Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Nobody asked for it???

Nobody wants it???

You're not paying attention.

I want it. I think it's our only chance at quickly solving huge complex problems like aging, cancer, Alzheimer's.



It's our only chance at quickly solving huge complex problems like how to cut expensive labor costs.


That too. If all/most labor is automated, humans will be free from the 9 to 5 grind. Products will be cheap or free.


That's never going to happen... when has that ever happened?


When most labor is automated, population will not have a source of income. That's the only way that I see.


When auto manufacturing jobs got shifted overseas leaving hundreds of thousands of people without a source of income did the government come in and provide for them? No, of course not.


Aging isn’t a problem. It’s normal.


That which is normal is not necessarily also good or not problematic. Racism is normal, but its a problem. War is normal, but its absurdly wasteful and problematic. Cancer is normal etc.


> Racism is normal

No, it is not. You ever observe small children in a multicultural group? They’re not racist, because they have no concept of it, nor why you’d want to do that. They see each other as peers. Racism is taught.


Small children typically ignore race, but there does seem to be (at the very least) a cultural tendency towards racism as people move into adulthood. There is a good book about it called "Why are all the black kids sitting together in the cafeteria?" https://www.amazon.com/Black-Kids-Sitting-Together-Cafeteria...

People's characters need to be formed, I would at least argue, to overcome a normal tendency to be uncomfortable with difference.


Just because a certain behaviour doesn't show up in a certain stage of life, doesn't mean it isn't an inherent behaviour. Babies don't talk. Talking is still an inherent behaviour of humans.

Young children's social environment is almost entirely defined by their parents. You'd expect very little aggressive behaviour there.


Animals eat their disabled children. Humans should also do it then, since it's normal or natural.

Around 100k people die from age-related disease every day. I'd be careful to dismiss that as not a problem.


And them not dying would be an even bigger problem on pretty much every single metric I can think of


Yes, people grow old, but I don’t think there’s a rational fear people will ever live forever. However, there are diseases and conditions that are age-related that if solved would improve people’s quality of life, especially the last decade or two of their lives.


I believe it would be a far greater problem if those people didn't die. Aging populations are a huge problem around the globe and unless we'd improve the quality of life to such a margin that octo- and nonagenarians are able to care for each other, I think we're all better off with people dying of old age.


It's a problem for me.

When you're 60 or 70, you will wish to have a body of a 25 year old.

Dying early from rotting teeth is normal in nature, yet you probably brush, floss, and go to dentist.


I don't know how to reconcile this point of view, which I understand, with quotes like "science advances one funeral at a time," which I also understand. I understand the individual desire for advancing anti-aging techniques, but it also seems to contain extremely real risks of stagnating our sciences & societies. I can't tell whether anti-aging science is ethical or not.


That's just a desire for a legal way to murder people (by waiting). It's entirely understandable, but we'll have to find another way.


> we'll have to find another way

You agree it is a problem, then. What are some solutions you can think of? I'm not convinced the gain is worth the pain, but maybe you have some ideas for me to think about.


I really don't know, unfortunately. It seems important that we go beyond merely maintaining people in a state of being not dead, and enable them to feel young as well. This may cure a certain entrenched resentfulness. It will not, however, prevent people from being asshats or provide a mechanism by which they will eventually go away. Perhaps the concept of limited terms in power would be made more general to fields outside politics.


Yeah I don't know either. Thanks to modern medicine, we already have enormous and growing problems with the accumulation of power and wealth in individuals. Currently, death is the only (legal/ethical) check we have on the people who have accumulated so much power that they are above the law. If we remove natural death, how can we ever get out from under the thumb of the uber-powerful? I'd really like to see more thought put into solving these problems before I could be convinced that longevity research is ethical.


It's only ethical because we have no power to prevent their death. If research could be done that looked likely to extend lives, but we chose not to because we'd rather people die earlier, that's equivalent to stealthily killing them. We might as well be upfront about it, and have them slain for being too old, except then they'd tend to resist.


Right, so we have two competing ethical problems: extending life is good, but also, it removes the one (ethical) tool we have to handle the problem of accumulation of power in individuals. This is the conflict I'm grappling with. I don't have a good answer to the problem.


You know what's stagnating? Being dead forever.


That puts you way deep in the wishful thinking territory, again there is 0 rational connection between LLMs and immortality, especially if you're already on the way out


That's why I said "chance".

And no, there's no zero rational connection. LLMs and other types of neural nets are already demonstrating they can solve complex tasks. Unless there's some intrinsic barrier to how intelligent they can become, we should surpass human level intelligence within years, maybe a decade.

Once you surpass human level intelligence, you can scale scientific research by orders of magnitude.

Heck, AI already assists scientific research. We basically solved protein folding problem with AI.


How many intelligent are they?


We define normal by what is common and familiar. But that has no bearing on what ought to be.

It's normal for half of all babies to die in the first year of their life. It's normal for an infected limb to commonly lead to death. Etc.


Dying because I cannot function adequately due to poor vision is normal.

