Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Fuzz
Jun 2, 2003

Avatar brought to you by the TG Sanity fund

Kavros posted:

I expect very little good from AI because of what it is being purpose-built towards. it's not going to be here to make our lives easier, it will be here to better extract profit for the benefit of a specific class of investor or owner-of-many-things.

One thing I expect it will be really good at in the meantime is further destroying "community," since so much of community is online and it will be even more numbingly difficult to navigate seas of inauthentic communication and signal-to-noise ratios blown into bits even more efficiently than it was in an age of mere botting

This.


My heart goes out to the graphic designers and small form illustrators who have made their livelihoods as freelancers doing a steady stream of projects to keep food on the table, because their entire profession just got nuked from orbit in under a year and the society is expanding pretty rapidly to devour a lot of those skilled workhorse type art and print production roles.

Corpos are gonna waste as little time as possible to trim the pesky money-wanting people from every facet of their profit treadmill as possible, and it's gonna destroy tens of thousands of people's lives. Their best bet is to get on board and aggressively learn to incorporate AI into their workflow and speed up their production quotas exponentially so they can be the first to stab all their colleagues in the back and be the overseer of the AI that will replace all of them... at least that handful of people won't go hungry.

Adbot
ADBOT LOVES YOU

Fuzz
Jun 2, 2003

Avatar brought to you by the TG Sanity fund

gurragadon posted:

It's like the industry of creative arts is going through what other manufacturing has been going through in a really fast time scale. All those people working on an assembly line are replaced by a machine and somebody to make sure it works. Now the same thing is happening to creative freelancers.

The advancements are coming really fast now though and it's going to hit white collar workers everywhere. Like I posted a bit earlier, GPT-4 can ace the bar exam and you can hook it up to other programs so it can perform accounting practices. Were quickly making most of the population's employment not worth the money. But if it goes beyond creatives, it will start hitting workers who command a pretty strong voice in the economy.

Will lawyers or doctors have a big enough voice when there turn comes? Since its so similar to what happened to assembly lines in the past in my mind, we know what NOT to do with people affected by advancements in technology. A simple example is just watching Roger and Me. Did any societies treat their redundant workers better?

Acing the BAR or even the MCAT or USMLEs is a far cry from actually practicing law or medicine. They're multiple choice exams where the AI has the equivalent of the entirety of the internet at its disposal to look up and eliminate incorrect answer choices.

Standardized testing is a dumb metric for gauging AI proficiency.



In terms of whose fault it was, I mean the tech sector thought it could reinvent banking with crypto and we all are seeing how that's going. The tech sector is literally that meme about wizards, when you look at everything in terms of being a math problem, suddenly everything has a solution... Let's just ignore that reality has a lot of remainders and trailing decimal points, we just shave those (people) off in the name of precision.

Fuzz fucked around with this message at 18:54 on Mar 25, 2023

Fuzz
Jun 2, 2003

Avatar brought to you by the TG Sanity fund

gurragadon posted:

GPT-4 took the Uniform bar exam. According to https://www.ncbex.org/exams/

The MEE is 6 essay style questions analyzing a legal issue and the MPT tasks are standard lawyering tasks.

The AI will always have the entire internet at its disposal, thats a feature of AI, not something that would changed unless delibrate.

What would be a good metric for you? Seeing GPT-4 actually being used in a courtroom would be convincing to me, unfortunately the law practice seems to be kind of reluctant to embrace technology. The supreme court still dosen't have TV cameras.

I kind of wish they would have just let this guy try out the AI lawyer thing.

https://www.cbsnews.com/news/robot-lawyer-wont-argue-court-jail-threats-do-not-pay/

Edit: Uniform Bar Exam not Unified

I was speaking mainly from the medical perspective, as I've taken and passed all those medical exams too.

They don't actually replace the clinical training at all, and all medical students and residents take those exams but you throw them in a solid 85+% of cases and unless it's super bread and butter stuff they won't actually diagnose the problem correctly, unless you as a patient enjoy the House style trial and error method of medical treatment instead of actually examining thoroughly and running focused tests and then treating the actual issue based on your results instead of jumping to a conclusion based on incomplete data because illiciting all that info from a patient and asking the right questions is like 75% of what experience teaches you.

Could it get there eventually? Who knows, anything is possible, but not in the next few years and there's a big gap between knowing the book and actually being able to take care of people. The same can be said of Nurse Practicioners, since they similarly lack a large amount of clinical experience and training before they're handed a prescription pad and told to just start seeing patients and figure it out as they go..

Fuzz fucked around with this message at 04:40 on Mar 26, 2023

Fuzz
Jun 2, 2003

Avatar brought to you by the TG Sanity fund

gurragadon posted:

I didn't mean to diminish the value of practical training, it's really important to put knowledge to use and that's when it's really reinforced. Earlier upthread Ohthehugemanatee said they were able to use the Bing chatbot to help diagnose a problem that doctors were missing or couldn't figure out. Do you think it has value in helping doctors diagnose patients? Practical knowledge is really important but you also do lose some of the book knowledge the further you get away from it and the more time you spend working in a field. Maybe just help as a refresher for doctors or nurses about medical issues?

I think it's real value in the medical field would be ways that it can lighten the burden on people.

Definitely could be useful in alleviating/eliminating the need for midlevel intake type stuff. Rather than have a physician with an advanced practice nurse or physician assistant backing them up to help with all the fiddly running around and screening results and double checking and following up stuff, an AI would be able to do a lot of that sort of grunt work and help formulate written records on the fly that are concise but still cohesive and directly centered around the patient rather than a form note with some blanks filled in because the humans don't physically have enough time in the day to get it all done.

I'd actually worry most for pharmacists. Pharmacy has been expanding its role clinically for decades as a liaison to physicians to help them with dosing (always been their role), learning about new drugs, knowing the normograms and resistance rates for things like bacterial resistances in the local population, etc. Literally all of those tasks could ultimately be pretty easily done via an AI, which is scary for both the patients and the pharmacists, but honestly if someone can get a system up and running and show that it makes fewer errors than humans in the same role, you can bet the pharmaceutical companies will push for it to cut costs and eliminate the need to get regional pharmacists "on board" with their new drug when some engi can just do a push update and fiddle with some weights and bam, everyone is getting recommended their new crazy expensive non-formulary drug that's only marginally better (according to internal studies) than the standard treatment.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply