Shadow C | 🍔Arbiter🍕 🌭of🥪 🌮Sandwiches🌯

@Just_a_nut69 @silver21096 @OrwellNGoode I'd argue that there is merit in the "life already in danger" angle as one needs to take an extra action to keep it. For example, stepping off the road, moving to shelter in storm, etc. A true "not in danger" means you should not have to do anything extra to keep your life.

It doesnt. OG question is simply, everyone on earth must press the button. Nobody says there is a danger to ur life. So trolley problem above is valid. The "danger" in this question is people who are not capable of understanding the difference for whatever reason that will press blue at random.

@Anonymous158746 @YourBudTevin In other words, I agree that your logic is valid given subjective value. The issue wasn't in the logic, itself, but in the completeness of the premise and consideration of all facts. Your logic works for the set of people that you consider, but does not for those that weren't.

@Anonymous158746 @YourBudTevin People who would otherwise be willing to help themselves if they could, but couldn't for reasons beyond their control (i.e. immature toddler mind, etc). It's not possible to categorise all blue pressers as a single category of those who don't *want* to help themselves. 2/2

I’ve seen like 800 tweets about this but my simple way of thinking is: If you choose blue, you don’t want anyone to die. It is impossible for anyone to die if you press blue. If you choose red, you are afraid of others choosing red. The only way anyone dies at all is if you choose red. When I read this, I actually don’t think about myself. I think about hurting the least amount of people. So I pick blue.

@UndoCopyPaste @Anonymous158746 @YourBudTevin You should let the question setter know that then. x.com/waitbutwhy/sta…

@ShadowC5 @Anonymous158746 @YourBudTevin You bring babies into it because you can't face the reality that no one is forced to press blue. It's purely performative.

@Anonymous158746 @YourBudTevin That's a subjective value judgement. Some people are willing to risk their lives to save babies who unfortunately put themselves in dangerous situations - e.g. pool, traffic, fire, etc. Such is the same here, for babies who choose randomly.

@YourBudTevin Wrong. Your survival is worth more than saving people who willingly chose to put thenselves in danger. It is far better to help those who help themselves. That isnt being selfish, or at least no more than to put yourself at risk and demand someone else do the same for you.

@Skolander2 @SeanCFoley12 @ChrisPacia Exactly

@ShadowC5 @SeanCFoley12 @ChrisPacia People being catastrophically wrong is nothing new. Flat earth, miasma theory, witch burnings, slavery, plato's theory of forms, all flavors of anti-realism, you name it. Any position you can think of, someone at some point in history, held. It tells you nothing of its merits.

The pattern played out like this: - Duh Red is the obvious choice. Nobody dies if everyone pushes red. There's not reason to push blue. - Blue button pusher think they have a bullet proof rebuttal... Some confused or irrational people will push blue. Therefore the knee-jerk red reaction is naive and the true enlightened position is that that you must push blue. Unfortunately, that turns out to be the naive position. Consider the following...

@SeanCFoley12 @Skolander2 @ChrisPacia Sure? Ethical systems are hotly debated. I'm just noting that it would be strange to speak of it like it is settled, such that it could expose anything in any crystal clear manner - it does not. Just that you have irreconcilable difference on what ethics is, with many others.

@ShadowC5 @Skolander2 @ChrisPacia For me that is an affront to my own ethics, which are based on solid principles. But hey, people can propose whatever ethical systems they like.

@SeanCFoley12 @Skolander2 @ChrisPacia Isn't this moreso revealing the dissatisfaction with kantian virtue ethics rather than any political thinking? There are other ethical systems, such as various flavors of utilitarianism, that happily accept that ethical/unethical can shift based on quantitative thresholds.

Yes, once you change the percentage threshold the ethics become crystal clear. You can even go as far to say, what if the threshold were that a single person choosing red kills all blue. Then everyone would agree with red. But the principle is no different if you are at risk from 1 person choosing red vs 51% choosing red. There is literally no difference. The risk you take does not suddenly become ethical if it's small enough. Nobody really even knows what the risk actually is. This exposes leftist thinking. It's a lot like communism, which is the belief that if everyone has the same communal approach as me, society as a whole will work much better. But what if people don't have the same approach as me? Well, you gotta force them. Can't allow any red-button pushers. Leftists say you must remove the red button entirely.

@IamthejuanG @Siddharth87 @micsolana If not Tim Urban, then who? Just curious.

@Siddharth87 @micsolana "Tim urban"didnt create the idea including babies is a bastardization of the original concept that completely mises the point

interesting how completely the conversation has shifted to kids. initially, most people assumed we were talking about adults, not babies smashing buttons, which kind of breaks the thought experiment. that this is all we’re now discussing implies blues understand they were wrong.

@Michael_Druggan Well.. as a fan of education...

There's a purple button and a yellow button. If you press the purple button nothing happens. If you press the yellow button everyone who smugly said "it's a prisoner'a dilemma, retard learn some game theory" during the red button/blue button discourse spends an hour in a magic room where subjective time is dilated 10,000:1 and a recording plays on repeat "the payoff matrix matters, it's only a prisoners dilemma if defect/defect is the unique nash equilibrium" on repeat after which they return home safely hopefully having learned their lesson Which button do you press?

