he11ow

If I may be so bold, the fallacy here is of time. Dare I say, you are already surrounded by friends, and not noticing it.

It seems like you measure "friendship" not by the time spent with them when you go to events or in school, but what happens afterwards. In other words, you are discounting to zero the moments that are actually happening in your life, for an imagined future of what a friendship might turn out to be. But life only ever happens moment to moment. That time that you spend with people at an event, that IS the thing.

There is no "friendship certificate": some people, you'll only spend ten minutes with; some, you might spend an afternoon with; others you might end up seeing once a week for a term. You might date a girl for a fortnight, or a couple of months, and then break up and never see each other again. It's ALL GOOD.

If I understand correctly, none of the above would qualify in your eyes as "real friendship", only the deep, deep kind. You might find it helpful to learn about Dunbar's number, and the size of the model's concentric social circles. Most people end up with 1-2 very close friends (possibly including their spouse). That is the outcome over a lifetime of making connections. Popular culture markets this idea that we should all be carousing with a tightly knit group of friends - no statistical social evidence bears this out.

When sinking into deep analysis about all your self-perceived inadequacies, your attention beam is directed inwards. That's unhealthy. When it's directed OUT, into the world, it takes you out of your head, and lets you see that everyone, every single one (person or otherwise), is struggling in their own way. Everyone's got a thing. Looking out builds empathy and kindness and affection towards things that make you happy, and gazing inwards makes you miserable.

You are surrounded by people in school, and are at a stage in life where you get to be interested in whatever. Any experience you have, even if it doesn't lead to lifelong friendships, is worthy in the space of time it occupies.

Throughout my undergrad I had friends come and go. I'm not in touch with anyone I met at Uni during those years. It doesn't make the actual time spent with these people any less special.

There are two books on the issue by a guy called Steven Pressfield, "The War of Art" and "Do the Work". I've read the latter and it's good. He talks about how resistance strengthens the closer you get to ending a project. I've found it useful in my work.

Don't know about "ideal" but YC have a video where Michael Siebel talks about the dev cycle they had at Twitch. Before they applied it, he says, they wrote lots of redundant code and also get stuck on stuff that didn't matter.

Eventually, what they did was set up two-week dev cycles. Before each, they'd write down suggestions for features/WIP and stick everything on the board. They would also estimate if a feature was large (one week of work or over), medium (2-3 days) or small (up to a day). Then, they'd pick the most urgent things off that list, and work only on them. Two weeks later, same thing. But you don't keep a list of the ideas, the other things don't go into a queue.

I applied this for a product I was building last year, and it definitely helped build faster.

Not sure about measuring things in LOC or commits.

Can You Say...Hero? Fred Rogers has been doing the same small good thing for a very long time... https://www.esquire.com/entertainment/tv/a27134/can-you-say-...

A supposedly fun thing I'll never do again (originally titled 'Shipping out') David Foster Wallace goes on a luxury cruise. There's a PDF version online, but the reading experience doesn't compare to reading it in a book. My copy is tatty by now, still keep going back.

A vast open source community. If you want to go fast, go alone. If you want to go far, go together.

I'll say something that sounds really unhelpful until one day it makes sense. You don't really choose the process, the process chooses you. You can't make yourself 'enjoy the process' for something that doesn't resonate, because it will always feel a chore. So the one and only trick is to really pay attention to the thing you keep going back to, and in a way accept that this is a path.

Some people might say "But I only enjoy video games!" and if that's the case, I don't really know what to say to that. All the people I know who have found great alignment with a path of progress are ones where the effort tickles an itch they would have regardless, so might as well scratch that itch.

And there are still goals, that doesn't change.

I used to be a big fan of LinkedIn. I've posted, here and elsewhere, expressing these views. But some of the value LinkedIn used to offer is definitely eroding.

The feeds have become really terrible. This wasn't always the case. Before, a lot more weight was given to posts from people you were connected to, vs. the weight given to their overall engagement. This offered two kinds of value: One, you were more likely to keep updated on what's happening within a professional niche (assuming lots of your connections were in that niche too.) Two, your own writing would be shown mostly to people in your business niche, which helped foster trust.

I've formed a lot of useful connections this way. A lot of client relationships, past and present, started out this way. A bunch of people helped me in a bunch of ways (intros, advice) because they felt I was a real person, not a rando profile image.

That's mostly gone now. A useful post is now so rare, and the rest is "engagement" fodder. Eventually, I turned off the feed, and that's how it stays most of the time. (I installed an extension someone once posted about here.)

