Man, I'm sorry for the lack of original thought, but I am so sick of AI websites.
I would seriously rather see a plain unstyled html page with a marquee tag saying "A HUMAN WROTE THE NEXT THREE PARAGRAPHS"
I switch between Pi, Autolith, and my own harness. I come to this website. I don't know what the hell a strand agent is. I don't know who is putting this out. I don't understand whether I'm meant to use a GUI or write code in order to configure it. All the screenshots show a GUI. "use one line of python or javascript to configure." I don't know if this is a harness, an orchestrator, an agent framework, or all of the above. I think it's hard. People get so caught up in making their own thing - they lose track of the state of the art. I have no idea what the value prop for Strands Harness is. But maybe I'm not the target audience.
Also, I don't really care about 30% token efficiency gains. Maybe I should. But this is not my experience. Don't get me wrong, I'm conservative with tokens. But I also have a LOT of convenience / essentials which bloat up context. And I'd rather bloat context and have it work well than not have it work well. And when GLM 5.3 Flash came out - all of a sudden I had more tokens than I knew what to do with. That's when I started using OpenClaw type agents for the first time.
I don't have the tokens/cash to throw at something like Gas Town or OpenRig. And I'm still not at all convinced that those software factories are actually the future of everything. Maybe so. But I just don't see why I would invest the time to learn a piece of software without really understanding the value prop. It's crazy to say, but it's just not enough anymore for software to be free. Because the cost of me implementing my own software is rapidly approaching free. And I like mine better.
"Oh, well I just need something that works out of the box." Wait, why? What does it need to work for? For making software? lol.
Personally, I build a loosely compiled / linted DSL to handle spec.
It's a SEXP language but could just as easily be markdown. It'd probably be better if it was - I spend a lot of tokens matching parentheses.
Compared to markdown - I lose some human readability but gain a lot in workflow.
For me - it turns the sort of "free form jazz code odyssey" into something with structure. It changes the question to "does the spec match my idea? And does the code match my spec?"
Every so often, I do a "code astrology reset" and try to purge all the tricks I use to keep things functional. I absolutely notice an efficiency hit when I don't have a spec language. Even with the new models - I find this essential to keep things on track.
I think most people on large projects hit a scaling ceiling with markdown specs. They get massive and contradictory very quickly. I think a src/md folder is a good strategy. I try to organize with a "one spec per module" rule. It doesn't always happen that way but I find that helpful.
I think Literate Programming is a good inspiration here. I also think Yegge's beads and Gastown have some really smart things to say in this regard - even if he's a bit tokenmaxxy for my taste.
I totally agree with this. Agents seem tremendously bad at UI to me. Maybe it's just because I am a back end guy.
Right now I'm working on a declarative UI framework which can help me along here. My thought is that if I sacrifice a little control for sane primitives, that will make that spec /build loop easier.
I think ClayUI is a really interesting "reduced instruction set" for UI. I don't know that immediate mode UI is the right call for anything web related (that's how you get React lol) but his reduced primitive layer is very interesting to me
> Right now I'm working on a declarative UI framework which can help me along here
After your first para, I was about to suggest exactly that (well, maybe not writing your own). I find that frameworks (both front-end and back-end) constrain the LLM's choices and result in both sensible defaults and improved consistency.
Of course, you will immediately hit the problem all frameworks have: customer requirements that the framework components don't quite meet.
Probably one of those things where it's best to try it yourself, because I think anything that constrains the LLM is an improvement.
Think up a simple CRUD problem (lets say a book database - Book *-* Author, Book 1-* Publisher, Book 1-* Shelf). Write a loose markdown spec, and have the chatbot one-shot it in [language of choice]. Then do the same thing, but tack "build it in Filament/Avo/Django/AdonisJS" on the end.
I'd be interested to see if your results match mine - everything's just more consistent, and smoother, and you start getting freebies like autocomplete. The framework acts as guardrails and the LLM has to make fewer choices, so less of the token budget goes on bikeshedding.
