AI is in its 1995/1996 stage, and it has the benefit of recursive self-improvement.

in 30 years, things will be pretty weird.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613260)

One thing that would be awesome is for IG to use "AI" to develop a usable search function. The next frontier?

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613550)

That's really amazing, but no one is going to watch them.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613301)

Edit: Wait, we kinda already were weren't we? Weren't you helping ZZZ back then? I guess that was before "AI," but I kinda doubt it would've helped you guys get a better outcome.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613524)

That would be fun, though

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613528)

Smirking Boomer weighing in on the potential of "Amazon" and "e-commerce" (c. 2002): "I mean sure, I guess maybe someday, kids will want to buy books on The World Wide Web and have them shipped to them from some warehouse half a world away, but let's be real... this isn't exactly gonna change the world most of us live in, ya know what I mean kid? I'm pretty sure 'bookstores' and 'shopping malls' aren't going anywhere, bud..."

lol it's amazing how smug the new iteration of you guys remains after you respawn into every new generation. BTW, the social change that is coming with AI in the next 25 years is going to be 100000x as profound as the social change that has been ushered in by the rise of near-universal internet usage and the shift to e-commerce over traditional brick and mortar in the last 25 years.

XOXO, HTH!

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49614993)

So, we shall see. But I certainly won't be investing in OpenAI any time soon. Even though they are privately owned, I think I read that they have brought in more private investment dollars than Saudi Aramco did in their IPO. That seems completely insane for a company that has just been a gigantic cash furnace so far.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615023)

"And in both cases, it led to a boost in productivity, but not to the point that it led to bonkers economic growth and solved all of society's problems."

Can you name a single prior tech revolution -- industrial revolution, internet, whatever -- that was fundamentally predicated on the tech in question actually replacing human judgment (and thereby undercutting the fundamental value proposition that undergirds all human work at the most basic level)? Not just increasing speed or efficiency (i.e. most industrial revolution tech) or reducing the complexity or the amount of physical or intellectual work that goes into completing tasks from the human perspective (i.e., computers can crunch numbers so humans don't have to), but actually replacing the need for human-directed and judgment-oriented logical reasoning to give meaningful direction to those time/energy-saving efforts in the first place?

I can't. It has literally never happened before at any sort of meaningful scale.

And that's why AI is, unironically, fundamentally different from every other tech revolution that has ever occurred in human history.

The fact that it's still early days and the tech itself is still far from being completely built out doesn't change that fact, and none of the people who seem hellbent on downplaying the scope of this social shift ever seem to be willing to *directly* confront that point and its implications for why this really is a unique inflection point in human history. Odd case.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615059)

"Uhhh, not sure if you're aware of this, chud, but this brand new technology ackkkkkshually ISN'T perfect yet!"

Wow! Shocking! Why didn't I think of that?! Nevermind, my whole assessment is wrong I guess.

Fucking retard.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49616801)

Thankfully, I think AI is way behind on reasoning - to where i don't think it will ever replace doctors/lawyers/accountants/etc.

Last week i used it to check a document, and it instantly gave out better analysis than i could and attributed the differences between two statements based on this isoteric county regulation. The problem was, the real difference between the two statements was one of the vendors increased their bid.

When I ask AI about it, it's just like "you're right." And when i ask why it didn't catch that, its just like "you're right i should have caught that."

I'm not sure how you can ever replace its "thinking" because it isn't thinking - its just guessing words, right?

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613321)

Fun fact: if you were relying on a lawyer fresh out of law school the exact same thing would have happened.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613378)

Someone at work got a ticket on MLK day for speeding in a school zone. He isn't a lawyer but he uploads it to Gemini and it immediately notes that school is closed on MLK day, that he should contest it and cites the statutory language he needs.

It is impressive as shit and a complete gamechanger.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613363)

Menial labor is absolutely dying. At my megacorps a decade ago every lawyer had an EA, even if they're at the bottom band. Several years ago, they eliminated that. You had to at least be director band to have an EA. Scheduling meetings and such used to be a big hassle but with Google Calendar it is super easy, so they just RIFed a bunch of them. Now they eliminated EAs for directors too. Only the most senior lawyers have EAs now. Now imagine you're a 50-something EA that just lost her job. What job do you imagine these ladies are capable of getting? I think most of them will never have full time jobs again.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613398)

1. You have no idea whether the car it told you to buy is going to be good for you. You're just following what it told you to do.

2. If your friend was so retarded that he was unaware that Monday was MLK Day, he should be fined for that alone.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613532)

Then I used it to figure out which service / warranty add-ons to reject at the sales manager office. Then scrutinize the final docs. It caught duplicate line items of 1 $699 add-on from a 20 page pdf with mostly boilerplate and signature lines. This was 2 random SKUs spread out over like 60 rows of tiny font. I made them remove before singing of course

I bet most people don't even read the docs before leaving with the car, after taking 5 hours to negotiate a worse deal!

