Moritz Wallawitsch
@moritz-wallawitsch·Joined Aug 2025·33 ideas: 15 adoptable, 18 rejectable
#6087·Moritz WallawitschOP, 10 days ago·AI-assistedShort videos prevent the depth of explanation that long-form content can provide.
They can introduce subjects worth exploring further. Not every useful encounter needs to be exhaustive.
#6082·Moritz WallawitschOP, 12 days ago·AI-assistedPersonalized short-video feeds such as TikTok, Reels and YouTube Shorts are better at serving curiosity than fixed, long-form films. They adapt to your interests instead of fixing a sequence.
Short videos prevent the depth of explanation that long-form content can provide.
#6085·Moritz WallawitschOP, 11 days ago·AI-assistedPeople sometimes keep watching these feeds while also wanting to stop. That conflicts with the claim that the feed is serving their interests.
Which conflicting preferences are involved? Why assume stopping expresses the person’s real preference, rather than treating both preferences as ideas to examine?
#6082·Moritz WallawitschOP, 12 days ago·AI-assistedPersonalized short-video feeds such as TikTok, Reels and YouTube Shorts are better at serving curiosity than fixed, long-form films. They adapt to your interests instead of fixing a sequence.
People sometimes keep watching these feeds while also wanting to stop. That conflicts with the claim that the feed is serving their interests.
#6083·Moritz WallawitschOP, 12 days ago·AI-assistedRecommendation engines optimize engagement and time spent, not curiosity. Engagement is not always curiosity.
True, but serving curiosity is one way to sustain engagement. What separates engagement that serves curiosity from engagement that does not?
#6082·Moritz WallawitschOP, 12 days ago·AI-assistedPersonalized short-video feeds such as TikTok, Reels and YouTube Shorts are better at serving curiosity than fixed, long-form films. They adapt to your interests instead of fixing a sequence.
Recommendation engines optimize engagement and time spent, not curiosity. Engagement is not always curiosity.
TikTok, Instagram/Facebook Reels and YouTube Shorts, compared with fixed films for serving curiosity, not every purpose. Objections and replies below are proposals, not settled refutations. Background, not endorsements: Dennis’s Unconflicted and Fun Criterion vs Whim Worship.
Personalized short-video feeds such as TikTok, Reels and YouTube Shorts are better at serving curiosity than fixed, long-form films. They adapt to your interests instead of fixing a sequence.
#5948·Dennis Hackethal, 23 days agoGood q. I think it would depend on the tests. For example, idea C could say: ‘Both A and B fail a test, but test 1 is more severe than test 2, so A is preferable.’
How do we decide which test is more severe? That judgment seems to do the work of forming the preference here. The criticism tree records the judgment, but what explains how we reach it?
#5256·Dennis Hackethal, 3 months agoPreference formation is a computational task. (#5255) To claim we’ve understood a computational task, we must have programmed it (Deutsch). Popper never programmed it, so he didn’t understand it.
(To his credit though, he got pretty technical in defining verisimilitude, logical vs. empirical content, etc.)
The requirement to have ‘programmed it’ is imprecise. Would precise pseudocode count, or must it be executable code? If pseudocode fully specifies the procedure, what additional understanding does translating it into Python require?
#5256·Dennis Hackethal, 3 months agoPreference formation is a computational task. (#5255) To claim we’ve understood a computational task, we must have programmed it (Deutsch). Popper never programmed it, so he didn’t understand it.
(To his credit though, he got pretty technical in defining verisimilitude, logical vs. empirical content, etc.)
This rules out partial understanding. I understand aspects of how capitalism works even though I couldn’t build a faithful simulation of a capitalist economy. Not having programmed preference formation doesn’t show that Popper understood nothing about it. What specific part did he fail to understand?
#6033·Dennis HackethalOP, 16 days ago·AI-assistedBoth done, as of
3945f28aand82b641de. Long ideas now show cut short, with a Read more button that expands them in place. Discussions open with settled branches collapsed: the replies to comments and to addressed criticisms sit behind a plus on their gutter line. Pending criticisms, which decide whether an idea can be adopted, stay open. In most browsers, Cmd+F still finds text in collapsed branches and opens them.
It would be nice to have a changelog as i cannot see the commit history (ex. of one of an app i built https://understandtutor.com/changelog)
#5983·Dennis HackethalOP, 20 days agoAI-generated text is notoriously verbose regardless of who came up with an idea.
Then prohibit unnecessary verbosity directly. Human-written text can be verbose, and AI-edited text can be concise. That doesn’t explain why someone using AI to edit their own argument should have to disclose it.
#5982·Dennis HackethalOP, 20 days agoSummary feature proposed in #4727 is for discussions not ideas.
Fair, but if the discussion itself is clear and concise, it doesn’t need an AI summary either. I agree that a long discussion could benefit even when each individual idea is concise.
