Live data from Hacker News

Anthropic Economic Index report: economic primitives

anthropic.com

81–82 of 82 posts

Re: Anthropic Economic Index report: economic primitives

#81

Earlier quoted context omitted.

Knowing exactly what you want to build is pretty rare and is pretty much limited to "rewriting existing system" or some pretty narrow set of projects In general, delaying infrastructure decisions as much as possible in process usually yields better infrastructure because the farther you are the more knowledge you have about the problem. ...that being said I do dislike how agile gets used as excuse for not doing any p…

If someone comes to you and says: "I want to build a platform that does WhizzyWhatsIt for my customers, it has to be on AWS so it's mingled with my existing infrastructure. It needs to provide an admin portal so that I can set WhizzyWhatsIt prices and watch user acquisition make my bank account go brrrrrtt. It needs the ability for my quazi-illegal affiliate marketing ring to be able to whitelabel and brand it as the…

Probably waterfall stuff that have actual clear functions and integrations (if you can extract all that system gets and what it does with it there is no reason to agile it) then slowly get thru the current mess, documenting it at each step while trying to replace it with something better.

Replacing existing system (and especially one you didn't write) is pretty much always the hardest case.

Re: Anthropic Economic Index report: economic primitives

#82
post #61

Earlier quoted context omitted.

I think we can surmise how bad that looked from the omission..

I wonder if it is even possible to get such measurements. With so many things affecting output, how can one establish a baseline or avoiding to compare apples to oranges?

the papers I've seen that tried just picked a group of developers at supposedly similar level and gave them tasks to do with or without being assisted by AI.

At the very least for programming the decline of improvement compared to difficulty of task and skill of programmer quickly got into "AI technically helps but time spent fudging with prompts is more than time saved".

However if your task is to write some SEO slop about your products for your company's webpage, or to make a bunch of business emails to customers, I can absolutely believe AI can double productivity there, there is nothing of value produced that requires skill, it's just a few bits of info filled with fluff around them

Post reply on HN