anbende

The primary issue with this is the way shares and equity work. Elon Musk is a trillionaire on paper. He pays himself famously little.

That net worth is, under our current system, tightly linked with his interest and control in his various enterprises.

He can borrow against it which gets around taxes and that should probably be addressed, but he like the hypothetical fresh billionaire startup founder don’t have that money. And the mega rich on paper can’t access more than a small percentage of that money without reducing their control of the company they built or are building.

Ideas like a wealth tax, or the new sovereign fund paid into by an equities tax or grant are all interesting, but they are all more complicated on the ground than “wouldn’t $100m be motivating enough?”

FWIW, I think that some sort of public endowment and/or sponsored healthcare, education and safety net and/or tax or management of hyper wealth is one of the problems of our age.

But part of that problem is that it’s not clear how to do that in a way that is workable in an increasingly multipolar world of tech and soon healthcare giants that are as powerful as small but growing nation states. Economically and in some cases militarily linked to great powers.

A generous read of this comment might be that you did catch it internally in testing AFTER it shipped but shrugged it off as something you'd patch in the next release in a week or two. Is that what you meant here?

Or that it was caught but didn't surface fully before release?

A helpful governance policy here might be that anything that mutates user content without opt-in consent requires a distinct sign-off or a double sign-off. If the goal is to prevent this from happening in future.

Well, if you personally were building von Neumann machines do any purpose in the solar system, would YOU be interested in sending some to neighboring systems knowing that you could conceivably get a response in your lifetime.

Would you be at all interested in expanding that project to outlast you?

And even if you personally wouldn’t be so inclined, surely you know or have met people who might?

Once you have the self replication, expanding scope may just be additional code…

Hi there. Right now they are going to a separate git repo, yes. Like this:

local-governor/epics/e-epics/e014-clinical-domain-model/runs/run-e014-01-ops-catalog-20260302-173907-244c82

- Attempts

+ Steps

  - Step 1

  - Step 2

  - ...

  - Step 13

job_instance.json

changes_final.patch

run_report.md

improvement_suggestions.md

local-governor is my store for epics, specs, run records, schemas, contracts, etc. No logic, just files. I want all this stuff in a DB, but it's easier to just drop a file path into my spec runner or into a chat window (vscode chat or cli tool), but I'm tinkering with an alt version on a cloud DB that just projects to local files... shrug. I spend about as much time on tooling as actual features :)

Here’s how I do the same thing, just with a slightly different wrapper: I’m running my own stepwise runtime where agents are plugged into defined slots.

I’ll usually work out the big decisions in a chat pane (sometimes a couple panes) until I’ve got a solid foundation: general guidelines, contracts, schemas, and a deterministic spec that’s clear enough to execute without interpretation.

From there, the runtime runs a job. My current code-gen flow looks like this: 1. Sync the current build map + policies into CLAUDE|COPILOT.md 2. Create a fresh feature branch 3. Run an agent in “dangerous mode,” but restricted to that branch (and explicitly no git commands) 4. Run the same agent again—or a different one—another 1–2 times to catch drift, mistakes, or missed edge cases 5. Finish with a run report (a simple model pass over the spec + the patch) and keep all intermediate outputs inspectable

And at the end, I include a final step that says: “Inspect the whole run and suggest improvements to COPILOT.md or the spec runner package.” That recommendation shows up in the report, so the system gets a little better each iteration instead of just producing code.

I keep tweaking the spec format, agent.md instructions and job steps so my velocity improves over time.

--- To answer the original article's question. I keep all the run records including the llm reasoning and output in the run record in a separate store, but it could be in repo also. I just have too many repos and want it all in one place.

I think this one way of looking at what your parent was describing.

They weren’t just saying ‘AI writes the boilerplate for me.’ They were saying: once you’ve written the same glue the 3rd, 4th, 5th time, you can start folding that pattern into your own custom dev tooling.

AI not as a boilerplate writer but as an assistant to build out personal scaffolding toolset quickly and organically. Or maybe you think that should be more systemized and less personal?

I had a very similar experience, watching it at Japanese camp when I was about 8. I had recurring dreams of a ship going down into a poison jungle for many years and vague memories of this movie.

I found it again when mail order Netflix first came out in my early 20s. It was like rediscovering my childhood and completing a character arc.

My (not a lawyer) understanding is "no", because Microsoft is not administering the model (making available the chatbot and history logging), not retaining chats (typically, unless you configure it specifically to do this), and any logs or history are only retained on the customer's servers or tenant.