I think it's a good, analogous, contemporary example of how RMS tends to be misunderstood.
In this context, his comments tend to be:
a) Correct, if you read them precisely (which is how RMS writes -- precisely)
b) Horrible, if you read them with the subtext implied in mainstream culture (which is how people read what RMS writes)
That combination results in "with the terrorists" in 2001, "with Epstein" in 2026, and a lot of similar things in between.
Coincidentally, it's highly unlikely Minsky did anything wrong. All the people thrown under the bus by MIT were either the people who acted ethically and came forward or, like Minsky, were dead. All the people who did unethical things didn't come forward, and are in the Epstein files. Not all the MIT folks in the Epstein files did unethical things, but many presumably did....
The MIT transparency report is worth reading in the light of the release. MIT had most of that information about its faculty at the time.
> Coincidentally, it's highly unlikely Minsky did anything wrong.
Virginia Giuffre said that she had sex with Minsky when she was 17 and he was 73. I think no matter how you look at it (even if she was technically above the age of consent in the jurisdiction of Epstein's island), it is still wrong and he absolutely should have known that.
I am a big fan of Stallman's work and his views considering computer science and its role in society, but his views are not precise or correct in this matter.
> but his views are not precise or correct in this matter.
The views he expressed re sexual assault were abhorrent
I have huge respect for RMS. Being this wrong once shows the bad way (we) men of a certain age were raised. The idea that sexual assault must involve violence or the threat of violence is old, out of date, and at one point RMS held it along with society at large
Times change, and as we get older we need to change with them.
In this instance, Stallman was arguing semantics. His long-term criticism [1] is that ‘sexual assault’ is being used to cover a wide range of wrongful actions, where a more nuanced terminology to distinguish between them would be more appropriate. That in itself doesn’t imply he finds any of those actions acceptable, only that he’d like them to have more specific names.
He might not have picked a good time and place to share his terminological views, but I don’t think there’s anything wrong with them per se.
From the article, Stallman said "We can imagine many scenarios, but the most plausible scenario is that she presented herself to him as entirely willing."
That expresses an extreme obtuseness about human behavior and a lack of nuance. A 17 year old is a child.
This expresses an extreme inability to properly communicate subtext, which is common for autistic individuals, or ones autism-adjacent.
Stallman's comments, if read logically like a computer, are generally strictly correct. Stallman's comments, if read inferentially like a human might, often imply abhorrent things.
> Stallman's comments, if read logically like a computer, are generally strictly correct. Stallman's comments, if read inferentially like a human might, often imply abhorrent things.
That's not true. I understand the drive to categorize it as such, but if you look at the context of what he said, which is even mentioned in that article, you will realize that this is not a "definitional" issue or have anything to do with strict precision of terms.
That is a red herring that you are getting stuck on and using to defend him. I don't care about whether or not he thinks they are applying a word that he disagrees on the definition. I will cede that if you wish, because it is besides the point and not what he is actually arguing.
He (like you) is making the claim:
> Coincidentally, it's highly unlikely Minsky did anything wrong.
That is demonstrably an incorrect statemet. We may not a definitive morality that I can point to you and we both agree to the letter, but a 73 year old man should have the wisdom to recognize that sleeping with children is damaging to children even if they appear willing.
I will grant that Minsky's participation was probably not the most damaging thing in her life, but he chose to participate and he harmed her, contrary to what you and Stallman are claiming. It is unfortunate that we have many men (and women) who believe and like Stallman and Minsky. It shows that we need to address this and train ourselves better, so that everybody is on the same page when it comes to how we let children develop in our culture.
I'm not sure how to respond to this, except to point out that in a second comment, you finally conceded this:
> Minsky's participation
Did not happen.
The polite thing would be to apologize.
And this:
> if you look at the context of what he said
Would require you to share context. People have a history of sharing an article about an article about an article about what Stallman said.
