I chose the lowest probability for “your quiz proposition “AI will destroy humanity.” Had the proposition been “Humanity will use AI to destroy ourselves,” I’d have chosen a higher probability range.
As well as autonomous weapons, I was also thinking of the use of AI fakery and massive disinformation campaigns to frustrate the measures we need to take for our survival and the survival of the biosphere. Social media disinformation has tilted the political playing field towards irrationality, and I fear AI will be used to tilt it further.
Greenwashing becomes quite terrifying. We live in the information age and it’s been very effective for the past several decades of it so far… don’t really want to give BP the power of iteratively mimicking a human at the cost of only precious metals, water, and unsustainable electricity generation. Literally corporations using LLM to spread pro-fossil fuel propoganda will be them burning the planet to be allowed to continue burning the planet.
A fascinating examination of the current problem of 'democracy' in our modern representative systems. I'm 30 years old and used to work as a manager for one of the 'Magnificent Seven' giants of Big Tech that briefly operated out of New Zealand. Dealing with the dangers of AI and LLMs will definitely be one of the biggest challenges for people of my generation that hold a rudimentary understanding of how it has developed so far...
I'm torn over whether AI can, or should, lead to a more Athenian form of democracy. My work as a manager showed me that the general public, even my peers, did not have a good philosophic understanding of computers. The current user interface is too easy to use, so most users have never seen the need to educate themselves about how it came into being.
When the AI revolution took hold post-COVID, I read up on the new technical advancements of LLMs and neural net algorithms. I've been playing around with Google Gemini, and found pretty much the same stuff you have with ChatGPT. It is a good source for broadly accurate data analysis, and more aware of its biases than many human interlocutors.
I personally have no issue with AI generated 'fake news', mostly because political polemics, satire, rumormongers, etc... have long been a feature of our representative democracies. In particular, I look back to the political war between John Adams and Thomas Jefferson in early American history - one where they both funded shocking amounts of 'yellow journalism' to discredit the other candidate and their ideas.
This war for what we would now call 'low information' voters affected their friendship at the time, but they both came out the other side and had a rich correspondence in their later year. So, based on this example, I'm not very worried about the spread of 'disinformation' as the U.S.A managed to survive this 'fake news' war between two of their Founding Fathers.
However, I'm only optimistic so long as people continue to trust our government and our political system, even if they despise many of the people who work in it...
I suspect we will probably see a massive increase in 'yellow journalism' over the next few years as political parties and activists become comfortable with the use of AI in their election campaigns, but voters will eventually get fed up and trust in our political system will continue to fall. This might lead to a revolt - but my concern is that, instead of revolting, people will entrust their governance to 'Big Tech' and other corporations because they don't understand the ancient philosophies of politics.
If this happens, humanity will drift away from democracy entirely and into the realm of a corporate form of government. Hopefully this doesn't happen, and the internet finally fulfills its promise of being a genuinely democratic marketplace of ideas, but who knows how this revolution in communications technology will play out?
There is a great series of history books about the fallout between the Founding Fathers post 1789 by Joseph E. Ellis including American Sphinx: The Character of Thomas Jefferson (1996). In the book, he discusses terms such as 'yellow journalism' or 'yellow rags' as the way commentators at the time referred to the political events of that election cycle.
Basically, Jefferson developed a technique of funding low quality print publications (hence the word yellow) with scurrilous gossip about John Adams personal character, then spread these widely in electorates which might take the gossip seriously. Adams did the same thing in retaliation.
Jefferson won eventually, but both candidates later regretted that they had crossed a line as politicians - in their letters later in life, they questioned whether opening the floodgates for ad hominem attacks through third party media had actually been a good idea.
At the start of 2023 I listened to a classicist give a talk on his analysis of the fall of Rome through our modern political lens, and he was especially interested in how the breaking of constitutional convention lead to deteriorating political behaviour that undermined the very constitutional soundness of the Roman political system, a cycle we are definitely watch play out in our own global and domestic politics right now.
So it’s somewhat reassuring to hear that these events have often been occurring in our politics without necessarily leading to the total collapse of civilisations. But it does also seem to be the case that underhanded plays by one party will too often lead to the other party “sinking to their level”, and that if that repeats a few times, politicians can find themselves standing on quite low-lying ground by the end of it.
How did Adam’s and Jefferson (or their parties) go about “pulling back” on these tactics in future, since they came to regret them?
Well, it was the spoils system for a while as I recall, then they reformed a bit in the 1890s?
Ad-hominem attacks become so popular that they were aired in the democratic house of representatives, including the infamous 'Golden Spoon' orations of 1840: https://en.wikipedia.org/wiki/Gold_Spoon_Oration
Both parties attacked each other's personal character more than their policies, and whenever a party won they would 'clean house' and take the 'spoils of war' by appointing their party officials across the civil service to collect paychecks.
AI is at its best when used as a supplement to human intelligence, rather than as a complete replacement for it. I've come to realise that if/when Skynet arrives, it's less likely to declare war on the human race, and more likely to provoke the human race into declaring nuclear war on itself.
As with any other technology, the biggest threat isn't AI itself, but the hijacking & monopolisation of it by the Elon Musks & Vladimir Putins of this world.
With regards to trusting AI: users have just discovered that attempting to make ChatGPT say the name David Mayer (a Rothschild, one of the most influentieatinl and wealthy families in history) will make the programme crash. A few others have now been discovered too. There is a layer of coding in the shell of the AI that is designed to remove information on specific people, and this is hidden from users.
To my mind, the programmers are just the employees of large corporations who benefit from their technology not being transparent and who I expect will have often a sizeable incentive to use AI in inaccurate or manipulative ways.
I don’t understand AI in terms of how the LLMs work on a technical level, but reading people reverse engineering the prompts on Reddit to try and work out what was even happening with the David Mayer “glitch”, let alone why it was happening (jury is still out), was very interesting. I wonder how much benefit we (humanity) would get from a legal requirement that a certain amount of an AI programme be open source, just to provide that transparency that allows the public to provide basic checks for these programmes and what/how they’re doing the things that they’re doing.
The Swiss referendum system is both a blessing & a curse. On the one hand, it's more responsive than a Westminster representative system. On the other hand, it meant Swiss women weren't allowed to vote until 1971 - yes, the same year John Lennon's "Imagine" was released.
Interesting thanks Grant... but on the question/subject of "democracy", I quote from your post: "...Later, the American colonists threw off King George III’s “tyrannical” government, partly on the principle of “no taxation without representation”..."
I suspect that that should have a riposte: "no representation without income paying tax!"
My view is that "democracy" as we know it today has gone past it's use-by date due to so many electors living off taxpayers instead of working for and earning a living.
JS Mill would agree with you, as he thought those on parish support should not vote. One of the things that supported a universal franchise was indeed income tax, but now everyone pays sales tax, don't they?
I chose the lowest probability for “your quiz proposition “AI will destroy humanity.” Had the proposition been “Humanity will use AI to destroy ourselves,” I’d have chosen a higher probability range.
Insofar as that Rome was a democracy, it was a representative democracy in which politicians were voted on to represent the interests of the people, especially the Tribunes who were elected by the common people to act on their behalf. So the idea that representative democracies didn’t exist in the ancient world isn’t strictly true.
But this is interesting of itself — Rome borrowed from and based a lot of its concepts on Greece and Greek scholarship, as they admired the Greeks particularly and sought to emulate them in many spheres. Unlike the Greek city states, and like the British political system, Rome was never MEANT to be a democracy, only a system of representation for the landed elite to have a say in government; it was very specifically a representative oligarchy and not a democracy. The Tribunes gave small representation to the plebeian classes, but this was a compromise and the result of uprisings and protests over the economic burden the wealthy elite had placed upon the working population, and their power was later curbed under the Empire.
