Death by KPIs. Management makes it too risky to do anything but benchmaxx. It will be the death of American AI companies too. Eventually, people will notice models aren’t actually getting better and the money will stop flowing. However, this might be a golden age of research as cheap GPUs flood the market and universities have their own clusters.
So big tech wants to court Trump with millions in donations and now that the big bully they supported is bullying them.. we’re supposed to feel some kind of sympathy? Am I missing something here? Why did Anthropic get involved with the military in the first place?
> What's the last 10,000+ employee org you founded and scaled?
A lot of smart and talented people could do this if given the opportunity. Jack was at the right place at the right time and had enough talent. Same with Elon and others. That’s kind of what happens when you have a population of hundreds of millions, a few get lucky and have enough talent to not screw it up.
It’s best to avoid being delusional and acting like billionaires are 5000 IQ geniuses. They’re regular people too, albeit, yes they are smarter than the average person you pull out of Walmart.
There are also plenty of smart people who simply do not care to run or start businesses.
Asking people to not use or develop specific types of apps is not asking for a stop to technological progress.
Developing and using apps like this is a choice, there’s no inherent force propelling them into existence other than people thinking it’s a good idea. I’m explaining why it’s not a good idea.
No, you have to understand that technology is neutral and that there aren't inherent design flaws because only the people are flawed. Just let developers develop in peace
‘Technology is neutral’ is one of the most cynical excuses, as it permits developers to remain in a child-like fantasyland where consequences do not exist, and shifts the blame to… nowhere I guess, since we also don’t like regulation or moral takes on what gets built.
That's what happened though? First humans built sheds, then we built 2-story buildings, then taller and taller, until we built skyscrapers. Obviously it wasn't a single structure, but we did have to evolve our thinking on how to build things, we didn't just start building a skyscraper before we built a shed.
But you didn't upgrade the shed into a skyscraper. The iterative process you describe involves a human respecifying from scratch using the knowledge developed building the previous instance and seeing it's limitations first hand. That part can't be automated, no LLM is going to challenge your design assumptions by itself. Hence people pushing agent-built projects way past what their inherent architecture should support, delivering an unmaintainable code spaghetti.
You can't do that. A small bike shed is often just put some concrete blocks on the ground, and then build on top of them with wood. A correct house needs a stronger foundation at higher costs (sheds larger than bike shed are build the same way), but is still made of wood. A skyscraper is built with a very different foundation, and needs a steel frame that would not be affordable in a house. In between the two there are also building made of brick which allows building taller than wood. (and there are lots of other options with different costs - engineered wood is different)
Point is though eventually some system runs out of ability. It works different in programming from physical construction, but the concept is the same, eventually you can't make a bad early design work anymore.
to put it in another way than the other replies: you will have 100x more pushback to an arguably-necessary ground-up rewrite instead of "just add this new feature to the existing codebase", even when you (as an engineer) know full well why "just adding a feature" is probably a bad idea.
That's exactly why software is so bad. No one ever knows their shed would ultimately have to become a skyscraper, and management doesn't allocate any budget to lay stronger foundations when expectations change; you make do with what you have.
See also: "there is nothing more permanent than a temporary solution"
It’s actually the opposite - you actually can. The feel I'm getting reading anti-AI sentiment is people are expect one shot results out of limited context.
I'm pretty sure that you can't gradually upgrade a shed into a skyscraper unless you pour a skyscraper-ready foundation before even starting on the shed. But if you're doing that, why start with a shed and not with a skyscraper?
Not sure why you're trying to bring AI development into this.
You can, start by clearing and grading the site - get a shed up over your head. Then you can start then start the skyscraper next to it and work out of the shed.
