Also remember that "tech" and SV in general are full of shit when it comes to basically everything, all of the startups making billions do so through means that are shady af or The Same Thing, tons of apps out there that do nothing valued at $50,000,000,000k+, also Sam Altman is gay, Jewish, a murderer, and fucked his sister or something, and got ousted but came back. Everything is shady, weird, and Jewish. Facebook makes money by Curating ads for boomer proles. Instagram makes money by fat shaming Fat Girls. it goes on and on. The one time they tried doing something irl with self-driving Cars, they Failed. SV tech needs a "closed system" for the product to succeed imo. AI needs so much more than that to succeed. Discuss.

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49412440)

The reality is that a lawyer who can barely turn on his computer, like EPAH, could write better code, and dominate a analyst/engineer job in corporate America if he was fully invested and balls deep with AI.

The problem is getting 85% into a project using innate knowledge, and then bringing in AI to solve a perplexing road-block without any context leading up to it. These types prompt like shit, and refuse to learn about different models or reasoning routing. They hit free-tier AI with a boor thread, and expect it to magically manifest a solution.

If you really want to "test" AI in your work, you need to stop being cagey with it, and provide excessive context. Feed all of the flagged doc review items (I'm not a lawmo so forgive me), relevant statutes, case law, etc. Think of the satirical Saturday morning biglaw threads where a partner is emailing you to start a new project, and at least suggests what areas/cases to focus on. You should be as specific as possible, and treat it as leverage--tell it what to comb through, and/or provide a templated example of a good citation and ask that it find more. Never giving it open-ended tasks without guidance like "analyze and quash this breach of contract" or something similarly retarded. You should not view it as an all-or-nothing replacement or laziness hack; it is best used to extrapolate on good ideas, or sift through shit to find pearls. hth

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49412540)

I will be the very first to replace my employees with AI if it’s possible, but it isn’t there yet and for awhile now hasn’t been getting materially closer.

And while I agree it’s mostly insecure white-collar workers shitting on it *ever* being able to do xyz, the biggest defenders/hypemen for the tech (who aren’t directly financially tied to said hype) are 120-125 IQ state-school retards with laughable overconfidence in their ability to understand varied professional applications, especially given that the live in a 3000-sqft house and send their kids to public school.

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49412570)

Its ability—not necessarily just its present ability, but its ultimate ability—to “do random jobs” is all that actually matters in the end though. Everything else is some pimply retard with grandiose illusions of depth and insight, but who everybody laughs at, preening like a fag about value that requires stupid fucking analogies and metaphors to convey (to a bored listener trying to look over your 5-foot-2-inch-off-the-ground shoulder to catch the eye of someone more attractive and interesting to talk to).

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49412792)

People have been creating predictive models since at least the 1980s. The ability to attain countably infinite training sets is a pretty basic end-of-chapter textbook thought exercise.

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49413005)

https://www.sciencedirect.com/topics/earth-and-planetary-sciences/supervised-classification

That technology has been in use since the 90s. The first time I saw AI in action this was all I could think about.

(http://www.autoadmit.com/thread.php?thread_id=5794931&forum_id=2)#49413006)