Wearing crafted pieces of glass on my head every waking moment is not normal, but I choose to do it.


It can be both


It sounds a lot like you want to cure life. We're mortal meat bags, go enjoy life instead of worrying about "solving" the inevitable.


Women and people of color also used to be treated like property, doesn't mean it has to be like that forever.

What's cool with humans is that we're able to reflect on our own being and our societies, and when we group together, we can enforce large-scale changes that improve the lives of millions.

We might be mortal meat bags right now, but maybe it isn't the only way, for better or worse.


Cure death, not life.

And I do enjoy life, that's exactly why I want it to be much much longer.


Cool, I like sci fi too. Now explain me how many dots I have to create out of thin air to connect LLMs to immortality ?


Increased intelligence -> automation and scaling of scientific research -> solving complex problems.

The chain is rather short. You don't even need immortality right away. Longevity escape velocity.


> I think it's our only chance at quickly solving huge complex problems like aging, cancer, Alzheimer's.

Machine learning is certainly a extremely handy in tacking these issues, thinking of breakthroughs like AlphaFold and similar. However, I'd like to push against your take:

1. There already are tremendous developments happening in those fronts you mentioned. Praising AI as "our only chance" is quite a stretch and possibly even a harmful statement, considering how severely under-funded these research projects are.

2. It seems to me the poem is more about LLMs/glorified chat bots than general machine learning. In that context, I wouldn't consider them as super useful in Alzheimer's research, certainly not "our only chance".


1. You need to read carefully. I didn't say it was our only chance. I said it was our only chance to do it quickly, I even emphasized the word.

I'm sure humanity can grind through these hard problems in hundreds or thousands of years without AIs.

But I would prefer to be alive, so there's urgency.


Let's hope solving aging is beyond the grasp of even the smartest AI. No surer way to end up in a world full of calcified ideas than having lot of calcified humans clinging to power and enforcing their ways on younger generations.

Not to mention the environmental impact of everlasting humans.


You don't have to participate, even if it gets solved. Feel free to die out.

The idea is that people can't take new ideas in as they age makes zero sense to me.


> The idea is that people can't take new ideas in as they age makes zero sense to me.

Hmm. My counter to this would be that I agree it isn't impossible, but would take such an enormous amount of voluntary effort that no one would spend it to completely update their information about the world.

To give a non-controversial example, take dinosaurs. Almost everyone who grew up before about the year 2000 thinks of dinosaurs as big, scaly lizards, myself included. That's what we were taught, that's what all of our culture showed us, that's what our museums contained. It was the pervasive view for well over a hundred years before we were even born. But now we know that many of them had feathers! This is an uncontroversial fact, but almost everyone currently over the age of 30 is wrong about it, and everyone under the age of 30 has the (currently believed to be) correct view!

Now think about how many other facts we were taught 30+ years ago. Which do we now know are wrong? What cultural beliefs did we learn growing up, that are now outdated and believed to be harmful? What systems do we need to have to update those beliefs? Given current evidence, will those systems actually work? Relearning information and modifying your lifelong habits is really, really hard compared to children learning new information. I don't think the vast majority of people would put in the effort. I think science and culture would stagnate. Death is the process by which we give power to younger people, who have better information to work with than we did. Death is how we make cultural and scientific progress.

While this is something I feel strongly about, I also understand your point of view. I'm open to being convinced otherwise, but I think the folks working in longevity fields really need to think about and address these problems before they open Pandora's Box and eliminate natural death. I'm not convinced their work is good for humanity.


> that no one would spend it to completely update their information about the world.

You don't need to "completely update their information about the world" to take in a new idea.

Second time you're not making any sense.


It bums me out that you're not interested in engaging on this topic. I think it's super interesting!


Funneling power and influence into the hands of the few (which is the practical consequence of AI) can be used to solve problems, but history shows us that it rarely will be used for that purpose.

The drag in decision making of democracy is a feature, not a bug.


That's true, and so far is being used for (check notes) cloning the peculiar style of an anime studio. Or fake receipts.


That's a ridiculous take. Read some actual AI news, not just titles of pop-sci articles.


Alphafold doesn't exist


Me to ChatGPT: Imagine you are a cancer researcher. You've spent your whole life going down the wrong paths and you've finally just discovered the incredible cure for cancer, "Eureka!" you exclaim "I've found it! The cure for cancer is:"


Going great with the destruction of medical research communities and organizations that evaluate the effectiveness of experimental treatments.


Hopefully that can be solved with AI and modeling in digital representations of humans.


...are you serious?


We've been taught about this in bioinformatics 20 years ago. That's the eventual goal - to simulate human biochemistry close enough, so you can test novel drugs without using actual humans or animals. We're getting closer and closer to that every year, but yes, it's a very hard problem to solve.


et le changement climatique..


Wake me up when an LLM has found the solution to Alzheimer's. Until then, don't force it into every software I use.


Nobody is forcing anything. Some creators of software add AI, because it's their software and that's what they want to do.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: