How could it possibly be the rational choice if the ev doing so is so ridiculously negative?

Having student debt doesn't justify throwing away the rest of your residual income

What's the ev for going to college once you factor in graduation rates?

People that get two or three years of college debt and no diploma have a big hole to fill and a small shovel.

Anyway, I think ev isn't the right tool to model gambling behavior; dollar utility isn't linear. It's more about a small spending for a large potential. But then you get into repeated small wagers and such.

College graduates make over $1 million in their lifetime compared to high school graduates.

> Anyway, I think ev isn't the right tool to model gambling behavior; dollar utility isn't linear.

You're right. The more money you have, the less utility it gives you, which makes gambling for a windfall an even worse decision. Worse still if you include taxes.

I always thought those numbers were dubious because you are mostly comparing people who are capable of getting a degree with those who aren’t.

What about college dropouts?

> What's the ev for going to college once you factor in graduation rates?

Very positive (IRR ~9%). It's been studied extensively: https://openknowledge.worldbank.org/server/api/core/bitstrea...

If I borrow a lot of money to start a business and then don't start the business I would also be in the hole. So, don't do that?

you can always go back and finish later...that's what I'm doing.

My school expires credits after 7 years.

-If I don’t gamble, I’ll live my whole life as a poor serf

-If I gamble and lose, the same wage cuck life awaits me

-If I gamble and win, I may actually ascend and have the kind of life I aspire to

No because open source models are close behind

Compute costs will fall drastically for existing models

But it's likely that frontier models of the future won't be released to the public at all, because they'll be too good

I have yet to see an open model that is close to the previous gen of frontier models.

Benchmarks are one thing, real world performance is very different.