4 ms·

That is a fair thought, but, of course, depends on a constant (approximately) per-unit labour cost. After all, businesses could poach those 10 additional engin

That is a fair thought, but, of course, depends on a constant (approximately) per-unit labour cost. After all, businesses could poach those 10 additional engineers from the company beside them with a $2Y compensation offer, without the need for any more engineers. But if you need to pay 2x more for each labourer to achieve $2X gross return then the appeal is quickly lost. However, if you can create 10 new engineers that didn't exist before, then they will be incentivized to fall in line with $Y as well, lowering the unit cost of labour and making adding 10 more engineers to achieve $2X in return much more appealing. The keeping of the price of labour down is exactly why businesses were willing to make such investments.

jrockway 2y ago
Everyone wants to lower the cost of labor, but you can only get so much blood from a stone. People will do something else if software engineering isn't profitable. That's how things are going in the post-0% interest rates world. Software engineers aren't individually getting paid less, but less speculative software is being written. In 2021, everyone thought that the way to get rich was by throwing 200 software engineers at a shopping cart app. In 2024, the way is to have 10 people making a shopping cart, 10 people making a database, 10 people running HR reports, ... I don't think this makes the size of the industry smaller, but makes "I am sure I will find something interesting to do at this super company" less likely. There are no super companies anymore. (Except Nvidia.) Anyway, the way we increase efficiency is by automating more things. 50 years ago, you had to have a person come to your house to collect letters from you, then they would be mailed across the country, and another person would deliver it to the recipient. Now we have email, and you can just send someone a letter with no other humans in the loop. That's the efficiency increase that engineering brings, and even if we can't envision what we're going to do tomorrow, it will continue. What that means is that we will always need more engineers, and the scope of our role will increase faster than we make more humans to be engineers.

jrockway 2y ago

Everyone wants to lower the cost of labor, but you can only get so much blood from a stone. People will do something else if software engineering isn't profitable. That's how things are going in the post-0% interest rates world. Software engineers aren't individually getting paid less, but less speculative software is being written. In 2021, everyone thought that the way to get rich was by throwing 200 software engineers at a shopping cart app. In 2024, the way is to have 10 people making a shopping cart, 10 people making a database, 10 people running HR reports, ... I don't think this makes the size of the industry smaller, but makes "I am sure I will find something interesting to do at this super company" less likely. There are no super companies anymore. (Except Nvidia.) Anyway, the way we increase efficiency is by automating more things. 50 years ago, you had to have a person come to your house to collect letters from you, then they would be mailed across the country, and another person would deliver it to the recipient. Now we have email, and you can just send someone a letter with no other humans in the loop. That's the efficiency increase that engineering brings, and even if we can't envision what we're going to do tomorrow, it will continue. What that means is that we will always need more engineers, and the scope of our role will increase faster than we make more humans to be engineers.