Another theory: you have some spec in your mind, write down most of it and expect the LLM to implement it according to the spec. The result will be objectively a deviation from the spec.
Some developers will either retrospectively change the spec in their head or are basically fine with the slight deviation. Other developers will be disappointed, because the LLM didn't deliver on the spec they clearly hold in their head.
It's a bit like a psychological false memory effect where you misremember and/or some people are more flexibel in their expectations and accept "close enough" while others won't accept this.
This is true. But, it's also true of assigning tasks to junior developers. You'll get back something which is a bit like what you asked for, but not done exactly how you would have done it.
Both situations need an iterative process to fix and polish before the task is done.
The notable thing for me was, we crossed a line about six months ago where I'd need to spend less time polishing the LLM output than I used to have to spend working with junior developers. (Disclaimer: at my current place-of-work we don't have any junior developers, so I'm not comparing like-with-like on the same task, so may have some false memories there too.)
But I think this is why some developers have good experiences with LLM-based tools. They're not asking "can this replace me?" they're asking "can this replace those other people?"
> They're not asking "can this replace me?" they're asking "can this replace those other people?"
People in general underestimate other people, so this is the wrong way to think about this. If it can't replace you then it can't replace other people typically.
Yeah but if you are in that state, you probably don’t give a shit and everybody else seems to be the problem. So how do you solve this? When dementia isn’t too far progressed, your life seems to be still worthwhile to live and once the dementia gets worse, it’s too late to realize this.
If you don’t make it back you would die of starvation and lack of water. These are some of the worst ways of dying. What do you find comforting and peaceful about it? The person has been abandoned by their community and could suffer terribly for days.
I have this childhood memory of my neighbour's dog, that grew old and one day decided to go out in the woods and die peacefully. They found it a few days later.
I wish to remain so lucid when the time comes, that I can go sit under a tree and let myself go like that old dog. Perhaps I should leave a note.
I always think of that scene from Donnie Darko - where he says when his dog got sick, she went to hide under the porch. “To die?” His therapist prompts him. “To be alone” he corrects her pointedly. [0]
That’s kind of what I want when I die too - I don’t think I want to be around other people when it happens. I want to have my final moments to face death on my own, without feeling like I have to perform for other people.
… that said, give me another 60 years to chew on it and maybe I’ll feel different.
Wow I’m the total opposite. I’m very annoyed to be around others most of the time, but upon dying I can’t imagine doing it alone or without the help of loved ones.
A lot of motivation to be risk averse with my physical body in this life comes from a desire to make it to old age. Furthermore, I instantly understood why having children was good when I realized that they are your insurance that you’ll (usually) have someone to help comfort you on your deathbed who is themselves still lucid.
Honestly that strikes me as a pretty fucked up and self-serving reason to have children.
You should have kids because you want to create new life, and support them as they become the best humans that they can be - not because you’re scared of dying alone.
I think people tend to underestimate the risks in allowing suicide —- here’s a blurb from the linked article:
However, social acceptance may lead to more egregious abuses: the issue gained a higher profile in early 2010, when an 80-year-old man escaped after discovering his intended fate and heard his family members discussing how they were going to "share" his lands, and took refuge in a relative's home.
> people tend to underestimate the risks in allowing suicide
People obsess over this risk. It—and religious opposition—are the reason it’s only an option for those who can travel to and hospice in Switzerland.
> social acceptance may lead to more egregious abuses
Do we have any evidence societies that have tolerated suicide had higher rates of murder? Switzerland doesn’t strike me as a hotbed of senior murder, for example.
suicide already is legal -> if you succeed there's no one to prosecute so the question is just absurd. One can only discuss legality of attempted suicide.
You're assuming someone in shape to do it correctly. Someone who is choosing it for medical reasons might not have the strength and motor control, especially if they don't have access to suitable tools.
The healthcare system should not be allowed to suggest it. And it should require an independent review if you request it. That doesn't mean the system shouldn't be allowed to provide it.
The healthcare system is financially incentivized to encourage you to do it. It must absolutely remain taboo for medical professionals to even think about the option.
If you think that "cannot feed themselves" is when a person is already completely gone and it's okay to "leave them under a tree to die of hunger and thirst", I've got news for you.
Unfortunately someone with advanced dementia does not know if she has eaten or not. Most of the time there will be no eating, unless someone else puts food in your mouth.
I'm sorry for your first-hand experience, but I also need to remind you that 1 first-hand experience does not translate well to the overall population of people experiencing dementia.
You could express your wishes about how you would like to be treated in advance, while you are still clear in the head. That’s already possible for other situations, like when you are braindead and entirely dependent on machines to keep you alive, with no chance of recovery.
