no one.
People have been collectively organising movements and spaces for many decades now, and the processes are pretty well worked out. Comparing utopian dropouts in the 60's running farms on vibes to dedicated activists establishing and maintaining consensus based social contracts is disingenuous. There's plenty to be criticised in terms of the practical aspects of these systems - from implementation to scalability, and good luck to the folks doing the hard work of improving them. But dismissing them out of hand is an argument from ignorance.
Small groups. That is the problem. It can work but doesn't scale.
As the number of people go up, the need and benefits of more complex coordination go up. People's interactions, dependencies, needs, understandings and problems multiply in quantity, complexity and diversity.
People O(n). Decisions O(n^2).
Direct coordination becomes too costly in time and effort. And people get asked to make decisions in areas where they have no skin in the game, so put little care into their decisions, or seek selfish benefits opportunistically. Things get difficult and ugly.
The default is a descent into anarchy, which gets counter weighted by emergent initiatives that centralize decision making around people who care, have expertise, are good at taking power, etc. Acknowledged or not, "direct", as the universal principle of decision making, no longer works.
Better to see it coming and as directly and openly as possible, prescribe how indirect governance is to be done. As openly and accountably as possible.
Organizations and societies, like computer memory and and processing cores in server centers, need to bifurcate and modularize into different levels, types and policies, to remain efficient and reliable at scale.
So no - they are not a direct democracy. For the unavoidable reasons I gave.
However yes, they are doing a fantastic and inspiring job of maintaining as much directness as possible. And at keeping the centralized and indirect processes formalized, open and accountable.
Vs. the US, where even the president isn't elected directly; lower offices are heavily controlled by whoever draws district maps, often the incumbents; only two parties have much power, and conspire together to maintain that - reducing citizen choices to only two preset menus of policies & governance styles; parties enforce litmus tests on their own candidates at all levels of office, vastly increasing party leadership power at the expense of eliminating 99% of potential alternative views. The result of all this power hoarding are many laws in significant opposition to majority views, and an institutionalization of two way service and influence with the rich. The highest court nomination process and members of the highest court have become aggressively politicized and brazenly unethical. Overall, the US system has become increasingly dysfunctional, corrupt, divisive and dystopian.
Switzerland is amazing.
We've somewhat arbitrarily settled on nation state scale geographic / population scales for governance which (a few less than democratic mega-nations aside), are pretty similar in size and complexity. This is a local maxima, rather than an inevitability. The ideal balance between federalised aspects of governance - which do need to be abstracted to some extent isn't obvious. It's reasonably likely that much smaller micro-states within stronger federalised economies could be more effective. That workplace level democracy can and does increase productivity. That in fact states where the level of consensus decision making is the business, the cooperative and the housing complex, could be more cohesive, less fractious and more stable. Could be more effective if we judge efficacy as best meeting the wellbeing needs of their population.
I'd take issue with the idea that centralisation of decision making is necessitated by having multiple layers of governance. I'd argue a canton style system of carrying forward the decisions and decision making authority of lower houses can also work. Where representation is absolutely essential, sortation can be used to prevent the creation of a distinct political class that is inevitably suborned by capital.
In terms of skin in the game - essentially all forms of consensus (which it's important to point out is radically different in theory and practice from direct democratic voting), require participation as the price of involvement. There is no voting without engagement, in fact in most consensus systems there is no voting at all. Participants become politically sophisticated through their involvement, and their level of involvement inevitably dictates their level of influence. For this to work, groups have to structure themselves such that a lot of the groups effort is spent engaged in decision making. Many of the specifics of the various consensus approaches push proponents of decisions into being responsible for the implementation of those decisions. There's no pretence that people don't affiliate, don't form minimal group paradigm loyalties, aren't selfish, or don't manipulate one another - an understanding of these processes is built into such systems. In fact they what leads to the cohesion that allows consensus to function.
But my key insight here is that many of ills of modern society arise from the idea that things should in fact 'scale'. Climate change and ecocide, the homogenisation of culture, the commodification of everyday life, the alienation that leads to polarisation and sensitisation and ultimately political extremism. All of it's rooted in an appeal to competing in a global scale market economy where goods and services move freely and people do not. Where all nations must grow or stagnate, where legal systems and free trade must be imposed at the barrel of a gun in order to allow the extractive system to continue to expand. This is one possible way for the world to work, in no means inevitable. Societies have existed with radically different affordances and compromises historically.
Somehow, one way or another, dispute resolution has to scale to the number of people in the world. That isn't a chosen scale, it is the scale determined by present conditions.