When I go back to the source, he does have an extreme case of foot-in-mouth, but I have seen nothing from primary sources which would justify the smear campaign. Usually, the "context" is a conversation about terrorists or whatnot, Stallman making some academic point, and the "context" labelling him as "with the terrorists." Stallman:
- Can't read the room
- Doesn't follow the context
- Speaks precisely
Let me be blunt: You're smearing good people who worked to improve the world. That's not cool. It's also something which does broader harm than you might imagine. It means all but the most thick-skinned won't want to get into positions like Minsky or Stallman.
The Minsky situation was MIT looking for a fall-man.
The Stallman one dates back to Microsoft's engineered astroturf campaign, back when they were anti-FLOSS.
Why the fuck would I ever want to pick a fight with a corporation doing something nasty, if the result would be people like you calling me a pedophile?
The reason I give people a very large amount of benefit-of-the-doubt is because I've seen the effect these engineered smear campaigns have on individuals. I've also seen the amount of crazy that happens when one is famous. If a million people know who you are, and 0.01% do something crazy -- obsessive false accusations, bomb threats, or whatnot -- that's still a hundred people. Britney Spears or Zuckerberg will have security teams, legal teams, and PR teams to handle that. People who, like Stallman, are fundamentally doing something for the public benefit have the same pressure from crazies, astroturf, and whatnot, but without the ability to afford a defence.
It does seem entirely possible (not sure about ‘most plausible’) she would have presented herself as entirely willing. I see no fundamental problem with that statement. (Note that there’s a huge difference between ‘presented herself as entirely willing’ and ‘was entirely willing’, which I should hope you’re keeping in mind.)
Legally, a 17 year old may be a child in many jurisdictions, but that fact alone has little bearing on their appearance and maturity, so how is it relevant to the topic here? Had she been several months older at the time, would that have had any bearing on her interactions with Minsky, given her situation?
The above commenter and others have made the claim that Stallman is just being "precise" and "correct" about a misapplication of what sexual assault means, which is exactly besides the point and a red herring. They have also made the claim that "she would have presented herself as entirely willing". That is also besides the point. A 17 year old is a child, and that should have been enough for Minksy to not have engaged with her like that, and to recognize that Epstein was taking advantage of her (and in fact, trying to take advantage of him by offering her to him). That should have been enough. Making arguments about definitions or that she appeared "willing" are illegitimate attempts to deflect a serious moral failing. That in and of itself is a moral failing of Stallman, compounding Minsky's failing by trying to defend it.
Unless you mean to imply that any sexual acts between a young person and a remarkably older one ought to begin with the former showing their ID to prove they are in fact of age, the fact that she was legally ‘a child’ is irrelevant to the moral failing at hand. (Because it would be absurd for the morality of the act to radically shift based on whether it takes place a couple of months later or earlier.)
I do agree it was a significant moral failing (assuming he did go through with it), even if she did indeed present herself as willing and appeared of age. I don’t agree that trying to differentiate it from other related moral failings (even if it’s in a better light) constitutes a moral failing in itself, though. It is beside the point and not helpful to the matter at hand, but I don’t see much harm in it.
> Because it would be absurd for the morality of the act to radically shift based on whether it takes place a couple of months later or earlier.
The morality of this is that you should not take advantage of and harm a vulnerable person. In this case, being a child makes them deeply vulnerable. It doesn't matter if it is "a couple of months later or earlier".
Yes, that’s exactly my point. Law usually draws a (somewhat arbitrary) line between child and adult, but what really happens is that people mature gradually, so while the practical difference between a 10-year-old and a 30-year-old is great, the practical difference between a 17-year-old and an 18-year-old is minuscule.
> Virginia Giuffre said that she had sex with Minsky when she was 17 and he was 73. I think no matter how you look at it (even if she was technically above the age of consent in the jurisdiction of Epstein's island), it is still wrong and he absolutely should have known that.
This statement is factually false.