It’s notable that between the system that brought democracy to the ancient world and the system that was implemented to ensure the continued power of the wealthy elite, we have gone with the latter.
A mix — it was a hybrid of direct and representative democracy, much more representative than direct. People who owned nothing couldn’t vote most of the time except for the people’s Tribunes who had power of veto over the senate.
Rome considered itself a democracy (republic comes from what Rome called itself, res publica, literally “for the people”) but it was a democracy where the people who wanted to give themselves power were rich, so it was a democratic aristocracy. This means there was an element of “direct government” from the perspective of the Roman elite — if you were a pleb though, that was meaningless to you. So ideas of direct democracy here are a bit complicated by Rome not actually being a very good democracy, in terms of our concept of the word.
Roman’s voted for censors who then selected senators, while magistrates selected their replacements for the approval of the people (or sometimes the senate).
Voting by the people was done in blocks (assemblies) based on ancient tribal alliance, and some of these centuries did vote directly on laws, but was also used to elect representatives. However the more powerful and important centuries got to vote first, so if a majority was reached, it was common for lesser centuries to not need to vote at all. Pre-augustus, priests (who held political power) were also elected by voting assemblies.
I chose the lowest probability for “your quiz proposition “AI will destroy humanity.” Had the proposition been “Humanity will use AI to destroy ourselves,” I’d have chosen a higher probability range.
Good point, Kai! Autonomous weapons are a real worry.
As well as autonomous weapons, I was also thinking of the use of AI fakery and massive disinformation campaigns to frustrate the measures we need to take for our survival and the survival of the biosphere. Social media disinformation has tilted the political playing field towards irrationality, and I fear AI will be used to tilt it further.
Greenwashing becomes quite terrifying. We live in the information age and it’s been very effective for the past several decades of it so far… don’t really want to give BP the power of iteratively mimicking a human at the cost of only precious metals, water, and unsustainable electricity generation. Literally corporations using LLM to spread pro-fossil fuel propoganda will be them burning the planet to be allowed to continue burning the planet.
A fascinating examination of the current problem of 'democracy' in our modern representative systems. I'm 30 years old and used to work as a manager for one of the 'Magnificent Seven' giants of Big Tech that briefly operated out of New Zealand. Dealing with the dangers of AI and LLMs will definitely be one of the biggest challenges for people of my generation that hold a rudimentary understanding of how it has developed so far...
I'm torn over whether AI can, or should, lead to a more Athenian form of democracy. My work as a manager showed me that the general public, even my peers, did not have a good philosophic understanding of computers. The current user interface is too easy to use, so most users have never seen the need to educate themselves about how it came into being.
When the AI revolution took hold post-COVID, I read up on the new technical advancements of LLMs and neural net algorithms. I've been playing around with Google Gemini, and found pretty much the same stuff you have with ChatGPT. It is a good source for broadly accurate data analysis, and more aware of its biases than many human interlocutors.
I personally have no issue with AI generated 'fake news', mostly because political polemics, satire, rumormongers, etc... have long been a feature of our representative democracies. In particular, I look back to the political war between John Adams and Thomas Jefferson in early American history - one where they both funded shocking amounts of 'yellow journalism' to discredit the other candidate and their ideas.
This war for what we would now call 'low information' voters affected their friendship at the time, but they both came out the other side and had a rich correspondence in their later year. So, based on this example, I'm not very worried about the spread of 'disinformation' as the U.S.A managed to survive this 'fake news' war between two of their Founding Fathers.
However, I'm only optimistic so long as people continue to trust our government and our political system, even if they despise many of the people who work in it...
I suspect we will probably see a massive increase in 'yellow journalism' over the next few years as political parties and activists become comfortable with the use of AI in their election campaigns, but voters will eventually get fed up and trust in our political system will continue to fall. This might lead to a revolt - but my concern is that, instead of revolting, people will entrust their governance to 'Big Tech' and other corporations because they don't understand the ancient philosophies of politics.
