66 karma · joined February 14, 2020
* obstinate: same responses/approach even when presented with new information
* persistent: updated responses/approach when presented with new information
I take comfort in participating in this huge system of life and death and in knowing that billions have lived ordinary lives and died. I don't dye my hair and don't ignore signs of aging. I'm happy to grow older and retire and reduce my needs and spend more time with friends and family and when I die I'll join the legion of billions of decent, hardworking, perfectly ordinary people. It's a fine group to be part of.
Show your customers where traditional ML beats LLMs hands down.
People are using LLMs as a hammer for everything. That hype will die down.
Similar to wordle. Thinking of putting this in App Store. Do you think it's worth it?
This method has also worked for me when building features in a large codebase. Write a unit test first and then keep checking where the code breaks and fix those until your test succeeds. This is effectively TDD. Note that you might have to refactor the code for better design but it gets you started towards understanding the flow.
If someone cares to fix problems with social media, then it makes sense to do it from inside as well as outside.
(I don't work for Facebook.)
Took down that Substack not out of cowardice but b/c:
(1) Though I've endured 100s of death threats and experienced zero actual deaths— I'm hardly fazed anymore— they affect people close to me._
(2) I discovered SS is YC. On moral principle, I will no longer be using Substack.
Your original comment reads: > African Americans are 13% of the US Population. Isn't it reasonable to aspire to have at least 13% of African American staff (proportionally adjusted per state/city populations)?
Your thought experiment does not apply here because it assumes the same percentage of every population desires the same career. That need not reflect what we see in the real world. What if 50% of one race and 10% of another desires a specific career, the outcomes will reflect those numbers.
Again, you are ignoring the statement regarding people's interests and who they consider role models and what they consider as rewarding. In a world of equal opportunities, humans are still the variables, they will choose what they like. If there is a systemic problem that prohibits a certain race from reaching their aspirations, that is abominable and should be cut out.
This question somehow only arises in tech. No one cares about making races equitable in other industries as much. What about farming? Or similar proportion of dentists, farmers, chess players?
So much of what we read, fact and fiction, stands on the foundation of an omniscient first- or third-person narrator. We do not even think of questioning what the narrator tells us. Even mystery or suspense stories only hide a secret with the protagonist "peeling an onion" to find answers progressively. Deceptive incidents, the doubts of the detective, inconsequential objects, descriptions of crime scenes serve to distract us and we know we are being distracted and it is part of the game.
But unreliable narrators are a different beast altogether. As the story progresses, there are hints of doubt created in our minds regarding description of events, feelings of characters, gaps in some story lines which we hope will be reconciled later (and which are). A great author can do this so subtly that if you do not follow along closely it is easy to miss or, worse, mistake this to be a problem with the author's storytelling.
Fight Club is one such story. It is non-linear which serves to disorient the reader and at the end we realize the narrator was also unreliable. There are sentences like "Today I almost ate Marla's mother" and we are left wondering what that means until a few paragraphs later it is revealed. The non-linearity masks the narrator's reliability. The book is well-done even though the story-telling is . But you have to read An Artist of the Floating World to see what a master can do with ordinary events, conversations, descriptions of places by (someone who we will later in the story consider as) an unreliable narrator. Stories like these serve as a reminder that every one sees the world differently. We consider others unreliable while forgetting that everyone sees what they want to.
So many of our physiological systems and the way our consciousness works will have to change unrecognizably that the being that exists after these transformations can no longer be called human. If a thing thinks, it will be finite and want to be.
site:news.ycombinator.com intitle:"some term"It is a sign of open-mindedness to be interested in many things. It is a sign of maturity to narrow that list down to pursue them a bit more deeply, then discard what turned out to be not so interesting and pick new ones. The idea is to remember nothing is a waste of time - worst-case, you learn something about yourself while you were pursuing what turned out to be uninteresting for you.
I think of knowledge in terms of concentric circles, rings on a tree if you will. You start from the core and every month or <pick your own timescale>, you keep strengthening that foundation on topics you chose. You might choose an end-to-end project: data in a nosql db, expose that with a rest service, write a react frontend to consume that data. You can do this superficially without knowing what is REST or deeply about React. But if these topics interested you after the first project, you can do the same project in a better way the second time and so on...
Again, many things are a matter of time and if you try to rush them and try to be an amazing engineer in a very short time, it might lead to exhaustion. Reading Hacker News might make you feel people are doing amazing things - but remember at some point in time, those amazing people chose something, one thing, and pursued it. And for every person that's doing an amazing thing, there are ten who are in the process of slowly making progress.
If the so-called millenials are not reading DeLillo, the onus is not on DeLillo to say things that appeal to them. Rather, the onus is on them to figure out what DeLillo is saying and see if they identify with him.
(Edit: I am not American and english is not my first language. I like American literature among others.)
Of course, there is a gimmick in the book. I didn't realize that at first. I don't remember at what point and like usual gimmicks it hits you and you react as you realize whose story this is.
As an engineer, I live in a world full of definitions and determinate actions and consequences. Literature is what I have access to that constantly reminds me that there is more that meets the eye and truths are relative. Don DeLillo's literature ranks right up there.
This is similar to how there are many rituals in every religion that people might follow without understanding the underlying concepts. The rituals are not the religion but we still see people mindlessly doing things that their ancestors did.
I would much prefer that we build systems that automatically create that knowledge graph objectively (like the wikipedia) and layer an AI system on top that will create new connections, even with brute force, that our minds in a lifetime cannot. This will not stop us from having opinions. People will still come to different conclusions based on the same underlying data but atleast the data and the surrounding context will be available to all.