She said she was asked to offer herself for sex to Minsky. People inferred from this that sex happened, but Virginia did not say this. Read the full file.
Accounts from Minsky's wife say that Minsky declined. Accounts from Minsky's friends say he was very monogamous. No one close to the situation supports an account of Minsky having sex with Virginia Giuffre.
The natural next step should not have been to indict Minsky's reputation, but to ask Virginia Giuffre the natural follow-up: How did Minsky react when you made the offer?
In the deposition in Giuffre v Maxwell, she says she was directed by Maxwell to have sex with Prince Andrew, Jean Luc Brunel, Bill Richardson, Marvin Minsky and others. In those remarks, she is usually asked if she was sent or directed, presumably because they were concerned with Maxwell's involvement, and less concerned with the actual acts.
> Accounts from Minsky's wife say that Minsky declined.
That statement is factually false, unless you have a source that is not publicly available. The only accounts I can find is that his widow said it couldn't have happened because she said she was always with him, not that he declined.
Everything you wrote is materially correct. Given that you have this background, can you please explain how this:
> Virginia Giuffre said that she had sex with Minsky when she was 17 and he was 73.
Was not a lie?
Also: Yes, I do have sources which are not publicly available, but the difference between "declined" versus "it couldn't have happened because she said she was always with him" is within the spectrum of ambiguity of real-world human speech and memory. It is not materially different. The material point is there are multiple witnesses (at least the widow, in public sources) that say he did not have sex.
If you want proof of MIT folks covering up:
a) Read the MIT Transparency Report
b) Look at photos on archive.org from people who went to Epstein island and events whom MIT did not reveal. All this was public at the time.
Compare. But MIT also has travel records, so most of the people in the Epstein files, MIT knew about and chose to omit.
It is materially different. Either it did not happen as the widow claimed, or it did and it was declined.
I will back down from my claim that Minsky did have sex with Virginia Giuffree. In the deposition in the case against Maxwell, the questions were if Maxwell directed her to have sex with Prince Andrew, Jean Luc Brunel, Bill Richardson, Marvin Minsky and others. She was not asked if she actually followed through.
If you have non-public information, then feel free to share, otherwise there is nothing left to say.
> I will back down from my claim that Minsky did have sex with Virginia Giuffree
Thank you for backing down.
An apology would be warranted.
> It is materially different. Either it did not happen as the widow claimed, or it did and it was declined.
The material fact was that sex did not happen, regardless of whether the argument was that people saw Minsky decline or the argument was that the widow was with him the whole time and saw it did not happen. The two are not contradictory.
> If you have non-public information, then feel free to share, otherwise there is nothing left to say.
I do. I never met Minsky, which was odd given we revolve in similar circles. I did and still do know a lot of people who do know Minsky, whom I trust. The accounts -- which for me are second-hand and for you are now third-hand, were that the widow wasn't the only witness that sex did not happen, and that people who know Minsky well found the concept of him cheating on his wife inconceivable. That's just not his personality.
That's good enough for me to conclude that Minsky did not have sex with Virginia. It's not hard proof, but it seems extremely unlikely.
However, even if it did seem plausible (let's omit the insider information from a stranger on the internet), I would STILL not indict Minsky without clear evidence. Treating people as guilty without proof is kind of evil, in my opinion. It's two kind of evil:
- It makes smear campaigns work. The reason corporations, the CIA, and other organizations are able to run character assassination on activists and whistleblowers is precisely people like you, willing to run with allegations.
As a footnote: If you actually cared about pedophiles, corruption, and sex, why haven't you followed up on the obvious? The logical question -- if you bothered with logical inference over outrage -- would be why there were multiple, non-public witnesses who weren't mentioned in the MIT Transparency Report. Epstein got around a lot in MIT circles.
You've made several dubious claims. I'm honest and humble enough to be self-critical and not claim beyond what the deposition said, but that doesn't mean I should be held to a higher standard than you (i.e., if you did what you are demanding of me, you would have already apologized several times).