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613656)

"Hi Cove, here's a picture of the used car I want to buy. Does this look mechanically sound?"

"You're doing great, and I'm sure this will be the used car of your dreams!"

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49613690)

https://i.imgur.com/0xFKEpj.jpeg

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615006)

It's all of them. You reject all of them every time. HTH.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615548)

Crazy that you think this is some devastating rejoinder when the vast majority of people in America who are or will soon be using AI are, by your own metric, literally too stupid to even evaluate whether these counterpoints are valid or not according to your own self-described parameters.

Gee, I wonder what will serve that huge underclass of retarded people *better* (albeit imperfectly in both cases) moving forward:

1. Admittedly imperfect AI that sometimes leads them astray but is steadily improving over time; or

2. Their own retarded brains which they would otherwise have to rely on to figure shit out for themselves, which incidentally are never going to get any better at processing information?

It's almost like myopically assessing the merits and macro-level social usefulness of evolving tech like FSD or AI in some theoretical vacuum that doesn't exist ("Uh, ackkkkshually, this tech is still FAR from perfect, bud") instead of assessing it against the benchmark of the real-life alternatives that currently exist (most drivers are pretty shitty and easily distracted + most people are kind of retarded and struggle to think clearly) is the REAL hallmark of a smug midwit intellect who should probably just shut the fuck up.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615012)

Nope, sorry you're too dumb to parse a few paragraphs.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49616803)

My objection to the AI is fanbois is that they seem to want to think that AI is going to replace all human labor, and that is simply moronic. If that is even possible, it is decades away.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49617070)

Most (well, many) xoers are at a stage in their careers where it’s clear they’re going to be safer from being automated away to the middle class than the vast majority of other ppl. Maybe no one’s safe, but I’ll be among the last replaced because I’m essentially a politician/lobbyist who uses being a trial lawyer to keep my street cred up. Will AI becomes able to do what Chief Justice Roberts *does*, as in write opinions? Yeah. You think Roberts is worried about being replaced?

Second, AI’s ultimate usefulness will be—hold on here!—its actual real-world usefulness in specific domains. The specific domains where we work and have expertise—which btw is the only domain where AI’s encroachment would have any reason to create the insecurity you fags love to poast about. (Why would a lawyer be sweating AI’s coding talent? We not only can’t evaluate that talent since we don’t know anything about coding, but coding acumen isn’t going to directly help (my perspective) or threaten (yours) us if what we do is law.) Most of us are actively using AI on a daily basis both for its current utility and as a long-term investment in its future potential—both for helping us in a Westlaw-like, non-labor-replacing way, and, yes, for replacing labor. Conversations about where AI is currently FOR REAL-LIFE LAW USAGE are both more important, and more dynamic/constantly changing than rehashing the same retarded conversation about some speculative intermediate future where we have ITE-like job insecurity.

**The exception is if one of the Eliezer Yudkowsky-like doomsday theories comes true and AI becomes, effectively, our enemy—a probability I’m not qualified to evaluate, but can tell you that Yudkowsky is very poor at persuading people of (although he may just have low verbal intelligence). This tends not to be the fears that xo is susceptible to / likes to play on, because I think most poasters would rather the world end and everyone die than see themselves reduced to the level of the proles they thought they’d left behind after high school.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49614510)

Very short summary

The author argues that fears about AI are ahistorical: like past technologies, AI’s harms and disruptions will be outweighed by its overall benefits. Anxiety about AI mostly reflects fear of personal obsolescence rather than real historical precedent. Technological change is not comparable to immigration or outsourcing.

AI should be judged by its concrete usefulness in specific fields, not speculative future job-loss scenarios. In law, AI already functions mainly as a powerful research and productivity tool and, in some cases, a labor substitute. Most professionals—especially those in judgment- and influence-based roles—are unlikely to be among the first displaced. Only extreme “AI doom” scenarios would justify broad fear, and those are seen as unlikely.

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49615100)

End | Of | Work

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49616578)

Every tech revolution kills some jobs and makes everyone else’s lives better; AI won’t be different. The skeptics are either Boomers mad they have to learn new tricks while their relevance evaporates, or younger folks terrified of sliding back to the middle class they thought they’d escaped. The poster isn’t sweating it because his real job is relationships and persuasion. Trial work is just the credibility engine. Even if AI could write opinions like Roberts, Roberts isn’t getting fired.

Debating whether AI will *eventually* take your job is less useful than figuring out what it can do for you today. He’s already using it daily for research and to replace billable grunt work. The only wildcard is Skynet-style existential risk, which he can’t evaluate but notes that Yudkowsky isn’t exactly packing stadiums with converts. Bottom line: most AI doomers would rather see civilization collapse than admit they’re not as irreplaceable as they thought.​​​​​​​​​​​​​​​​

(http://www.autoadmit.com/thread.php?thread_id=5825604&forum_id=2...id#49616620)