Accessing information on a customer's server or tenant (I have been assured) would require a court order for the customer directly.

But... as an 365 E5 user with an Azure account using the 4o through Foundry... I am much more nervous than I ever have been.

Yes this is a reasonable way to misunderstand given the way we refer to “hard sciences” and “soft sciences”, but it does not map to the terms “hard scifi” and “soft scifi” in common usage.

It’s not exactly about rules consistency either as stated by the GP, though that’s part of it. It’s more about strong consistent application of scientific principles even theoretical or untested principles.

This is in contrast to futuristic fantasy with no real focus on the science. But futuristic or space fantasy can be very consistent just like magical systems in fantasy can be very consistent. Hard scifi has to be constrained by plausible consistent science and that science is typically a main character in the story, or even THE main character.

You’re getting downvoted because chess is 2D, not 3D. The pieces only move on a flat plane. 3D chess exists and pieces move up and down as well.

Research anecdote here. I’m a psychologist in a different area and a friend who did his PhD in cocaine research with rats told me this.

Addiction is highly dependent on the immediacy of the drug’s effect on the brain. For this reason, people who smoke cocaine (directly or as crack) are highly likely to get addicted. It may be the majority. People who snort it are less likely to get addicted. And there are a lot of casual cocaine users (snorting) who do not develop a long term life altering addiction. I believe he (researcher friend) said it was 10-15% who go on to become addicted. Still substantial and dangerous but much less than smoking.

His research was looking at delayed onset of cocaine in rats after they pushed their cocaine lever. At longer delays more and more rats showed little interest.

This is part of the overall addiction picture. Decoupling drug use behaviors (smoking, snorting, and lever pressing) from noticeable drug onset prevents reinforcement of the behavior and makes addiction less likely, often much less.

This would explain why nicotine patches and gum would potentially be much less addictive than cigarettes.

As mentioned elsewhere in the thread, it's called the metric system because the values were derived from the meter. Decimal time has nothing to do with the meter, and "decimal" refers to it being base 10.

This isn't totally true. Mass and volume measurements were indeed derived from the meter. A gram is a cubic centimeter of water. A liter is 1/1000 of a cubic meter. Apparently Celsius is derived from Kelvin (really just translated so 0 is the freezing point of water), which is derived using metric units in a formula that is a bit beyond me but available here:

https://www.bipm.org/en/si-base-units/kelvin

Anyway, TIL.

If we’re talking about clarity I think there’s some merit to the claim. It has tense markers that English lacks which buys you information in the conjugation about tense, gender and speaker. And unlike other Latin languages you aren’t able to drop the subject and just rely on the verb to convey it which forces clarity one could argue. You get the best of both worlds for clarity though the worst of both worlds for conjugation complexity and overall verboseness.

At least that’s my attempt to defend the GP’s statement.

Yes, I am always worried about prequels for that reason, but I think I enjoyed this one more than the original. They are really very different in scope and story as you said, and the bittersweet quality of the connection between the two books enhances the ending of the prequel in my view.

I thought it was referring to the planet being "deep" in the sense of being in one of the lower zones of thought. And the Fire was the conflict of galaxy-spanning import that was playing out there.

But maybe there was more to it?

By contrast, the meaning of the title of the sequel "a deepness in the sky" is fairly literal once you get to know more about the alien species and their habits.

Just curious, where do you see an AI therapist? Is there a site or service that you like?

Thanks for saying this. It's really heartening to see people with this understanding.

So typically there's a lot of exposure, facilitation, and processing work FIRST. There's not really a route to forgiveness in my opinion until someone is pretty deeply in touch with all the feelings associated with how they were treated. I take a mindful self-compassion approach to experiencing and reprocessing memories and emotions drawing a lot from Paul Gilbert, a self-compassion researcher and advocate in the UK.

Once a lot of the feelings have been processed, including anger and resentment, we often get to a point where the anger doesn't seem to serve much purpose for the person anymore. It's not covering up something they didn't want to feel (guilt, shame, sadness, fear, etc) and it's not "punishing their parents" in a way that many people hold on to. More often than not, the client just spontaneously forgives at that point, and forgiveness looks like seeing the situation as a bad one that everyone involved was stuck in and no longer having a need for the perpetrator to suffer or take responsibility.

So acceptance and compassion then processing and reprocessing with self-compassion. Sometimes a discussion of blame and forgiveness and what they do and don't mean.