For that to happen, given how many disputes exist between 8 billion people and all their groups and subgroups, there is going to be non-direct decision making.
You can point at subgroups with direct decision making, which is great when it works well, but that only works between those people, and they will invariably care about many things that extend beyond their group.
This isn't the case today. A small number of superpowers resolve territorial disputes mostly through court cases and negotiations, sometimes through violence. In all of these cases a tiny number of people are involved in the decision making process. What's happening is not dispute resolution at scale, its capricious and stochastic decision making corrupted by the same forces of transnational capital that put those decision makers in charge.
The global south isn't making any of the key decisions around climate change, AI, international trade etc. Neither are most developed countries. It's at most a cohort of G8 leaders and billionaires. It doesn't work, and moreover its failing - as the collapse of the liberal political order demonstrates. Our choice is to allow it to collapse into a new form of feudalism, or to actively participate in replacing it at every level of governance through civil dissent, and consensus based approaches.
That isn't anything like direct is it?
And even more disputes are settled in very indirect implicit ways. i.e. military capability is an implicit threat that tilts all kind of disputes in one direction or another.
I agree that global decision making isn't keeping up with the world's needs and problems. But pointing out something isn't working well is the easy part.
Improving on it, as apposed to simply critiquing it, is the first hard part. Or worse, tearing down what we have that works because it is imperfect, without a better system to replace it. As happens throughout history.
Much more difficult to design an alternate system that works - so unless someone has a serious alternative, improving what we have is best/only path forward. What does work today took millennia of slow progress.
Then the second hard part is the challenge of getting changes accepted and implemented. Since in the short run, everything is pushed and pulled by vested economic interests. Non-direct decision making not in the open, not any agreed upon process, and with little accountability.
The latter is the biggest problem. Especially when it corrupts what we have, as far as open and formalized decision and accountability systems. I.e. when it corrupts our politics, politicians, justice system, etc. Creating dense areas of hidden self-interested decisions trading in misdirected public power.
The most basic answer for most global problems is to start removing corruption. You can't do much if you don't get good sleep and healthy food and water. Likewise, massive global problems being mismanaged are going to need a less corrupt environment to make faster headway.
Otherwise, reality will eventually convince enough people to solve problems. That is a cost maximizing path, but surprisingly popular.
Day to day their government is run via elected representatives, just like almost anywhere else.
Looking at it that way, though, I find myself concerned with the people who are setting themselves up as the arbiters of my fate. Regardless of party affiliations, they don't seem to think very highly of me or care about my personal happiness.
Seeing that is the case, I'd rather they stay out of the business of pursuing my happiness and instead support my freedom to pursue happiness. I can be responsible for my own happiness.
I find this meme, in the original sense, rather odd. Speaking as an american, over the last 40 years the right wing of our political institutions have been extremely hierarchical and authoritarian. Right wing ideologies are almost entirely based on which group should obey which other group and why.
The term left-wing has gotten a little vague recently, but I think you could say that the common premise of most of their political theory is that there's already/always going to be a powerful government that everyone has to obey so we might as well make that government the best it can possibly be for as many people as possible.
(I have a personal theory that political power, much like energy, can never actually be destroyed, merely moved.)
This journal article discusses power in two forms:
- 'Power to' (freedom, from others): a variable-sum game
- 'Power over' (others, ultimately denying freedoms): a zero-sum game
I'm no scientist and this is just from the abstract: https://www.tandfonline.com/doi/full/10.1080/00323187.2020.1...
You are being way to generous here. You are suggesting that no matter how much we deny freedoms and give some people power over others, the sum will not change? (And in reverse, no matter how much we stop denying freedoms etc, the sum will also still stay fixed?)
I doubt that.
I mean, no I really don't summarize it that way at all.
>Conservatism consists of exactly one proposition, to wit: There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect
Is a better, but much more politically charged saying.
Moreso, self determination is more of a libertarian thing rather than right directly.
This is The Tyranny of Structurelessness: https://www.jofreeman.com/joreen/tyranny.htm
The Tyranny of Tyranny by Cathy levine
https://theanarchistlibrary.org/library/cathy-levine-the-tyr...
Dream lovers are invariably machine lovers. A dream lover lives entirely in dreams. To the degree that he absolutely cannot avoid interacting with reality, he does it through machines.
So that's something I think about.
This image of an electronic garden - leptospirosis from cat sitting - has flooded nature with server calls. Now, a tree does not grow of its own accord, but in the structure of a system. Furthering its seed, the redwood pollinates through incineration. The technological is constantly in flux.