If this happens, humanity will drift away from democracy entirely and into the realm of a corporate form of government. Hopefully this doesn't happen, and the internet finally fulfills its promise of being a genuinely democratic marketplace of ideas, but who knows how this revolution in communications technology will play out?
Great thoughts, thanks Stephen. I'll come back to political trust soon.
Interesting term, ‘yellow journalism’. I wonder how it compares directly to clickbait journalism and its effects.
Before Elon Musk, there was Rupert Murdoch. And before Murdoch, there was William Randolph Hearst. And so forth.
It seems to be a similar concept.
There is a great series of history books about the fallout between the Founding Fathers post 1789 by Joseph E. Ellis including American Sphinx: The Character of Thomas Jefferson (1996). In the book, he discusses terms such as 'yellow journalism' or 'yellow rags' as the way commentators at the time referred to the political events of that election cycle.
Basically, Jefferson developed a technique of funding low quality print publications (hence the word yellow) with scurrilous gossip about John Adams personal character, then spread these widely in electorates which might take the gossip seriously. Adams did the same thing in retaliation.
Jefferson won eventually, but both candidates later regretted that they had crossed a line as politicians - in their letters later in life, they questioned whether opening the floodgates for ad hominem attacks through third party media had actually been a good idea.
At the start of 2023 I listened to a classicist give a talk on his analysis of the fall of Rome through our modern political lens, and he was especially interested in how the breaking of constitutional convention lead to deteriorating political behaviour that undermined the very constitutional soundness of the Roman political system, a cycle we are definitely watch play out in our own global and domestic politics right now.
So it’s somewhat reassuring to hear that these events have often been occurring in our politics without necessarily leading to the total collapse of civilisations. But it does also seem to be the case that underhanded plays by one party will too often lead to the other party “sinking to their level”, and that if that repeats a few times, politicians can find themselves standing on quite low-lying ground by the end of it.
How did Adam’s and Jefferson (or their parties) go about “pulling back” on these tactics in future, since they came to regret them?
Well, it was the spoils system for a while as I recall, then they reformed a bit in the 1890s?
Ad-hominem attacks become so popular that they were aired in the democratic house of representatives, including the infamous 'Golden Spoon' orations of 1840: https://en.wikipedia.org/wiki/Gold_Spoon_Oration
Both parties attacked each other's personal character more than their policies, and whenever a party won they would 'clean house' and take the 'spoils of war' by appointing their party officials across the civil service to collect paychecks.
Eventually civil service appointments were removed from the ongoing ad hominem battles of politics by the Pendleton Act (1883): https://en.wikipedia.org/wiki/Spoils_system
AI is at its best when used as a supplement to human intelligence, rather than as a complete replacement for it. I've come to realise that if/when Skynet arrives, it's less likely to declare war on the human race, and more likely to provoke the human race into declaring nuclear war on itself.
As with any other technology, the biggest threat isn't AI itself, but the hijacking & monopolisation of it by the Elon Musks & Vladimir Putins of this world.
Scary!
With regards to trusting AI: users have just discovered that attempting to make ChatGPT say the name David Mayer (a Rothschild, one of the most influentieatinl and wealthy families in history) will make the programme crash. A few others have now been discovered too. There is a layer of coding in the shell of the AI that is designed to remove information on specific people, and this is hidden from users.
We should already be asking how much we trust AI.
Interesting! I'd say we put trust in programmers, rather than in AI itself. So the programmers should be open about any such bugs or features.
To my mind, the programmers are just the employees of large corporations who benefit from their technology not being transparent and who I expect will have often a sizeable incentive to use AI in inaccurate or manipulative ways.
I don’t understand AI in terms of how the LLMs work on a technical level, but reading people reverse engineering the prompts on Reddit to try and work out what was even happening with the David Mayer “glitch”, let alone why it was happening (jury is still out), was very interesting. I wonder how much benefit we (humanity) would get from a legal requirement that a certain amount of an AI programme be open source, just to provide that transparency that allows the public to provide basic checks for these programmes and what/how they’re doing the things that they’re doing.