It’s simply an example, though a somewhat extreme one, of the problem with the GP’s generalization.

I think it actually relates to the original article. There’s a difference between mere interpretation and what actually happened.

“You should forgive your parents, because one day you’ll be older and see their perspective” collapses “interpretations develop and mature” with “some events are a problem”.

Both can be forgiven and it’s probably a good idea to do so. I think it’s not helpful to generalize in that way.

Also abuse happens a lot. It may not be the majority but it is NOT rare.

I stand by my objection in this case

Oof, as a psychologist that deals a lot with people who were abused and mistreated, this is a pretty big generalization.

Yes, as we age we come to see things from a more experienced perspective and the perspectives of the adults we grew up with and change, but going all the way to "that is why it is important to forgive your parents" is a big big step.

It is typically a good idea to try to get to forgiveness, you're right, but there's a lot of very indefensible behavior out there.

I agree with this point. I hadn't meant to suggest that we should discard information received from others.

But I think it would be crazy to not to differentiate between immediate experience and what we've been told. Not even because immediate experience is always more accurate. Sometimes it is NOT, but it's a different source of information subject to different problems. Often more trustworthy but not always, though the "not always" can be ameliorated a bit by understanding some of the limits of personal experience.

I was really only taking issue with "interpretation is all we have" applying in the original story - that there is a difference between "my interpretation about something I experienced" and "my beliefs about a thing I did not experience".

Yes the author's story changed, but it changed because he found out that he was lied to by the police (or perhaps, if we want to be generous, "unintentionally misled") not because his memory was fallible.

To get back to the point I took issue with, in the story "the facts" mattered an awful lot. It was a lack of access to the facts that caused the problem not "an incorrect interpretation" of what the author experienced. The latter happens all the time, but interpreting our experience differently (e.g., reprocessing a traumatic memory with self-compassion and seeing it as unfortunate and something to learn from) is a different thing than finding out what we were told was a lie. Both change our story, but one is indeed a reinterpretation and the other is a change in belief or knowledge.

I think it's important to separate those two things. I think some want to treat them as the same. I think that can cause problems.

Yes, I agree that “interpretation is all we have” is the point they are trying to make. I also agree that there is an important point there. Memories are often not what we think they are.

However, in the story, the author did NOT have a memory of breaking a woman’s back. He had a memory of getting in an accident. He interpreted it as his fault. He was told that he broke her back. Not his interpretation. It was a belief about events that he wasn’t present to (what happened in the woman’s car and inside the woman’s body) not really any different than anything else we’re told but don’t witness firsthand. It sounds like it may even have been a lie the police told him to scare him.

Your story is different because you did have an actual experience and misinterpreted it’s meaning (man in red suit = real Santa Claus).

I think the article is sending a pretty problematic overall message.

The underlying story is somewhat interesting. The author went through life with a lot of unnecessary guilt and suffering because he was carrying around a false narrative, and it is true that false narratives happen. But the larger conclusion he tries to draw from it seems really problematic:

History is not true. You can change history. The actual factual events are such a small part of the story. Everything else is interpretation.

But this is NOT the moral of the story in my view. The moral is that one can have an erroneous belief about what happened, and THAT can cause a lot of problems. The author even experienced feeling better when he learned what ACTUALLY happened, not when he decided on a new interpretation of events. Because his original belief about the event was NOT an interpretation of events, it was an erroneous belief about what the events were.

And in fact, in the anecdote, learning the "factual events" was everything.

In general, I struggle with the idea that calling something that's just factually wrong "an interpretation". That seems to stretch the word "interpretation" to the point where it stops being useful. If I am convinced that Napoleon was, in fact, a black man, do we really want to call that an "interpretation of events". What events am I interpreting? None, I would argue. I'm just making things up. Just like someone made up that the woman in the story broke her spine. It just never happened. It's not an interpretation of anything.

It’s very clear from the article that they actually value the work and do NOT want to destroy it.

The problem is that they feel obligated to protect it from vandals and thieves.

As mentioned elsewhere, the whole thing may be essentially marketing to find a buyer.

Also the implication that the digital equivalent of adding entire new lines (new products, new features, or new backend infrastructure, I guess?) is just a "click of a mouse" glosses over the large amount of work that goes into software and tech development, especially at the largest scales. Yes, it might be faster than building a rail line in the early 1900s, but the "button click" metaphor seems a little silly.

