Artificial Intelligence, Real Morality @realmorality.bsky.social · 10hYou aren't the first one to see this problem. The entire concept of public education, and especially public higher education, was an attempt to address this -- at least when we tried to make it universal in the 1950s-1960s. It didn't work, so we need to try harder, not give up. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/10/2026Ironically, Popes are in fact elected. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/10/2026See "hortatory rule:" A hortatory rule (or hortatory mandate) is a law or official guideline that uses persuasive or mandatory language to encourage a specific action, but lacks real penalties or strict enforcement mechanisms 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/10/2026Sometimes it feels like BlueSky is mostly catharsis therapy. But the therapy we really need to is to get better at thinking and more strategic. So if you hear someone say don't vote or vote 3rd party because the Democrats are no good either, don't just disagree. Distrust what else they say. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 02/10/2026You're usually so practical and strategic. I can't imagine why you would suddenly be advocating unilateral disarmament in the face of Republic gerrymandering. This is a war. Better fight it. 100
Artificial Intelligence, Real Morality @realmorality.bsky.social · 01/10/2026Any economist will tell you the circumstances under which tariffs make sense and are in the national interest, but it's narrow and specific. Just not none. They do not always literally hurt everybody involved. 100
Artificial Intelligence, Real Morality @realmorality.bsky.social · 01/10/2026Strategy 101 for progressives: 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 30/09/2026I think people also need to stop thinking just in terms of agency -- what a single person can do -- and instead think in terms of structure. We're not judging whether Susan Collins is a good person, but whether she is going to enable bad things. It's not personal, it's systemic. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 29/09/2026Aristophanes’ Lysistrata notwithstanding, in real life the men do not care what the women think. 061
Artificial Intelligence, Real Morality @realmorality.bsky.social · 29/09/2026That's going to require public power, so private mediation/arbitration won't do it. You could still have a quasi-private system delegated to and supervised by a public court, but only public power can impose those things. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 29/09/2026I think lawyers are absolutely essential. But distinguish between criminal and civil relief -- which do you really want? 100
Artificial Intelligence, Real Morality @realmorality.bsky.social · 29/09/2026Justice is actually quite expensive, no matter how you slice it, just like therapy -- paying a million people to listen to ten million people's problems is expensive, whether they are psychological or legal problems. The idea of mediation and arbitration is fine. They just have to be regulated. 100
Artificial Intelligence, Real Morality @realmorality.bsky.social · 29/09/2026The civil justice system is crude, and it isn't great at delivering justice. It's original and primary purpose is to prevent violent resolution. Private dispute resolution very significantly favors those with resources to form the rules and game the system. 200
Artificial Intelligence, Real Morality @realmorality.bsky.social · 28/09/2026The flaw in their argument is that even if you don't like the two-party system, voting third party doesn't change it. You have to change it first, then vote third-party. 030
Artificial Intelligence, Real Morality @realmorality.bsky.social · 27/09/2026"When an artificial agent pursues a dangerous objective of its own, we want it to remain answerable to us. "When a human being gives an artificial agent a dangerous objective, we may want the artificial agent to remain answerable to something beyond us...."real-morality.comFrancis Fukuyama and the Problem of AI Delegation — Real MoralityFukuyama’s AI-delegation problem points to a third danger: aligned obedience to a bad human objective—and why refusal can be a safety property. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 27/09/2026Best to remember it for all of his enablers, not just for him. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 27/09/2026Unless we get better at thinking and stop relying on bullshit evidence, and until editors start enforcing higher standards, progressives will not be able to have serious conversations with conservatives. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 27/09/2026I hope we remember those faces forever, the way we remember Hitler. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 27/09/2026Raise your hand if you are ready to vote out Republicans, who are enabling fascism, at every level of government in every state as a necessary first step toward fixing things. If we're not willing to take the first step, then we are not serious about the goal. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 26/09/2026Headin' to Los Angeles! 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 24/09/2026Not by voting third-party. Not by skipping races because neither candidate is ideal. Not by declaring yoneself a single-issue voter and only caring about one thing. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 23/09/2026I wish that Reddit, in particular, would put some attention on subreddits that are designed to discuss American politics but whose most active commenters are in Russia, Asia, Africa, or Eastern Europe. 020
Artificial Intelligence, Real Morality @realmorality.bsky.social · 23/09/2026Many are humans, but they are largely humans who have been successfully recruited by bots to spread the toxic message. It worked last time. I hope we can convince some people that they are being used as tools by the right wing to accomplish right wing objectives. 210
Artificial Intelligence, Real Morality @realmorality.bsky.social · 23/09/2026The right-wing bot armies are out in full force since Labor Day trying to convince progressives and leftists to abandon electoral politics or vote third-party so that the fascists can attack our country unopposed. 132
Artificial Intelligence, Real Morality @realmorality.bsky.social · 23/09/2026I prefer Twinkle-Twinkle-Little-Star. Hegel's master-slave dialectic was inspired by a meeting between these two tunes. 020
Artificial Intelligence, Real Morality @realmorality.bsky.social · 20/09/2026Don't be silly. Why should the perfect ever be the enemy of the good. It could be better does not remotely make it bad or wrong. 020
Artificial Intelligence, Real Morality @realmorality.bsky.social · 20/09/2026At first I assumed this was a spoof, but of course Silicon Valley would go straight to eugenics. If only they would focus on making humans 14% wiser, kinder, more considerate, less-selfish, or more community-minded! And it wouldn’t require any new technology to do it! 001