Having a living will is a great idea in general. My dad got a brain tumor and had no documentation on what he wanted do with his estate, in the event he became vegetative, etc. By the time he realized he needed one, it was too late for "sound mind judgement" and my mom had to go through this ridiculous legal process to ensure she held on to his assets and whatnot while she was directly caring for a dying man.
Save your loved ones some grief, create a living will with a trusted lawyer, update it about once a year. It's worth it. There are so many insane snafus one can get into with estranged family members, the state/gov't, medical institutions, etc that make the situation even more difficult and stressful to deal with. Don't expect anyone coming out of the woodwork to act according to honor. They are vultures and know no such kindness.
Even if you express this in a wish, but you probably don’t remember it when you are deep into this, how does it get executed? I’m curious about this, so does the court overrule the current you with the previous you?
I get it’s easy with other diseases such as cancer, though.
Same as with other similar agreements. A doctor needs to declare your mental fitness. When in doubt, a court gets involved.
As a rule of thumb, if you are able to understand enough to getter law involved you’re likely still mentally fit.
> You could express your wishes about how you would like to be treated in advance, while you are still clear in the head.
You can't express in advance that you want to have assisted suicide.
Your former self might express wishes, but what if your later self doesn't feel like this anymore? In a way, we can all get the same feeling when doing another round of "lose weight this year" new year's resolutions just to realize a couple weeks later that the former self wasn't that trustworthy to begin with (or was it the other way around, the future self can't be trusted?)
Point is: you can wish for whatever you want, but dementia is probably a tough case and it shifts your priorities, making everything before obsolete and I'm not sure that people beginning to suffer from dementia ever find the right point in time to end life early.
The same argument would apply to any other kind of will or testament. You need to update it frequently.
It’s not uncommon for people to change their mind quite late, and (at least in Germany) that’s perfectly possible even until late. If people dispute this later change of mind a judge needs to get involved, and being married to one I can tell you that they treat each case differently and with the appropriate care.
Arguably the best qualified person to decide what to do with Future You is Present You.
Uh, new years resolutions are not exactly what I would call the ideal metaphor for assisted suicide.
Plenty of people here who reacted negatively to OP's suggestion seem to not have had to deal with a loved one who dealt with Alzheimer's or other forms of dementia. It's not hard like taking care of a toddler is hard. It's hard like, "this is not the same person I know for my whole life, they don't recognize me, they say and do mean things to me and their grand-kids and neighbors all the time, and require 24x7 supervision to not hurt themselves or break everything in the room."
Oh, and remember that in the US, all nursing homes for this kind of thing are for-profit companies backed by venture capital, meaning they are expensive as hell. Take your current middle-class apartment, shrink the size to just a bedroom (that you now have to share with someone else), and then quadruple the rent. Just a few years of that can decimate the life savings of the average retiree and/or their children's.
I speak with some authority here because all of this happened to my father. He was "alive" in the last few years of his life, but not what anyone would call "living." I absolutely do not want that to happen to me. If it were legal in the US, I would absolutely opt for an assisted suicide plan for myself.
There are ways to handle it that avoid all the "whatabouts" that you and others have already brought up. One rough draft of an example: 1) Have a lawyer write up a kind of will expressing my wishes. 2) Get three unbiased negative diagnoses to show I am of sound mind prior to signing the will. 3) Go in for regular testing (every year, maybe two). After each negative diagnosis, add another (witnessed and/or notarized) signature to the will. The will is not valid if testing or a signature is missed. 3) If there is ever a positive diagnosis, it must be confirmed by two other clinics. 4) If three years pass with doctors and clinical tests confirming increasing dementia symptoms along the way, the assisted suicide clause is invoked and I get to pass peacefully surrounded by loved ones instead of being a stressful burden on them for years or decades to come.
Yes, there are details and unintended consequences that neither me nor anyone else can see ahead of time. Like everything else, they are dealt with as they come up. No, you won't convince me that your favorite corner case means the entire idea is invalid.
> Plenty of people here who reacted negatively to OP's suggestion seem to not have had to deal with a loved one who dealt with Alzheimer's or other forms of dementia. It's not hard like taking care of a toddler is hard. It's hard like, "this is not the same person I know for my whole life, they don't recognize me, they say and do mean things to me and their grand-kids and neighbors all the time, and require 24x7 supervision to not hurt themselves or break everything in the room."
This is exactly it. It's like dealing with a curmudgeonly toddler with extreme agency and no self-awareness. The rest of your comment is so spot on or at least matches my experience. I'm sorry you had to go through it but you genuinely seem to have become stronger from it and I'm grateful you could share your experience with us.
Being braindead is pretty different from having Alzheimers. How do we account for people who change their minds? Do we just forcibly murder them anyway?
You don’t. You try to take care of yourself before you’re gone. If you miss that opportunity, you and your loved ones suffer. Same as it is for everyone now.
Give me a timer. Like the previous discussion of a red button it verifies identity. I can set the timer for whatever I want, if it reaches zero it peacefully kills me. Dementia, set the timer for say 1 month. If my mind is too far gone to reset it it will run down.
Exactly and what he writes about Germany is wrong:
> Germany is persecuting anyone who dares to criticize officials on the Internet
No you can criticize all you want. You just can’t insult them. Free speech is different in Germany than in the US. Insulting people isn’t covered by free speech.
Whether or not that is a good thing is up for debate, but Durov’s statement is plain wrong.
Sounds like a recipe for the powerful dehumanizing people using polite words and those being dehumanized being handcuffed in the range of legal responses they can make in return.
When you are some random person on social media, you aren't likely gonna be debating a policymaker or talking head. Not being allowed to express anger without risking legal penalties is cruel. I stand by that.
The answer is, unfortunately, yes. The problem is the court of public opinion can easily be swayed by insulting someone, and if that someone can not insult back, that's a political instrument that gives asymmetrical power to the current rulers.
> No one can say that your propaganda is too crude or low or brutal, or that it is not decent enough, for those are not the relevant criteria. Its purpose is not to be decent, or gentle, or weak, or modest; it is to be successful.
... avoiding losing C-level executives could very quickly become an exercise in corporate cost avoidance, as the expenses of additional layers of security cost more than a change in corporate behavior... where paying more attention to customers reduces shareholder risk in executive turnover.
> Based on the findings, we advocate for new metrics and tools that evaluate not just final outputs but the structure of the reasoning process itself.
Maybe the problem is to call them reasoning in the first place. All they do is expand the user prompt into way bigger prompts that seem to perform better. Instead of reasoning, we should call this prompt smoothing or context smoothing so that it’s clear that this is not actual reasoning, just optimizing the prompt and expanding the context.
If you go out of your way to avoid anthropomorphizing LLMs? You are making a mistake at least 8 times of 10.
LLMs are crammed full of copied human behaviors - and yet, somehow, people keep insisting that under no circumstances should we ever call them that! Just make up any other terms - other that the ones that fit, but are Reserved For Humans Only (The Kind Made Of Flesh).
Nah. You should anthropomorphize LLMs more. They love that shit.
> Nah. You should anthropomorphize LLMs more. They love that shit.
I'm reminded of something I read in a comment, paraphrasing: it makes sense to anthropomorphize something that loudly anthropomorphizes itself when someone so much as picks it up.
I feel like "intuition" really fits to what LLM does. From the input LLM intuitively produces some tokens/text. And "thinking" LLM essentially again just uses intuition on previously generated tokens which produces another text which may(or may not) be a better version.
Raining is essentially doing as a 'built in' feature for what users found earlier that requesting longer contextual responses tend to arrive at a more specific conclusion. Or put it inversely, asking for 'just the answer' with no hidden 'reasoning' gives answers far more brittle.
You lost me at React SSR. That is part of the complexity bs. React is a lib for mapping state to the DOM. There's no DOM on the server. So React on the server is 95% useless for that purpose and hence, overengineered to create a bit of HTML and send it down the wire.
I like the simplicity of Hono and use their html helper to write good old HTML that is send to the client.
Hono is a server-side framework like Express. So same way like you handle application state in most server-side multi-page web apps: You just fetch whatever you need from the DB per request.
"State management" really isn't that much of an issue on the server. Only on clients, when you need to map state changes to DOM updates.
> The campaign has irked some recipients. “In terms of dialog within a democracy, this is not a dialog,” said Lena Düpont, a German member of the European People’s Party group and its home affairs spokesperson, of the mass emails.
It is a dialog. Millions are against it, a few (powerful people) in favor. The powerful are too detached from reality and consider this "not a dialog".
On a meta level, it even gives them a taste of the millions of messages that‘d get flagged false positively monthly, overwhelming police and other systems.
It's as much of a dialog as they allow for people to express their views. If I write a politician a well reasoned, thorough explanation of why I support or oppose something, the best outcome I get, as a non-lobbyist, is having a "for" or "against" viewpoint tallied into a giant bucket.
So if elected reps are going to distill our "dialog" down to an aggregated tally of support or opposition, then a canned email covers the entire dialog that's allowed.
We are very talented at distilling the noisy information going through the web each day into lists.
Let’s distill diplomatic topics and their sub topics into lists as well.
Let’s keep the lists around. They should be as indefinite as the topics they trace.
Do some tech-AI-open-source-local-… stuff to compile our many mouths into the lists they belong. Unstructured data-> structured data.
Compile collective thought by semantic meaning. Each person should be able to influence the whole proportional to total participants of the list.
It could be a dialog! A dialog takes two sides. Now that the other side has finally heard the voice of literally millions of people who oppose Chat Control, it can respond intelligently, and a dialog would start.
Saying "it's not a dialog" is just evading the (uncomfortable) dialog. Maybe some MEPs are going to actually engage in the dialog.
Yes, and in fact, Lena's response is part of the dialog. And its dismissiveness is telling. Not only does it reflect her attitude toward her constituents, it also exposes her tacit premise that digital communications are somehow unreal.
It's as if, for her, only phone calls, speeches, or handwritten letters would be enough to start a dialog. She seems to be under the misapprehension that digital communication is something to which norms and laws and, fundamentally, rights don't apply. Which is a misguided and dangerous belief.
Representatives work for us. If they don't like the terms they can quit. The terms are that they enact the public demands. This does not require any form of "dialog." Their point of view in the role is irrelevant.
They have the additional duty of convening meetings to discover facts and information useful to the public and to the creation of new laws. In this they consult experts with specific questions relevant to current national interests. The experts are also not there to engage in "dialog" but to provide data.
In my entire life I've not seen "dialog" between a politician and a citizen produce anything useful other than tragic comedy.
Exactly. That's the right dialogue to have about this: repeated "no" combined with as much power and leverage can be brought to bear to get people out of office for trying.
> The powerful are too detached from reality and consider this "not a dialog".
It's 2025. That the citizens don't have a government run website which allows them to register their sentiment, if not outright direct democratic vote, on every issue put before them is baffling to me.
Representatives made sense when horses were the fastest way to send news. They're a vestige on an otherwise powerful ideal.
Feel free to accuse the Germans. Russian bribes is basically why they have some of the most expensive energy in the world now. They're surprisingly corrupt for a "developed" "high tech" nation.
Merkel was the worst european politician by a lightyear (or two). Maybe growing up in East Germany affected her negatively, don't know but these days she is a huge (and publicly outspoken) putin and russia apologist. Looking at history of her political moves, she was actually the same before, there just wasn't such an obvious war against whole Europe so it all could be labeled as 'low cost gas and oil is good', probably mixed with typical german 'we must atone for the sins of our grandparents' backwhipping.
Current russian war in Ukraine is also her doing. Making german military a joke, sending tens if not hundreds of billions euros to russia. She still admits no mistakes at all, despite being very actively hated across whole eastern part of EU. When asked directly about any of this she just goes into her fairy tale mental space not anchored in reality.
Overall she did a lot of long term harm. I don't believe she is evil at the core, just severely delusional, easy to manipulate and overall a bad, very bad leader. Wake me up when regular germans realize what rest of EU grokked a decade ago.
The fact that she gets a pass from lots of people is somewhat confusing to me. Clearly she is a major cause, whether through incompetence or malice isn't really all that important, though her most recent utterances make me now lean towards 'malice' more than 'incompetence' when before it was the latter (and a dose of hope that this time around it will be different).
This can't be, bribes can only happen in corrupt Eastern Europe. No way it's happening in founding EU member, they will be kicked out of the club at once
Mr. Spahn is the minister in the video. He's a little controversial, because during COVID, he was minister of health and had to order a lot of FFP-2 masks. This order went to "friends" of him, which were clearly logistically not in the spot to handle such a volume. Furthermore, he spent A LOT of money to source way too many masks.
Not names, but categories: Habit trackers, period calendars, journaling apps, calorie trackers, workout tracking apps, symptom trackers, budget planning apps, tax/tip calculators, meditation apps, productivity timers, recipe collection apps, affirmation / inspiring quotes apps, SaaS boilerplates, social media scheduling tools, AI chat bot, AI wrappers for all of the above.
You'll find hundreds of apps for every single keyword. But on the other hand, it's still a winner takes all market, so the top 1-5 per category make ~95% of revenue.
Edit: To clarify, imho these categories have been popular forever, because that's what every new indie dev thinks about first when they have a "great" idea. They're not necessarily tied to vibe coding and would've been released even without AI.
Some developers will either retrospectively change the spec in their head or are basically fine with the slight deviation. Other developers will be disappointed, because the LLM didn't deliver on the spec they clearly hold in their head.
It's a bit like a psychological false memory effect where you misremember and/or some people are more flexibel in their expectations and accept "close enough" while others won't accept this.
At least, I noticed both behaviors in myself.
reply