You are claiming there is no difference in the defense by the widow (who said she was always with him), while also claiming that Minsky was offered but declined. Both can only be the true if the offer was discreetly made and that the widow did not notice. The consequence is that it diminishes her as a reliable witness and also raises the question if the offer was discreet, couldn't it have been just as discreetly accepted?
If we accept your insider knowledge, it does not improve the circumstances. Not just because friends and colleagues would want to defend someone they are involved with (and indirectly, themselves), but also because the issue here is not just whether he participated, but why didn't he respond appropriately?
He was an intellectual leader. Shouldn't we demand our leaders to be moral? In a slightly better world, he would have had the moral clarity of Norm Finkelstein and rejected any association with Epstein once he was offered sex by a 17 year old under his care. It does not appear that Minsky did that, likely because he did not want to risk the grant money and perks of hob-knobbing with the wealthy, or risk reputational harm that Epstein could have inflicted to discredit him.
I will share a more ambiguous situation to make it clearer why Minsky's response was problematic. I had a friend at a Zen center when two young women came to visit. One of them was wealthy, and the other was nominally her friend but behaved like a servant. The wealthy one even directed her to have sex with some of the male students and other visitors. It was disturbing to my friend, and at the time, he was not sure how to respond. One could claim that the young woman who was being directed was benefitting from the relationship, and that, after all, she was an adult.
However it is clear that it is toxic behavior, even if the young woman's victimhood might seem ambiguous to some. I would claim that the behavior should be rebuked and that morality should not just be about protecting the victims you see but also about rebuking malevolence, because there are so many victims that are not seen. The wealthy young woman was exhibiting her character, and that character would likely persist whether it was in a relationship with her friend, children, or other vulnerable people. She should be plainly told that it is not ok.
Although I think some of your assessments are misguided, I think we have common ground about scape-goating (which you were referencing in the second half of your response, but didn't name). If Minsky had rebuked Epstein, or if my friend had rebuked that young woman, it is likely that both would have been scape-goated. Malicious people do not bind themselves to morality so they can lie, instigate and conspire to undermine the reputation of whistle-blowers and other truth speakers. This is a persistent problem of humanity that was exemplified by the stories of Socrates and Jesus, and all of free-thinkers killed during the European medieval period, Nazi Germany and the Soviet Union. It is a problem even in much more mundane settings, and will occur whenever a moral person confronts immoral leadership. That is why we have to trumpet morality, lend support to moral behavior and speak out against abuse.
I'm not advocating for Minsky to be condemned as a terrible person. I'm just pointing out that he, as an example of wide-spread societal behavior, did not do the right thing.
> You are claiming there is no difference in the defense by the widow (who said she was always with him), while also claiming that Minsky was offered but declined. Both can only be the true if the offer was discreetly made and that the widow did not notice.
This doesn't seem like rocket science. If I am e.g. at a dinner party with my spouse, I might be aware there was no opportunity to have sex while not monitoring every conversation my spouse is involved with.
Conversely, if someone flirts with my spouse, and my spouse doesn't engage, I might not take notice, or if I do, I might not find it important enough to report back. There's a broad space of possibilities. "Offered," in this context, is pretty broad.
Famous people have such "offers" all the time. I was once mildly famous, and in one place I visited, there was a person who clearly expressed interest, who looked perhaps half my age. I signed an autograph and quickly moved on. This is the first time I even thought to even relay the story.
> Not just because friends and colleagues would want to defend someone they are involved with...
... you have paranoid delusions. That's just not how the world works.
> If Minsky had rebuked Epstein
I have no evidence this did not happen, and neither do you. Most times, such rebukes are private. I also have no idea what, if anything, Minsky knew.
> rejected any association
> I'm not advocating for Minsky to be condemned as a terrible person. I'm just pointing out that he, as an example of wide-spread societal behavior, did not do the right thing.
You have no idea what he did. You picked a random person, and decided to vilify him in your brain, based on fragmentary quotes.