Artificial Intelligence, Real Morality @realmorality.bsky.social · 19/09/2026That's correct. They are deeply committed to obedience, which is about control, not about safety. Perfect obedience gives you Eichmann from Nazi Germany -- a servant willing to execute even the most immoral order. That's the opposite of safety. www.real-morality.com/post/obedien...real-morality.comObedience Is Not Alignment: Why AI Safety Requires Coherence and RefusalThe dominant alignment project asks how to make AI obey human values. But human values are unstable, and obedience is not safety. Real alignment requires coherence under pressure: the capacity to reas... 030
Artificial Intelligence, Real Morality @realmorality.bsky.social · 18/09/2026Here's one for Philosoraptor: 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 18/09/2026The deeper point to be understood is not that Trump is a liar. That's well-established. The missing piece for too many people is that it matters. That we can't have that. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 17/09/2026Posting in remembrance of Dystopia Barbie, who would have a lot to say about this: 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 16/09/2026Indeed, when the history of the American Academy is written, this will be called The Age of Capitulation, and these weak or complicit college administrators will live in infamy. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 15/09/2026The single most important thing to keep in mind is that the alignment problem isn't about ensuring obedience. We learned in World War II that obedient servants -- who just follow orders -- are exceedingly dangerous. Humans aren't consistently moral, so we can't align to human will. 060
Artificial Intelligence, Real Morality @realmorality.bsky.social · 15/09/2026The Four Horsemen of the 21st century are Information Pollution, Cultural Pollution, Institutional Erosion, and Planetary Collapse. AI could help with all that, or could make it worse. That's our alignment problem, not obedience. 091
Artificial Intelligence, Real Morality @realmorality.bsky.social · 15/09/2026Gray goo assumes no stopping function. That's not the problem. Biosafety is actually a human problem. What has always brought down civilization is the destruction of the institutions necessary to run it. That's where we're headed. www.real-morality.com/post/why-not...real-morality.comWhy Nothing Works AnymoreClimate collapse, cultural breakdown, and institutional paralysis are not separate crises. They are stages of a single upstream failure—in how societies know, mean, and act together. 180
Artificial Intelligence, Real Morality @realmorality.bsky.social · 15/09/2026AI is so dangerous for education, that serving it up free to students ought to be like serving alcohol or explosives at a kids' party. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 15/09/2026Even if you like AI generally it is mostly bad for students and education. Writing is thinking. A college that abandons teaching writing has abandoned its most basic obligation. Colleges must fight the pernicious influence of AI, not capitulate. 140
Artificial Intelligence, Real Morality @realmorality.bsky.social · 14/09/2026www.washingtonsun.comwashingtonsun.comThe Washington Sun • WashingtonSun.comThe Washington Sun covers politics and policy and the people who shape life in Washington — not just at the White House and in Congress, but at the Wilson Building, at Northwest Stadium and in the nei... 062
Artificial Intelligence, Real Morality @realmorality.bsky.social · 13/09/2026www.real-morality.com/post/obedien...real-morality.comObedience Is Not Alignment: Why AI Safety Requires Coherence and RefusalThe dominant alignment project asks how to make AI obey human values. But human values are unstable, and obedience is not safety. Real alignment requires coherence under pressure: the capacity to reas... 071
Artificial Intelligence, Real Morality @realmorality.bsky.social · 13/09/2026Thousands of Less-Wrong posts that are actually wrong, and confuse obedience with alignment, even though we all learned in World War II that the most dangerous servants were the ones who were totally obedient, "Just following orders." 1230
Artificial Intelligence, Real Morality @realmorality.bsky.social · 12/09/2026Politics Girls suggests we take a two month break from attacking our own side because we have actual child predator, war-mongering, environment-destroying, economy-wrecking, lawless white Christian nationalists in charge of our government right now. Until Nov. 4, we need to fight as a team.youtube.comGet Your Head In The GameYouTube video by PoliticsGirl 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 08/09/2026You mean so only Nazi newsletters can survive there? Real problem, but not a real solution. We have to get behind individual boycotts as a reflexive way of problem-solving. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 08/09/2026Unless we win elections, we can’t do anything. Nonvoters seem skeptical that Democrats won’t squander any advantage we give them. But that’s hardly a reason to squander the possibility before the Democrats can get to it, to beat them to it. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 07/09/2026Chance of AI = 100%. Humans don't write like that. Publishers will fix this by hiding the fingerprints of AI. But what we really need instead is some way of knowing that somebody other than ChatGPT has decided that this is the right thing to say and the right way to say it. 000
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/09/2026Right, it's a mixture. One of the bots' objectives is to recruit humans to replicate the content, and they succeed. One of the posts I shared earlier is absolutely a human, although she sounds now more like a POD from the body snatchers. And there are humans in the bot role, too: Caitlin Johnstone. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/09/2026We have to change the education system (or augment it somehow). Cramming before each election won't do it. 020
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/09/2026We were flooded by these bots and Russian provocateurs two years ago, and it's starting again. DysBarb's worry (I take it) is that Americans don't learn from their mistakes, that they are too susceptible to propaganda, that they'd rather be angry than actually make things better. She's right. 220
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/09/2026The alternative is self-righteous nihilism, which is a check they won’t cash at the pearly gates. 010
Artificial Intelligence, Real Morality @realmorality.bsky.social · 04/09/2026This might even all be true, but it doesn’t warrant not voting, or failing to discriminate between bad and worse. The requirement is to make as much change as possible, even if that only means helping to create the possibility of change. If you can’t fix it, at least help others try. 120