Perhaps we need to look to Switzerland - where enhanced democracy rules.
More referendums then?
Yes!
The Swiss referendum system is both a blessing & a curse. On the one hand, it's more responsive than a Westminster representative system. On the other hand, it meant Swiss women weren't allowed to vote until 1971 - yes, the same year John Lennon's "Imagine" was released.
https://www.swissinfo.ch/eng/society/political-history_the-controversial-minaret-ban-ten-years-on/45399822
Interesting thanks Grant... but on the question/subject of "democracy", I quote from your post: "...Later, the American colonists threw off King George III’s “tyrannical” government, partly on the principle of “no taxation without representation”..."
I suspect that that should have a riposte: "no representation without income paying tax!"
My view is that "democracy" as we know it today has gone past it's use-by date due to so many electors living off taxpayers instead of working for and earning a living.
JS Mill would agree with you, as he thought those on parish support should not vote. One of the things that supported a universal franchise was indeed income tax, but now everyone pays sales tax, don't they?
Yes, but some pay sales tax from taxpayer-funded "benefits" so they more than cancel out... that's why I suggested "INCOME" tax as the qualifier.
Ok then. You could start up a citizens' initiated referendum on that.
Won't work (yet): the electors in question will not like being disenfranchised and the workers haven't yet realised what's going on!
Or establish a micro-nation. Oh wait, it's already been tried before...
https://www.vice.com/en/article/atlas-mugged-922-v21n10/
https://washingtonmonthly.com/2020/08/30/libertarians-took-control-of-this-small-town-it-didnt-end-well/
https://slate.com/news-and-politics/2022/05/michael-oliver-republic-of-minerva-history-libertarian-micronations-tonga.html
Mine’s flourishing
I chose the lowest probability for “your quiz proposition “AI will destroy humanity.” Had the proposition been “Humanity will use AI to destroy ourselves,” I’d have chosen a higher probability range.
Insofar as that Rome was a democracy, it was a representative democracy in which politicians were voted on to represent the interests of the people, especially the Tribunes who were elected by the common people to act on their behalf. So the idea that representative democracies didn’t exist in the ancient world isn’t strictly true.
But this is interesting of itself — Rome borrowed from and based a lot of its concepts on Greece and Greek scholarship, as they admired the Greeks particularly and sought to emulate them in many spheres. Unlike the Greek city states, and like the British political system, Rome was never MEANT to be a democracy, only a system of representation for the landed elite to have a say in government; it was very specifically a representative oligarchy and not a democracy. The Tribunes gave small representation to the plebeian classes, but this was a compromise and the result of uprisings and protests over the economic burden the wealthy elite had placed upon the working population, and their power was later curbed under the Empire.
It’s notable that between the system that brought democracy to the ancient world and the system that was implemented to ensure the continued power of the wealthy elite, we have gone with the latter.
How democratic or representative was the Senate?
A mix — it was a hybrid of direct and representative democracy, much more representative than direct. People who owned nothing couldn’t vote most of the time except for the people’s Tribunes who had power of veto over the senate.
Rome considered itself a democracy (republic comes from what Rome called itself, res publica, literally “for the people”) but it was a democracy where the people who wanted to give themselves power were rich, so it was a democratic aristocracy. This means there was an element of “direct government” from the perspective of the Roman elite — if you were a pleb though, that was meaningless to you. So ideas of direct democracy here are a bit complicated by Rome not actually being a very good democracy, in terms of our concept of the word.
Roman’s voted for censors who then selected senators, while magistrates selected their replacements for the approval of the people (or sometimes the senate).
Voting by the people was done in blocks (assemblies) based on ancient tribal alliance, and some of these centuries did vote directly on laws, but was also used to elect representatives. However the more powerful and important centuries got to vote first, so if a majority was reached, it was common for lesser centuries to not need to vote at all. Pre-augustus, priests (who held political power) were also elected by voting assemblies.