@bbishdotdev @johnpalmer Even as a rational adult, I would have the "hope" that by picking blue, my kids will survive by either picking red randomly, or by my meager contribution to the blue pool in hopes of crossing 50%, that there is some hope they also survive if they randomly pick blue. 2/2

@bbishdotdev @johnpalmer I agree that blue is a "hope", but I think there is much more to explore along this angle. The poll isn't waiting for everyone to grow up into adults with firm rational minds. If the poll runs while my hypothetical kids are still age 3 and 1, and might like the color blue.. 1/x

- If all humans are rational, everyone picks red. - But we know many humans are not rational - Therefore, red means 100% survival, but greater chance non-rational humans die - Blue increases chance all humans live, including non-rational ones So it just comes down to if you think it’s good for non-rational humans to exist.

@Chirag_1313 @theo If you are not part of the poll, would you hope the other country presses blue because 50% is "easy enough" for 100% survival, as compared to requiring 100% to ensure 100% survival?

@Chirag_1313 @theo The poll can be reframed like this: if >50% press blue, all survive. If =100% press red, all survive. If you take yourself out of the poll - say - you are watching some other country do this poll. What choice would you rather they make?

If >50% of people press the blue button, everyone survives Red button pressers always survive, but they’ll get a “red button presser” badge on their Twitter profile. What do you press?

@Pierozi @theo Output becomes input on the next turn. The main difference between cache on/off is whether you pay cache read+write or repeated input for the conversation so far.

@theo How come output are not increasing if cache read is down? There is something wrong from Anthropic or I'm missing something?

Here's our token utilization for those curious

@MilksandMatcha @cerebras I want it because I am in beeg doubt

Giving away 5 Windsurf Max ($200/month) plans Each person will get 3 months of free Windsurf Max (highest tier). Try out SWE 1.6, Cognition's latest, fastest, and most intelligent model, powered by @cerebras. Winners will be selected from comments in 48 hours, comment below why you want it.

@thenoblesimian @_Sagiquarius_ @atvictor @theo No, that is not what Codex is. Harbor is the evaluator. Codex is the evaluated. Scroll up. Read through harbor's code. Harbor calls codex cli to evaluate codex: #L31" target="_blank" rel="nofollow noopener">github.com/harbor-framewo… You have confused the examiner for the examinee.

Not backtracking. "Harbor is a framework from the creators of Terminal-Bench for evaluating and optimizing agents and language models." If that's what Simple Codex is, it's not an accurate representation of Codex CLI. If you go by the numbers on Opus 4.6 in Claude Code CLI and GPT-5.2 (where is 5.4? 👀) in Codex CLI, there is a ~3 percentage point spread with Codex CLI inching out Claude Code. But it's also noted on Terminal Bench that there is a +/- deviation of 3 percentage points on each of those. Both Codex CLI and Claude Code CLI are far from the top. I'm assuming @theo was using Opus 4.6 in OpenClaw because it's a better model for being a personal assistant. GPT-5.4 is cold... aloof. Bottom line: I wish I was able to use the thing I pay $200 a month for in whatever I want and however I want. Appreciate that OpenAI does that and that their harness is open-source. So much so that I pay another $200 a month for them as well. But it's not something that subscribers should automatically assume is allowed because the other guy allows it. They're businesses and Anthropic is winning on running a business right now. OpenAI is winning on public relations.

Can't stop thinking about how Claude Code is in LAST PLACE on TerminalBench for harnesses using Opus 4.6. There are TEN separate harnesses that use Opus better than Claude Code

@BullStern @Gregorein I actually found the entire thread amazingly productive. *Somebody* needs to audit LLM work for common failure points with a human eye at some point. He did it. Now I can add them as post validation when I give LLM work to do.

@Gregorein Like what was the point of this? To just shade the guy? Doesn't seem like a productive use of your time as you seem intelligent (not meant as sarcasm.)

so... I audited Garry's website after he bragged about 37K LOC/day and a 72-day shipping streak. here's what 78,400 lines of AI slop code actually looks like in production. a single homepage load of garryslist.org downloads 6.42 MB across 169 requests. for a newsletter-blog-thingy. 1/9🧵

@SealOcybe @SnooziUzi @theactualrob @TS9gbn @conquistadorof1 @TheRabbitHole The main idea here is that the same line must cut both ways. If failed economy is often implicated as a consequence of communism, then that same rule also cuts capitalism above. If we want to apply a rule to demarcate capitalism from its corruption, then same rule on communism.

@SealOcybe @SnooziUzi @theactualrob @TS9gbn @conquistadorof1 @TheRabbitHole No, I was arguing that the use of the term is almost always w.r.t. long term economic systems - think along the line of socialism/communism tending towards authoritarianism/failed economy. Yes, the line matters to say what is what in both cases. But not in most applications.

Capitalism creates so there’s more for everyone. Socialism is the weaponization of greed and envy making everyone worse off.