The chat feature, which was always really powerful, is still there. But it used to be that the flow was written insights -> chat -> in person/Zoom. Without that slower process of getting acquainted, the connecting itself has less potential to become really useful.

For anyone saying "I definitely won't get in" here's a counterpoint that has little to do with whether your application succeeds or not: Do it for yourself, not for the approval.

I found that filling in the application, without even sending it, is a useful exercise. It makes you think hard about what it is you are trying to achieve with whatever it is your building. It's very helpful when you come to talk with clients.

You know how they say launch fast, launch early? That's part of it, because launching means talking about it to other people. Those people will have questions or objections or an incomplete understanding of pre-existing solutions. A lot of the application is about this kind of stuff.

Personally, I wanted to fill the application irrespective of YC. I figured, if I'm serious about building something useful, I shouldn't just build it and que sera, sera. I should know what to focus on. And since they've learned a bunch about building companies, if they care about something enough to ask about it, that's a good pointer that I should care enough about it to be able to answer it.

For example, I filled it in when a solution was just an idea, as if it were already built. So it made me set expectations, and clarify what I want this thing to be. Then, once it was built, the first thing that happened when that solution met with clients was this: I'd give them the short description I've written in the application, and they'd ask - Oh, but does it have to do just that?

Without the application, there wouldn't have been the short description, without which there wouldn't have been the useful client feedback...So, again: do it for yourself, not for the approval.

From everything you've described, it seems the company is structurally built so that you would be on a need-to-know basis. I'm going to use the analogy of horse blinders: processes around teams have been built so that ICs would focus on their tickets, just as you describe. This is not an act of malice - in a company this big, to steer the ship in any one way, you don't want noise coming from half your workforce, nothing would ever get done...

But it does mean that all the gaps you are describing are, in many ways, there by design. So going against them is going against the grain of the organization. I think it generally helps to bring something to the table. A point of view on something external to work that could add value inside it. In many ways, though, it still requires hustle, in the sense of reaching out to people and creating opportunities.

For example, you might take an interest in a certain type of antipatterns, and want to propagate the knowledge internally about it, and how to resolve it. (I'll assume the company doesn't actively seek to cultivate this anti-pattern deliberately. You never know.) You might say, oh, I can put together a colloquium about it. So you find the person in the org who's in charge of the events, and talk to the person who books the rooms, and talk to people around you to get ideas on how to promote it internally...you get the drift, it's not all that different to doing these things outside a company. (Except the part where you risk treading on more toes.)

One book that's really useful for just about anything is "The Goal". It teaches you to think in terms of processes, bottlenecks and constraints. With this perspective, you learn to look at the work you and others do not as a series of discrete tasks, but as part of a flow of production. It broadens the perspective on the types of questions you can ask.

But that's the thing: I wasn't looking for this, and even having found it, I won't bother looking at "the competition". In my mind I don't even see "missing features" because I have zero awareness of what features could be.

I'll be honest, would I pay for it? Probably not, because I already pay Spotify. But I did look you up on LinkedIn earlier today, so as a calling card it works.

And, judging by the parent comment, you clearly have a knack for marketing it, which tends to be the differentiating factor...

I'm glad you posted about it here! I've not come across it before and really enjoying it right now.

The label, like any label, is a tool, it's not who you are. It might be a good tool for you to better understand yourself (your struggles, your preferences), but still not be a great tool for others to understand and communicate with you.

For one, unless people have had direct experience with spectrum diagnosis, I think most still think of autism as the extreme end of the spectrum, of severe disability. For another, the label itself doesn't offer any knowledge of what to do differently.

In my mind, there are two things that you can do that might help:

First, it helps other people understand better when you make it about vulnerability. When you say things like "I dislike crowded spaces, it stresses me out." or "It takes me a little longer to come up with a retort, I've always been this way" it does exactly what you want: allows you to be your authentic self in a social way. Everyone has things that are outside their comfort zone, so being honest about yours is very relatable.

Second, you mention people "viewing hints of antisocial behavior". It's very challenging - to yourself and to other people in the situation - to be both "social" and "antisocial" at the same time. Social behavior requires effort from everyone. That's why advice columns have always existed, that's why there are so many self help books. Recognize that, as a baseline, practically everyone is walking around with some internal sense of inadequacy, and that in social situations most everyone puts on a front. Granted, most will understand the codes to putting on the front better than you. But the effort is there. Which is to say, if you decide you are going to be in a social situation, there isn't a way to square the circle, be anti-social and have it glide. But you can manage the effort to your capacity: step aside, limit the time, set boundaries - whatever works. It can help to know that really everyone else is doing it too.