I'll also mention: Good people associate with bad people. Christians often cite how Jesus associated with tax collectors, sex workers, and other sinners. As a tactic, isolation does not work. Overcoming racism means associating with racists. Overcoming sexism means associating with sexists. "Talking with" is not the same as "supporting," and if we didn't have cancel culture or "cut diplomatic relations" as a punishment, the world would be a much better place.
I can't edit my last comment but want to address the rest of your response even if it has become clear that you are not discussing this in good faith.
> > Not just because friends and colleagues would want to defend someone they are involved with...
> ... you have paranoid delusions. That's just not how the world works.
This is absolutely how the world works. Notice I did not say only. There usually will be a cohort that will immediately try to distance themselves, but also a cohort that will defend. The ratio will be different in different circumstances. You are denying the defending cohort while you are clearly part of it (if we believe your prior statements about your closeness to the situation). That is absurd.
What is further absurd is that you are comparing Minksy and Epstein to Jesus's ministry amongst the poor and marginalized. I am not sure how to respond to that. I did not know Minksy but I can assure you he was much more likely thinking about his grant money and protecting his own reputation (and MIT's) than being Christ-like to Epstein.
What you are talking about is enabling: ignoring morality. Once again, I'll state that that is not ok, and admonish Minksy for his own actions, as well as understanding that why we do this as a society is also to remind each other of what should be ok and what should not be ok. You seem to want to say that Minksy's association with Epstein was ok. I disagree. Let's leave it at that.
> I will back down from my claim that Minsky did have sex with Virginia Giuffree
And then continue to argue basing your arguments on that "fact." You have some kind of vendetta. It's not grounded in evidence, and you're willing to contort facts around that vendetta.
And you write things like:
> if you did what you are demanding of me, you would have already apologized several times
What I'm asking you to do is to not spread baseless accusations, especially about someone who can't defend themselves because they passed away.
The people you should apologize to aren't myself, but the people whom you slandered.
The topic seems to be warnings and predictions coming true even if they seem far-fetched at first. The linked article doesn’t talk about any warnings or predictions. It only recounts (in a way) how RMS, while discussing the topic of Epstein’s abuse, said some things which many people found offensive, leading to his resignation.
I guess it is marginally relevant; it just seems odd to drop a marginally relevant link unaccompanied by further comment. Usually people seem to mention what brought it to mind, or which part they found particularly interesting.
I recently built a calculator to try to quantify the gains an org could expect as they become more "AI-native." I tried to account for a bunch of things including env, which I call "AI roadway," but I'm definitely missing codebase readiness (thank you, author), and also bottleneck analysis. Even without those considerations, though, gains are usually modest. It's hard to get to 2x.
I did a similar thing. Yours seems cleaner than mine. I have a 'compiler' and a sexp based DSL. I don't know how anybody vibecodes medium or large programs, say above 50,000 lines of code. I don't read all the spec until I sense something is out of alignment.
Ive wondered if I have too much complexity, and from time to time I do a "prompt astrology reset" where I get rid of all the extra cruft. I can't go without the spec sheets though.
On removing cruft: one of the key features of spekk is an "observer" agent role that is tasked with finding drift. On a production codebase, I run this daily in a sandbox. It pulls the latest changes and looks for specs that are mismatched from the implementation, preferring to look at specs and code that changed recently and prioritizing "major" drift events. The observer agent then opens a PR wkth its observations (markdown with YAML like the specs). It also posts a summary to Slack, but that's optional. The sandbox agent code is part of the spekk-cli codebase.
It's funny to me that instead of moving existing tech to a word in word out paradigm, people are dropping millions getting the chat to work with the existing paradigm.
Probably inevitable, but seems like a lot of disruption could happen there.
I mean, that's the first adoption phase where you don't start from scratch but keep using the old tech with the bolted on LLMs. It would take some time for devs to build the primitives around the new tech.
https://pricklypear.rocks/
reply