TechTakes

Deep Research is the AI slop of academia — low-quality research-slop built for people that don't really care about quality or substance, and it’s not immediately obvious who it’s for.

it's weird that Ed stops there, since answer almost writes itself. ludic had a bit about how in companies bigger than three guys in a shed, people who sign software contracts don't use that software in any normal way;

The idea of going into something knowing about it well enough to make sure the researcher didn't fuck something up is kind of counter to the point of research itself.

conversely, if you have no idea what are you doing, you won't be able to tell if machine generated noise is in any way relevant or true

The whole point of hiring a researcher is that you can rely on their research, that they're doing work for you that would otherwise take you hours.

but but, this lying machine can output something in minutes so this bullshit generator obviously makes human researchers obsolete. this is not for academia because it's utterly unsuitable and google scholar beats it badly anyway; this is not for wide adoption because it's nowhere near free tier; this is for idea guys who have enough money to shell out $whatever monthly subscription and prefer to set a couple hundred of dollars on fire instead of hiring a researcher/scientist/contractor. especially keeping in mind that contractor might tell them something they don't want to hear, but this lmgtfy x lying box (but worse, because it pulls lots of seo spam) won't

OpenAI's next big thing is the ability to generate a report that you would likely not be able to use in any meaningful way anywhere, because while it can browse the web and find things and write a report, it sources things based on what it thinks can confirm its arguments rather than making sure the source material is valid or respectable.

e: this is also insidious and potent ~~attack surface~~ marketing opportunity against clueless monied people who trust these slop machines for some reason. and it might be exploitable by tuning seo just right

116 million

There’s no way that what they’re buying is worth that much.

thanks! It might be uncommon because it's a real pain in the ass to keep it short. Every time I make one I stress about how easily my point can be misunderstood because there are so few details. Good way to practice the art of moving on

will automate millions of jobs if it works, might kneecap the economy

will kneecap the economy if it works, too. Because companies certainly aren't going to keep people employed in those millions of jobs.