I'm sorry you feel so scared. Even if the fears aren't real, they feel real, and our body responds to them as if they were real. So your fears may be irrational, but the pain is very real.

I think the first thing worth saying is that, in my experience, this anxiety is more common than you'd think. The common narrative is that a young man (and it's always a man, right?) has only a few short years to make it big before it becomes clear to the world they're worthless. Never mind that zero data backs this up, this is about narrative, and storytelling always trumps data.

So you come out of college, you've taken two breaths, got some kind of job, you're 24 and time is running out. Naturally, you begin to question everything you've done up to this point, and feel you have one last ditch to fix it...or else.

The thing to understand is that this is not life. This is not how life works. This is a story that is so evidently fictional, that the only thing stopping you from seeing it is sheer panic.

You say you're scared for your future in software engineering. I am assuming you see, right now, a future you want in software engineering? (Because maybe you don't actually want it, just feel it's 'the path' to somewhere?) Because if you do want to be in software engineering, school is not the fix. Learning and doing is the fix. (I am massively formally educated in CS. And I still stand by everything I just said.)

Life is long - especially from where you're standing now. You've just started. You're not doomed. If you lose your job, then you'll have lost your job. You may be out of a job for a while, and then you find another job. And then you leave that one, because you got a better one. You may get an opportunity to relocate and seize that. You may meet someone and decide it's a good idea to move halfway around the world. All along, you're picking up skills and know how. If you're even just a bit methodic about building up your skills, you'll find that jus five years down the line, you already know some actually useful stuff.

It's just the way it goes.

According to the latest Stackoverflow developer survey, 40% of developers use React, well above any other web framework. (42% for "professional developers").

It doesn't answer to your question, but it does answer to mine on these flexes in HN around niche frameworks and the difficulty of putting them into context.

I'm glad people are working on alternative solutions, and maybe ten years from now everyone will be using something else. But right now, 235 MB is an awfully small price to pay for something that just works, especially as a lot of it doesn't actually show up in deployment.

I'm surprised no one has pointed the obvious, which is using chatGPT as a coding buddy. I've done pair programming with developers who were much more capable than I was, so can compare the experience, and I have to say - it works just as well. The solutions it offers are not infallible; when it comes to pair programming, that's a feature not a bug. Often times the solution will emerge only through the back and forth. But it does a tremendous job of helping locate local bugs, introduce functionalities you may not be familiar with etc. etc.

There was a system like that. Roll up, roll up, I'll treat you to a story:

It all started in 1957, with the Sputnik. The US was entirely taken by surprise. The belief that the Russians were way ahead instigated an internal crisis, which, in turn, led the US to re-evaluate its national maths curriculum. Thus was born a think tank called the School Mathematics Study Group (SMSG). They developed a radical reform in mathematics education known as 'New Math'.

This was rolled out nationally, to great criticism. The teachers were ill prepared and the parents felt clueless. Look it up, there were 'Peanuts' strips from the period mocking it.

Now, New Maths focused initially on the early years, but then, in the mid '60s came a second round, and one specific initiative was the Secondary School Mathematics Curriculum Improvement Study (SSMCIS), and the guy who heads it happened to be a professor at Columbia University Teachers College. This last tiny fact probably means exactly zero to you, but it is central to why I even know all this stuff.

Anyways...The program's signature goal was to create a unified treatment of mathematics, so that instead of studying the normal curriculum you'd basically study maths the way you're taught it at university: set theory, group theory, axioms proofs and logic, all the way up to calculus. The programme was intended for grades 7 through 12, and was rolled out initially in the NY area and then later in select schools in other affluent cities in the states. It only ever targeted the top 15-20 students in the a class body. That was for sure the right call - this ties to your question, so more on that in a minute.

Eventually two things happened: one, the programme ran out of funding. Two, by the mid-seventies there was a massive backlash against New Maths and the US decided maybe it's okay to just leave it, since the Russians didn't end up winning the space race after all.

I would have known diddly squat about this whole affair were it not for a curious corollary. In 1953 one very specific individual happened to be on a mission in New York. He was a former Russian Jew who studied maths in Canada, served in WWII for the US military and eventually made it to Israel. He was an educator and had somehow caught on to what was happening in Teachers College, and upon his return to Israel, he started devising maths curriculums and translating the original SSMCIS textbooks. This was now dubbed 'The Columbia Programme'.

Fast forward almost 40 years later. In a way I've never managed to uncover, that programme survived, and was still being taught in one of Israel's gifted programmes. I entered my first maths lesson at seventh grade never realizing just how much this would end up influencing the person I'd become. Our textbooks were literally photocopies of the typewritten texts. The teachers has added to it bits of the regular curriculum plus more practice exercises, which the original textbooks lacked, but they left most of it as is.

