![Sean Goedecke, Glue work considered harmful (emphasis here, and in the rest of the quotes, in original): So why doesnât glue work get you promoted, if itâs so crucial to shipping projects? Are companies stupid? Are they deliberately leaving value on the table? No, I donât think so. Companies donât reward glue work because they donât want you prioritizing it. And they donât want you prioritizing it because they want you shipping features. Glue work is hard. If youâre capable of doing glue work well, they want you to use that ability to deliver projects instead of improving general efficiency. This is an argument to settle for mediocrity. If youâre like most people who have worked in big tech, it is plainly obvious to you that they are leaving value on the table. (Whether it is deliberate or not is a matter of whatâs in peopleâs hearts.) What big tech employers want is largely incoherent. [... rest in blog post]](https://pbs.twimg.com/media/HW8oC8pbMAAGgp7.jpg?name=small&format=webp)
ALT Sean Goedecke, Glue work considered harmful (emphasis here, and in the rest of the quotes, in original): So why doesnât glue work get you promoted, if itâs so crucial to shipping projects? Are companies stupid? Are they deliberately leaving value on the table? No, I donât think so. Companies donât reward glue work because they donât want you prioritizing it. And they donât want you prioritizing it because they want you shipping features. Glue work is hard. If youâre capable of doing glue work well, they want you to use that ability to deliver projects instead of improving general efficiency. This is an argument to settle for mediocrity. If youâre like most people who have worked in big tech, it is plainly obvious to you that they are leaving value on the table. (Whether it is deliberate or not is a matter of whatâs in peopleâs hearts.) What big tech employers want is largely incoherent. [... rest in blog post]

ALT When you disagree with someone, there are two paths forward. The first is to try to engage with what theyâre saying: what their reasons are, whether their position is true, what evidence could be brought to bear, and so on. The second is to diagnose them: to engage with the second-order question of why theyâre saying it. For instance, if youâre arguing with somebody about microservices, you could say âwell, you only like microservices because you want to polish your resumeâ, or âyou only like monoliths because youâre scared of distributed systemsâ. This happens a lot in online arguments. On the question âare AI models improvingâ, some anti-AI people will argue that people only think so because AI psychosis is rotting their brain, and the dumber they get, the smarter AI models seem in comparison. Likewise, many pro-AI people argue that people only disagree because theyâre Luddites, or scared of change, and so on. None of these arguments are about the quality of AI models.

ALT It feels so good to diagnose your opponents because thereâs no possible response. If you say âI think AI models are improving because these benchmark scores are risingâ, somebody might point out that the AI labs are training on those benchmarks, or argue that they represent isolated skills that donât transfer, and so on, making you feel stupid. If you say âyou only disagree with me because you want to roleplay as a revolutionaryâ, how could you be proven wrong? The more someone protests a diagnosis, the more it sounds like it must be right. Otherwise why would they be so upset?





