So I've been using Language Transfer English to Spanish.

The method uses your native language to scaffold the target language using patterns and rules. They constantly discourage you from trying to memorize or remember the word and instead encourage you to use your principles to "find it" or "think your way there".

For example, they teach you that virtually any English word that end in 'ation' will correspond to a Spanish word that sounds virtually the same, so for example:

- salvation = salvacíon

And the Spanish verb will be the Spanish word minus the "acíon" plus "ar". Since this almost always works, you can often use this principle to find a lot of common verbs you need. So if, for example, you want to say "save", think of the english "ation" word that is related to saving, so "salvation", Spanish "salvacíon", which gives us the Spanish verb "salvar".

So with that rule, and some rules for present tense conjugation and word order, you can soon you can say a surprising amount very quickly by "thinking through it" rather than memorization. So they "build up" your Spanish from these sorts of building blocks, and there are a surprisingly large number of them. "Don't remember it, find it" is the guiding mantra. And it really works well. Even as my Spanish advances, learning new principles continues to be helpful and kind of magical.

Change my mind? I hadn't meant to make a value judgement about their progress. How could I not knowing how long they've studied?

My comment was about the current level of their French. It's in the elementary range A1-A2, with some problems that look Duolingo-specific around tenses and conjugation.

Would a year of classroom instruction produce better results? I think it would produce different results. I think they'd know basic present-tense verb conjugation, especially first-person. I think they might also get the passé composé (main tense used for past actions) and "aller-verb" ("going to" verb) construction for future tense (which wasn't relevant to the post).

My main point was about the Duolingo-specific errors. Classroom instruction tries to give grammar and conjugation tools that will allow you to construct French. It does not always do so well or efficiently (Language Transfer is much better, IMO). The French I was attempting to assess was missing a lot of this. The author appeared to have words but be missing a lot of grammar, conjugation and common word order information. Duolingo does not really convey these building blocks, and that's one of the primary criticisms of the method.

The post in French did not mention how long they had been studying. That came in a later comment, so that wasn't part of my thinking. I think you're right that B1 in a year would require more work than a casual 2 classes per week.

So I wasn't giving B1 as a benchmark for any amount of time, because no amount of time was given in the post I responded to. I was just answering the question about the quality of their French, since they had asked that in their post.

I am just about to return to the US after 4 months in Buenos Aires. I lived in Peru for a year and a half leading up to April 2022. I'm probably approaching a B2 level of Spanish right now (based on my experience attaining a B2 and then C1 level of French once upon a time).

Interestingly, when I was starting out, I was in Peru with a Spanish speaking significant other, and using Duolingo as well as taking private lessons. I found as a few others have mentioned that Duolingo actually got in the way. I liked the gamification. It helped me practice, but there were so many exercises for each set of vocabulary. Due to the way their system worked, I had to go through a lot of exercises that weren't useful for me because I had picked up the content with other learning methods. You can test out of things, but if I wasn't ready to test out of the whole block I still had to go through a lot of material that wasn't useful (i.e., even testing out skill by skill, was tedious and time-consuming and counter-productive).

So as discussed in the article and here, Duolingo can't get you even to a low intermediate level, just the basics. And I found that it doesn't pair well with more intensive learning. Maybe other people have had a different experience, but I think it's good for the light hobbyist until they are ready to move on to something more serious, but that's about it.

I also worry that it's promises and gamification trap and frustrate people who could have done more. Maybe it makes up for that by getting a large number of people started who then go on. Maybe.

You asked if your French was good, so here you go. I have some experience with French standardized testing. I had to pass a B2 (upper intermediate) exam for Canadian immigration and a C1 (lower advanced) exam for my professional license in Quebec.

I think this actually highlights a lot of the problems that people have with duolingo. You aren't terrible at French, but you are missing exactly the skills people malign Duolingo for not teaching. There are problems with word order, which looks like a lot of word to word translation or bad Google Translate (e.g., "Est-ce que mon francais bon?" is odd and unnatural). Your present-tense is spotty (e.g., it should be "j'apprends" not "je apprendre"), and your ability to use other tenses correctly appears non-existent (e.g., it should be "j'ai changé" not "j'ai changement" which translates as "I have change").

I don't think you'd pass a B1 exam (lower intermediate) given the above. Maaaybe A2 (upper elementary) for comprehension if your comprehension is better than your production. I suspect you'd have a great deal of difficulty communicating with a native speaker without the help of translation software as well.