In the first three years, no one in the class was allowed to drop out and take 'regular' maths. For many, even in a cohort that was already pre-screened for academic achievement, this was a struggle. For sure, once highschool rolled along, anyone who hated it could switch back to the regular national curriculum.

Of the people who stayed, nearly everyone went on to study Maths, Physics or Computer Science to graduate level. This tended to happen in the years above and below as well. Over the years, though, the programme got smaller and smaller. I'm not sure it still exists.

To your question:

You absolutely CAN get highschool students to leave secondary school with advanced-undergraduate level of mathematical maturity. (And, BTW, the Russians are still ahead there...) But you can only do it for a small minority. Not because of elitism, but because most people aren't a good fit for this path.

At the time I intended to write up all of this into a nonfiction essay. But other things took greater priority, and I just left it there. In a way, it's been nice telling this story here.

My perspective here is shaped by two things: One, my kids are a little older than yours (so I've seen a little further ahead), and the experience to have changed up my career not through promotions but through upskilling.

Prioritize spending time with your kids. You already know that you will never get that time back, but what you don't realize is how much you'll value in retrospect having spent that time.

Career advancement doesn't happen because you put in 20% more hours. It just doesn't. No one appreciates you more because of doing 'more of the same'. In each of the jobs I've had, I look back, and some of my former colleagues are doing pretty much the same thing for a little bit more money. Some get promoted, not everyone. MUCH WORSE is, in having spent all this time doing 'the job', windows have closed to do anything else.

Real leaps happen when you're able to be X but then suddenly also Y. This means upskilling. Finishing your regular tasks a little early means nothing; automating your tasks means a radically different state.

The great thing is, there is more time than you think. People think their careers are set by the time they hit their late 30s. But for many others, that's when things get exciting.

I use it when I'm writing code.

Grandiose, large-scale automation has to start with a simple system that just works and is miles better than what it is replacing. To shrug off how AI changes the act of coding is, in my mind, to miss everything that matters about this technology.

When I use it, I'm in two halves. One part is engaged in the task itself (whatever I happen to be building). The other part is observing why the process feels comparatively so much better.

I was fortunate enough to become friends with a Japanese person in my community this past year, and I had the joy of experiencing the everyday quality of Japanese aesthetics. It was inspiring. It goes well beyond choice of décor or clothing. It's a profound commitment to aesthetics, one I struggle to express using English vocabulary. It's like, if something isn't thought out - why even bother?

What problem are you solving and for whom?

If it's a real problem for real people (and not an imagined one for imagined B2Bs), what are they doing right now? Is it good enough?

Why do you, of all people, want to solve this problem, of all problems? What's the fire emanating from inside?

With good answers to these questions you'll find it a whole lot easier talking with clients. If the problem is real and painful enough for them to have cobbled together a solution, they won't care if you made your software out of matchsticks.

Honestly, I feel that every discussion of this sort should have at least one Memento Mori comment. I don't know what to say to your situation, because we're different people.

But I just always had a strong sense that life is too short to be spent on aggravation from people who stop mattering the instant you leave a job.

https://www.telegraph.co.uk/christmas/2023/11/15/what-i-know...

My introduction to the web was, I'd say, about a year earlier - and I don't miss it at all.

Sure, it was nascent and full of promise. But that's mostly what it was - promise. Yes, there were all these offbeat websites you could find in the Yahoo directory, but that was not much different to aimlessly scrolling (or, fine, clicking on HN stories).

When I was studying CS one didn't have a hope in hell of building anything of substance independently, because Every.Single.Little.Thing had to be built from scratch - in C or C++ at best. And you'd better have enjoyed those late nights at the lab, it's not like you had anywhere to turn to at 1am when your code is buggy...

The ability to just build is, I feel, far more democratized today - and there's still a long way to go. So to me, the web is still very new, and still full of promise.

The Internet was also incredibly elitist back then. In 1998 only 17% of people in the developed world had access to it. Globally, that number was 3%...You could meet anyone in the world! So long as they were rich. Today these corresponding numbers are 90% and 63% - I think it's better that way. Certainly, that has been the promise, hasn't it?

And walled gardens? AOL wanted to own the Internet point blank... I think if anything, our generation was dumb the way all people are dumb in the face of something new. I think our kids would look back at the time we wasted on FB and Twitter and think - "You morons. You spent hours scrolling reading random people's posts? Were you THAT bored?"