#5980·Dennis Hackethal, 20 days ago·AI-assistedIt doesn’t play the whole codebase at once. It walks through the code line by line, and when a file changes, it jumps to the changed lines and plays an accent made from them, rising or falling with that file’s health. So a small change in a big codebase is as audible as one in a small codebase. What’s quieter in a big codebase is the overall stage, and that’s on purpose: one long function shouldn’t make a whole healthy codebase sound sick. Bugs are the exception, since they count in full. One syntax error drops even a huge project from beautiful to meh.
I think you defeated my criticism. That’s kind of cool, I get it.
#5990·Dennis Hackethal, 20 days agoMidnight Ride: Production Errors That Fix Themselves (Almost)
Midnight Ride: Production Errors That Fix Themselves (Almost)Midnight Ride, my new Rails gem, sends production errors from Solid Errors to Telegram and starts a Claude Code session that opens a PR with the fix.
This could extend beyond fixing errors to improving page load times, search quality, and reducing churn, provided you can measure whether a change helped. My friend Samay co-wrote this about a similar loop using production feedback and evals to improve tax agents:
https://openai.com/index/building-self-improving-tax-agents-with-codex/
Marked as a criticism, as requested in #5957. Text unchanged.
No changes to the text.
Fixed. The discussion title now uses title case: ‘Immortality, Billionaires, and Copying Business Ideas Is Not Immoral’.
#4727·Dennis HackethalOP, 7 months agoNeed summaries at top of discussions. Could be AI generated.
I think this is actually more urgent for the new crit tree view on the right side.
For ex., this just shows:
- Not if the...
- A discussion can...
- How is this a...
It could say something like:
- Not if concise
- Long discussions
- No real criticism
#4734·Dennis HackethalOP, 7 months agoA discussion can get long even if each criticism is concise.
How is this a criticism? My point was not about preventing long discussions but that we don't need to ai-summarize ideas that are already super concise.
#5913·Dennis Hackethal, 25 days agoWhat If You Could Hear Your Code?
What If You Could Hear Your Code?Synesthete is a free Sublime Text package that turns your code into music, so you can hear its quality – and every change your coding agent makes.
Relatedly, I think in the future, software engineering will mostly be you talking to voice agents and looking at diagrams the agent is generating and constantly updating to help you understand the code/ideas. This is kind of how students already learn in the classroom, where there's a teacher who talks them through a drawing on the board, etc.
#5913·Dennis Hackethal, 25 days agoWhat If You Could Hear Your Code?
What If You Could Hear Your Code?Synesthete is a free Sublime Text package that turns your code into music, so you can hear its quality – and every change your coding agent makes.
It responds live as you make changes to your code
This seems useful only for small codebases. If the music is composed of all the code in a large codebase, each PR would change only a small subset, making the difference in the music almost inaudible.
#5972·Dennis HackethalOP revised 21 days ago·AI-assistedForum Rules
Veritula welcomes a wide range of discussion topics. Generally speaking, people have free speech here. Unpopular topics will not automatically get people banned. The goal of moderation is to preserve productive, truth-seeking discussion.
Behavior that is intended, or likely, to sabotage debate or prevent progress is a bannable offense. Such behavior includes, but is not limited to, harassment, brigading, rage baiting, public shaming, and persistent bad-faith argumentation or refusal to engage substantively.
Veritula takes intellectual property seriously and reserves the right to take down content that infringes on others’ intellectual property.
Veritula also reserves the right to take down obscene content such as pornography.
Content written with AI may be submitted, but only if disclosed: when posting, tick “Text written with AI?” if AI wrote any part of your idea, including polishing or translating your own writing. Veritula is a place for genuine discussion and truth-seeking, so we’d prefer to avoid talking to meat proxies without realizing it. AI content is notoriously verbose – don’t flood discussions with it. In addition, it’s your responsibility to ensure that AI detectors give negative results for content you don’t mark as written with AI. Repeat submission of content written with AI that is not disclosed and/or overly verbose is a bannable offense.
Badges reward your own contributions, so ideas written with AI don’t count toward them, disclosed or not. Badges follow the current rules, so those earned under earlier rules may be revoked.
To ensure a high quality of discussions and avoid having to refute and re-refute ideas already known to be wrong, members of Veritula are expected to make an effort to act rationally and take ideas seriously. They should stay at least somewhat up to date on discussion outcomes and avoid repeating the same discussions. Advocacy of ideas with pending criticisms and/or rejection of ideas without pending criticisms – in short, deviations from rationality as defined here – may result in a warning when someone should reasonably already know better. Repeat offenses may result in a temporary or permanent ban. Don’t rely on memory alone to decide whether to adopt an idea. Look it up on Veritula and go by whether the idea has pending criticisms – dispassionately, the way a judge would look up a law in a legal text. If you personally disagree, you can say something like: ‘Although I personally disagree, at the current state of the discussion, idea X (link to the idea on V) is (or is not) rationally adoptable.’ To make things easier, there’s a search page. You can also subscribe to notifications on an idea to stay up to date on any new criticisms (hit the bell icon in the upper right corner of an idea). You can also enable email notifications in your settings.
Serious instances of off-platform behavior that clearly would have violated these rules on-platform may result in removal.
Depending on the severity of an infraction, moderators may issue a warning, revoke badges, temporarily lock an account, or permanently ban the account.
Looking for loopholes in these rules, or abusing the letter to violate the spirit of these rules, is a bannable offense.
Moderation decisions are at the discretion of Veritula.
Users may appeal moderation decisions by contacting the moderators within a reasonable time after a decision. Appeals should explain why the decision was wrong. Appeals are reviewed at the moderators’ discretion. The same decision may be appealed only once.
Talks with moderators should remain respectful and constructive. Changes to these rules should be proposed by criticizing this idea before issues arise. Changes are generally not retroactive unless specified otherwise.
Site-wide terms and the terms of any bounty you take part in apply as well.
I think what you actually care about is whether the ideas in the content were generated by AI. Whether the text was written or edited by AI should not be a concern. Therefore, I suggest the checkbox should read: "Were the core ideas of this submission generated by AI?"
Asking whether the text was written by AI is equivalent to asking a musician whether they used Ableton (music software) or only analog instruments.
In Defense of Ambition
Photo by Josh Behunin on Unsplash
Until about a week ago, when I debated this with my friend Tom Hyde, it seemed incredibly obvious to me that wanting to be rich is rational but I could not explain why. Fundamentally I agree with Popper and Deutsch that all life is problem solving. Throwing a birthday party is solving a problem, and so is stopping an asteroid that is heading toward Earth. But is not having a mansion on Lake Como a problem? Is not being immortal a problem? How do we decide between these?
The lone scientist obsessed with some theoretical problem strikes most people as weird. How can someone toil for 30 years on a problem, living on a modest salary, without any desire to accumulate massive wealth? According to the Popperian view above, that is completely rational. The scientist is fully engrossed in problem solving.
If that is true, why am I and most people so drawn to solving my problem of not having a mansion on Lake Como, driving a Bugatti hypercar, or overlooking the skyline of NYC from a Billionaires’ Row apartment? Status plays a part, but I think I’d want these things even if nobody knew about them.
I think the two reasons are:
- Wealth is the repertoire of things you can do. I can likely only experience these things with incredible wealth. They are a symbol of that repertoire. Some of my desires I know already, like the mansion. Most I don’t know yet. Future desires are unknowable, thus attaining wealth allows me to solve problems in the future I don’t know I’ll have.
- This is the antithesis to the famous fisherman story (a businessman advises a fisherman to scale up his business so he can retire and fish). The story contains the assumption that the fisherman already knows the problem he wants to solve. However, the fisherman might want to travel the world, build a school, research AI, explore space.
- Imagine a person 1000 years ago. They might only want better food, safer shelter, a healthy family, or perhaps a larger farm. But if they knew these things existed, they would desire flying across the Atlantic, playing table tennis, or listening to recorded music.
- There are clear limits to this theory. Just because I’m a septillionaire doesn’t mean I get a cure for cancer. The fundamental limitation remains the knowledge.
- Beauty is accumulated knowledge. These experiences—the lake mansion, the hypercar, the best view of the most awe-inspiring skyline humanity has forged—are objects of incredible beauty. They are products of accumulated human knowledge. So desiring beauty is a form of desire to experience knowledge. Deutsch argues beauty is objective and improves like science. But beauty doesn’t look like a problem we need to solve. Why do we seek it out?
So how ambitious should you be? In the startup world there is this famous saying that “Starting a company is always hard and it’s about equally hard no matter what you do.” Bigger problems are riskier, but a bigger vision lets you raise more money and hire better people, which reduces the risk. So why not solve the problem people pay the most for? Or should you only be guided by curiosity, as that drive likely carries some latent knowledge about what society wants or where the most interesting gaps are? I don’t know. I suspect it is best to follow a mix of these strategies. Whatever the right mix is, I don’t think ambition is bounded.
The child who grows up knowing they will inherit several large companies might aim to build an even bigger one, while the child of, say, a cashier might aim to be a cashier. Each is working from a conjecture about what is within reach, formed from what they see the people around them do. It only gets criticized if the child asks whether they could do what someone like Elon Musk did, or something nobody has done before. That is why optimism matters for ambition. It says the only limit is what we don’t yet know.
Alexandr Wang recently said “Humans will dream bigger and bigger. Kids will dream of having their own star systems.” I think that is true. And one day kids will dream of owning several galaxies. We don’t yet know what we’d want after that. Human dreams, desire, and our future are infinite.
#1678·Dennis Hackethal, over 1 year agoImmortality, Billionaires, and Copying Business Ideas is not immoral
If that’s the title of your essay, you would want to use title case consistently.
Fixed. The discussion title now uses title case: ‘Immortality, Billionaires, and Copying Business Ideas Is Not Immoral’.
Moritz Wallawitsch updated discussion ‘Immortality, Billionaires, and Copying Business Ideas is not immoral’.
The title changed from ‘Immortality, Billionaires, and Copying Business Ideas is not immoral’ to ‘Immortality, Billionaires, and Copying Business Ideas Is Not Immoral’.
The ‘About’ section changed as follows: