rubicon33

I doubt it. There’s a very hungry work force ready to take any job available.

You will literally build nothing but the most primitive of devices unless you accept black boxes. In fact I'd argue its one of humanities great strengths that we can build on top of the tools others have built, without having to understand them at the same level it took to develop them.

There is a huge difference between greenfield development and working with an existing codebase.

I'm not trying to discredit your experience and maybe it really is something wrong with the model.

But in my experience those first few prompts / features always feel insanely magical, like you're working with a 10x genius engineer.

Then you start trying to build on the project, refactor things, deploy, productize, etc. and the effectiveness drops off a cliff.

I totally get these are very hard problems so solve and that we're on the bleeding edge of what's possible but I can't help and wonder when someone is going to crack real video understanding.

sure, maybe it's still frame-by-frame but so fast and so often that the model retains a rolling context of what's going on and can answer cleanly temporal questions.

"how packages were delivered over the last hour", etc.

Is there anything unique here happening for the video aspect or is it just taking snapshots over and over?

I’ve been looking for a good video summarizing / understanding model!

Can someone explain what one might use this model for? As a developer with a casual interest in biology it would be fun to play with but honestly not sure what I would do

Don’t have to prove it to me… I was asking genuinely.

What do you mean when you say “do a gut cleanse”?

Why do you do in production?

What? Where do you live?

I often ask people when I first see them, “ hey how’s it goin’? “

And 99.9% of the time it’s a short “good, u?”

Very much just a hi/hi type oft thing.

Genuinely curious as someone who wants to improve… what the hell do you talk about?

I’m trying to picture the random people I encounter while walking my dog, for example. Neighbors who I don’t know. That’d be a great place to start up a convo as I walk by but I draw a complete blank.

0 skill in this.

Fascinating post for THIS forum.

I don’t know about others but a big reason I went into software development (whether I want to admit it or not) is because I am “naturally” not very skilled in talking to, or dealing with , people.

While I’ve done very well in tech, the truth is, lacking this skill has only been a detriment.

The problem at this point is I truly don’t like random people, and I have no idea how to fix that. Often people are rude, inconsiderate, selfish, etc. the canonical example is being in a row at the grocery store, walking down the row, someone’s cart is just way out in the center blocking your path, and as you approach, they do nothing.

Anyways “people skills” are definitely a SKILL and one I’ve avoided to my detriment over the years. Need to work on that.

It all depends on how deep you want to go.

Building wrapper apps on top of LLM APIs takes barely more knowledge and training than normal frontend and backend dev works.

If you want to develop your own models, that’s probably the next layer where you need to learn about training, etc.

Even deeper is Ai research… once upon a time LLMs were not the bleeding edge… it takes research often times theoretical and steeped in physics and math to develop your own “ai”

Eh, sure, but marginally better if not the same as Claude 4.6, which itself was a small bump over Claud 4.5

But actual progress seems to be slower. These modes are releasing more often but aren’t big leaps.

there’s something of a paradox there. Reduce the context window and work on smaller/tightly scoped requests? Isn’t the whole value proposition that I can work much faster? To do that, I naturally try to describe what I want at a higher, vaguer level.

I’m an “old school programmer” just like you, but still use Claud code.

For greenfield projects it’s absolutely faster to churn out code I’ve written 100 times in the past. I don’t need to write another RBAC system, I just don’t. I don’t need to write another table implementation for a frontend data view.

How Claud helps us is speed and breadth. I can do a lot more in shorter time, and depending on what your goals are this may or may not be valuable to you.

A share a similar frustration as you, that it seems “people” don’t care about / never question things, but for me it’s really about one big question:

Why the f*ck are we here? Why does ANYTHING exist? What IS this reality?

How “nobody” (very very few) people are trying to figure this out or are bothered by the question and open to talking about it blows my mind mind.

Honestly you need to try Waymo. It’s in a league of its own.

That pretty much nails it. Starting something from 0? AI can get you from 0 to something in blistering speed.

Working on a real product, in production, at scale?

Tread carefully..

Gosh, given you’ve been there I have to ask what allowed you to get out of that and pursue only things that interest and excite you?

Unless that polish has value. Thats the hard part, knowing what features to really spend that extra time on, and which to 1 and done.

Ultimately the vision for the company and the product have to come from SOMEWHERE.

Small companies often have the CEO in a product position in my experience.

Still isn’t ideal though and more to your point, how they actually do the product role is really what matters. I.e. chasing shiny objects.

Oh my god you just perfectly described the frustration of working in enterprise SaaS. It’s been fun in some ways but the constant churn of almost-but-not-entirely-production-grade software is soul crushing. We celebrate and reward speed to market and lack of process in a way that feels unhealthy and unrewarding.

I’ve worked at consumer facing companies but also other enterprise SaaS and have to say I’ve never seen it done like this before. Just ruthless pursuit of features over polish, craft, etc.

How does it compare to Cursor with Claud? I’ve been really impressed with how well Cursor works, but always interested in up leveling if there’s better tools considering how fast this space is moving. Can you comment to how Codex performs vs Cursor?

LLMs are actually terrible at generating art unless they're specifically trained for that type of work. Its crazy how many times I've asked for some UI elements to be drawn using a graphics context and it comes out totally wrong.

Wow. I’ve never actually heard this take before but I think you’re spot on. Vibe coding has serious limits as the scale of the project grows. As the complexity grows, human language becomes a crude tool compared to just coding and you will find your efficiency falls off a cliff trying to get the LLM to understand and effectively change or implement what you could have done by hand.

If you’re asking this question then you have really no idea how the sales engine works at a company. It’s inherently incentive driven.

> I’ve always ended up in a position where I was pretty miserable after a couple years no matter what company I was working for at the time. I’m annoyed at all the process, I’m annoyed with the people, I’m not interested in what the company is doing, and it just feels like such a grind.

This should absolutely be your red flag, your canary in the Coal Mine, your alarm bell.

If you are CONSISTENTLY not enjoying the process then that tells me you don't actually like software development enough to overcome the negatives. There will be negatives at every company, and certainly software as an industry has its share of downsides, but your enjoyment of the craft should easily make day-to-day worth it, even fun.

> Earlier on, I thought that this was just a symptom of not liking the product, culture, or just getting bored of what I was working on at the time and could be fixed by going to a different company. But now after working at 3 different companies of slightly different sizes, I’m starting to think that maybe I just don’t enjoy this field as much as I thought I would.

Sounds like you get it. Smart to work at different companies, products, sizes... Since you've given it a shot, I think you've done your due diligence here.

> All of the companies I talk to are only “hiring the best” and I always think to myself “that’s not me” so I’m wondering if it’s time to just pursue other options.

Toxic aspect of the software industry. I work at a company doing exactly this. We're instructed to only hire people we would "die on a hill" with.

It's SUCH a software / engineer minded thing to do. You're either GODLIKE PROGRAMMER or you're GARBAGE. It's either a 1 or 0. You're either going to take us to the moon, or you're burning our cash and a waste of time.

The reality is so much more nuanced and colorful. People have a wide range of skills and what they don't bring to the table in engineering prowess, they might bring in other ways like product creativity, or passion for design, etc. Moving a company forward can happen in more ways than 1 but this industry has a toxic fascination with developers.

And that's coming from someone who is a developer, loves coding, and has been quite successful at it. I wish we (the industry) had more willingness to train, more willingness to tutor, and less obsession with the 10x autistic kid.

“Religious nonsense” is such an engineer minded thing to say.

I’m with you for sure, but the truth is systems like religion, art, design, etc all serve a functional purpose to trick the mind, calm the mind, etc.