Besides, we ARE, right now, part of something very new and incredibly big. Anyone can play with transformer models, with about the same relative ease and accessibility as you could with the Internet 25 years ago. (Which is to say - as a user, easy enough. As a builder, requiring a healthy dose of determination.)

I think of technology as, like, a tidal force, that just sweeps everyone. It does good and bad. It creates and destroys. Pining for any moment in time is like pinpointing a moment in a stochastic process...it was only ever just a moment in flux.

Thanks! When it comes to choosing where to work with these models, which compute platform do you recommend (assuming locally doesn't really make sense with my resources)? Colab? AWS StudioLab?

Which is your go to?

This may be an unpopular opinion, amidst a crowd who are maybe more comfortable with the whole deployment side of things:

I use super boring tech (React, Node) for fairly simple apps. When I was looking to deploy one recently, I looked at a bunch of alternatives, including the ones listed.

I couldn't get answers to even basic questions from the docs, and was in no way convinced I'd actually be able to deploy my apps. Plus the pricing differences were negligible.

Heroku was clever in that it worked to make deployment so easy every basic tutorial chose to use it. Its documentation is robust. I found it easy to use, and that counts for a lot with me, because that's not where I'm looking to spend attention. I want the damn thing up, and if someone can make life a bit easier, yay.

I'm only writing this because I think the vast majority of these companies' clients are probably closer to the profile of someone like me. Meaning, they're a little apprehensive about dealing with deployment trouble, and trying to minimize that specific risk.

You're like the person who's struck gold and believes it to be tin foil.

You're being paid like a developer, but spending your days in direct contact with real customers and their real pains? You say it's a fintech, so odds are this is a specific niche of customers, with specific needs, and you get to learn about what they are?

Ask yourself if by "haven't learned much" you mean "coding stuff".

As for the "things are pretty much disorganized" - again, it's a mindset thing. If stuff is holding back making customers happy, and you find a way to make it better, that's a direct impact on the bottom line. Whether it's through better documentation, or a set of processes, or writing a bunch of scripts...that's what ownership looks like.

And sure, you're not an owner in this company, but assuming you see yourself taking ever more senior roles (with compensation to match), this kind of experience will probably help you more than writing a couple of extra buttons on an app.

This is absolutely marvelous! I've been looking for something just like this. The batteries included aspect matters a lot if you just want the damn thing to work within a larger workflow, and the text editor is just one step along the way. Looking forward to trying it out!

Thanks! Will check out these services, MLOps is definitely where the biggest pain points are right now.

This looks great!

Can I ask, are you using a HuggingFace endpoint to hit the LLaMA models, or deploying it yourself? Still new to this and trying to understand how putting the large models in production works...

I've tried every shortcut possible only to realize the long way is the short way.

Sign up to a Udemy web-development bootcamp. Take it class by class by class. It won't necessarily be quick. Some things won't work, because things will have changed since the bootcamp was recorded - that's practice onto itself in how to fix things - and the Q&As on Udemy courses are very good for that. Afterwards, sign up to an intermediate course that actually builds a project in a given framework. It doesn't matter which, because once you know one, it's a lot easier to contextualize a different one. Anyhow, the framework doesn't matter, at least not for the first year or two; being able to create and ship something does.

What I realised in my journey to pick up web development was, the long way is the short way. YouTube, blogs, forums etc. look like they might make things easier. They don't. The only thing that I found has made a difference is to fully surrender to the process.

I bought a webcamp course on Udemy. And I sat through each and every class. And I did each and every exercise. I took notes. I aimed wherever possible to try things first before seeing a solution, even if it was during class and not officially an "exercise". The course was a bit old, so some things didn't work. When something didn't work, I sifted through the Q&A to look for a solution, because that in itself is a skill you have to use as a developer. I'd wake up early in the morning to cram a couple of hours before the day began. Some weeks, I was able to devote more time. Some weeks, I took a break. All told, it had taken me about six months.

By the end of which, I still didn't have enough knowledge to build a full webapp. So I bought an intermediary Udemy course covering the MERN stack. I was just as diligent with it. And by the middle of it, I could confidently build a demo version of the app I wanted to build. It looked ugly.

My intent was to improve my HTML and CSS skills, so you can guess what I did next. Yes, bought another Udemy course. I ended up not doing it, because it was better value-for-money to get a professional developer to rebuild it, but now I had my knowledge, my demo and I could manage this effort.

My point is, it took several months of humbling work just to get to the point where I say - gee, if I want to improve I need to do all this other humbling work. I deliberately say humbling and not hard because, frankly, the coding wasn't hard. The hard part was decisively letting go of the ego and accepting the no-shortcuts, heads-down, do-the-work way was the only way through.