Wealth Existential Dread/Hope With AI

[TheEcho]

Hey all!

I'm becoming rapidly consumed by the belief that AI is going to flip EVERYTHING on it's head within 2-3 years. Having an "intellectual" skillset, it's freaky what ChatGPT 3 can do. I always was great at wording and my mom (who has an English degree), had me proofread and improve her work materials. ChatGPT destroys my level of wording. It needs a little tweaking to get the ideal, but it makes me feel like a 5th grader. If you haven't run your LinkedIn/Resume descriptions through it, you should ("make it professional"/"optimize it" worked best for me. Usually professional first, then make it shorter second).

I feel almost terror to put together a business and power it by AI asap (I have a solid idea of the what, just need to learn some how, which surprise! AI can help with). Went to college for real estate and have studied outside of school to develop an advanced Excel skill set for analytics, which is a very intensive activity, and I honestly believe that it will be gone in 1-2 years (because it literally already has startups automating it NOW). My goal is to be on the rising wave, rather than drowning underneath.

ChatGPT 4 will be literally 1000x more "knowledgeable" compared to 3, and I don't think AI has yet entered the advancement of AI in earnest yet. Once that happens, it's over y'all. Not much point to this, I just have to get it off my chest and hopefully warn others to become obsessive about finding a way to leverage AI in the coming year to get a slice of the cake. My 2-3 year forecast is intuition based on observing the trends: AI has passed medschool entrance exams, parts of the bar, already seeing startups automating highly intellectual tasks. Capitalism is going to do it's thing and those that leverage AI will demolish those that don't, which is why I think things will change so rapidly. Brokers that can mulch through 1000 properties a day will erase those that take 4 hours to underwrite 1 (what my job would've been). Lawyers that offer services for dirt cheap where they oversee a largely automated system will demolish offices that charge by the hour (and will be more effective than them).

Buckle up y'all...

Here's a niche, but seriously helpful idea: build an AI that evaluates conversation recordings and coaches you on ways to perform better. Maybe even off visuals, though that would be harder to be subtly recording yourself visually vs audibly. Dating coaches, your thing might be the "wearer of the camera" for your students and you help break the recommendations down for them.

Update: Put this in chat, but I had come up with a more descriptive forecast of how I think things will go yesterday on discord, so adding it here now:

"I believe intellectual work will experience 3 phases:

1. AI will empower skilled workers to become monsters in their field. Things will work similarly but market share will shift to those that best use automation, most companies will die due to not adapting soon enough. (try to capture a lot of market share)

2. AI will become mass-market and programs that do entire high-level projects will cost like $2/month, may see a resurgence of smaller shops able to operate under such low barrier to entry. (be ready to pivot with a popular product)

3. ???????? endgame. Pretty much only extreme results at this point. Extreme diversification or extreme accumulation to the top with nothing left for the rest. (hope you got rich earlier)"

 

[Conquistador]

I think this is overblown.
Look, ChatGPT is a breakthrough. But I'm still pretty circumspect about what it can really do in the real world.

Unless AI hallucination and related problems can be solved (and I'm skeptical), the possibilities for real-world applications will be somewhat limited.

AI tends to go through boom and bust cycles, for good reason. The big advances always turn out to be a little less amazing than people think at first.

Also, there's the little problem, common to all conventional machine learning AIs, that no matter what kind of results can produce, no matter what kinds of tests it can pass, ChatGPT and its cohorts don't actually understand what you or they are saying, in the way that humans understand (or think they do). This is the main reason why it can produce extremely convincing hallucinations ("ChatGPT, give me a summary of the regimental history of the Ruritanian 14th Grenadier Regiment"). All it is is extremely advanced pattern matching.

 

[rr2021 aka DEVENCI]

I think this is overblown.

its way under blown.

i work in this field.

back in 2018 i managed a team of like 10-15 people.

1 guy stood out so i told him i would give him a big raise for him to automate all non creative work being done by the team

in less then 1 month he had set up a fully automated process and i fired everyone else on the team, gave him a massive raise and reduced my stress from having to manage 10-15 people to only 1 person.

this was in 2018.

forget chatgpt.

well before then if you know this field then you know machine learning, ai,automation ,etc...

all this stuff is already 100x more advanced then the average human is aware of.

you have no idea how many things in your day to day life are already impacted or fully controlled by these things.

 

[rr2021 aka DEVENCI]

Hey all!

I'm becoming rapidly consumed by the belief that AI is going to flip EVERYTHING on it's head within 2-3 years. Having an "intellectual" skillset, it's freaky what ChatGPT 3 can do. I always was great at wording and my mom (who has an English degree), had me proofread and improve her work materials. ChatGPT destroys my level of wording. It needs a little tweaking to get the ideal, but it makes me feel like a 5th grader. If you haven't run your LinkedIn/Resume descriptions through it, you should ("make it professional"/"optimize it" worked best for me. Usually professional first, then make it shorter second).

I feel almost terror to put together a business and power it by AI asap (I have a solid idea of the what, just need to learn some how, which surprise! AI can help with). Went to college for real estate and have studied outside of school to develop an advanced Excel skill set for analytics, which is a very intensive activity, and I honestly believe that it will be gone in 1-2 years (because it literally already has startups automating it NOW). My goal is to be on the rising wave, rather than drowning underneath.

ChatGPT 4 will be literally 1000x more "knowledgeable" compared to 3, and I don't think AI has yet entered the advancement of AI in earnest yet. Once that happens, it's over y'all. Not much point to this, I just have to get it off my chest and hopefully warn others to become obsessive about finding a way to leverage AI in the coming year to get a slice of the cake. My 2-3 year forecast is intuition based on observing the trends: AI has passed medschool entrance exams, parts of the bar, already seeing startups automating highly intellectual tasks. Capitalism is going to do it's thing and those that leverage AI will demolish those that don't, which is why I think things will change so rapidly. Brokers that can mulch through 1000 properties a day will erase those that take 4 hours to underwrite 1 (what my job would've been). Lawyers that offer services for dirt cheap where they oversee a largely automated system will demolish offices that charge by the hour (and will be more effective than them).

Buckle up y'all...

Here's a niche, but seriously helpful idea: build an AI that evaluates conversation recordings and coaches you on ways to perform better. Maybe even off visuals, though that would be harder to be subtly recording yourself visually vs audibly. Dating coaches, your thing might be the "wearer of the camera" for your students and you help break the recommendations down for them.

you are 100% correct to be worried.

the biggest problem is there is no oversight in place.

the government is run by old people who lack the understanding to put systems in place to protect the general public from the world ending real dangers AI could pose.

remember the early days of the internet with things like limewire and piratebay how it took years for the government to catch up and protect copyright owners from having their work stolen.

the problem with AI is that we cannot afford the slow process the government takes on these things.

never forget that car seatbelts where around for 10+ years before the government forced car makers to use them as a public safety policy.

10 years with AI is like 1000 years; the world could end by then.

there needs to be clear oversight in place.

the same way someone cannot just go start making nuclear weapons, there should be clear rules in place as to who and what can be done with AI.

i believe it's as dangerous if not more dangerous then nukes..

 

[TheEcho]

All it is is extremely advanced pattern matching.

That's most intellectual work. A handful of people will replace entire fields leading up to full automation. This is not a maybe it will do these things well, it already does.

rr nailed it with a startling example.

My goal is to get as rich as quickly as possible before things are totally upended.

 

[Conquistador]

all this stuff is already 100x more advanced then the average human is aware of.

Exaggeration. Besides, the average human lives in a developing country.

you have no idea how many things in your day to day life are already impacted or fully controlled by these things.

Sort of true, but phrased way too sensationally.

That's most intellectual work.

What do you mean by that term?

A handful of people will replace entire fields leading up to full automation.

Which fields?

When I was a kid, everyone said that by 2023 things would be way more futuristic than they turned out to be.

Sure, a whole lot of office type jobs will get cut. So what?

And about automation...full technology shift takes quite a long time. Older systems tend to be more cost-effective, and where they're good enough, there's no point replacing them. Whether it's Neolithic plows in 1970s Greece, wheelbarrows in our garden, or Windows XP at my uncle's small factory, tech tends to layer over older generations rather than replace it outright.

This is not a maybe it will do these things well, it already does.

Which things exactly?

My goal is to get as rich as quickly as possible

Hope u have a plan then

before things are totally upended.

It'll be way less dramatic than you imagine.

 

[TheEcho]

Which things exactly?

Hope u have a plan then

It'll be way less dramatic than you imagine.

Like analyzing large documents for key metrics, plugging them into Excel, scraping the internet to find 100 comparables and analyzing them in minutes, then producing a recommendation of it all. This happens now. Sure you want an analyst to greenlight it, but AI just let 1 analyst be as effective as a team.

 

[rr2021 aka DEVENCI]

Older systems tend to be more cost-effective, and where they're good enough, there's no point replacing them.

lol.

wtf.

older systems more cost effective in the world of automation.machine learning, etc...

hahahaha

this is so hilariously incorrect that i cannot believe any human adult would say such a thing.

it's literally one of the biggest concerns with all this.

so many things can already be done so cheaply with automation that mass job losses are in the very near future.

right now there automation tools that allow one human to do things that before would require an entire floor of developers/programmers.

this is NOT the future; this is now.

google,amazon and facebook have all fired 10 000+ people each in recent months and a lot more is to come.

what planet do you live on where older systems are more cost effective.

hahahha one of the funniest things i have seen all week.

 

[Conquistador]

Like analyzing large documents for key metrics, plugging them into Excel, scraping the internet to find 100 comparables and analyzing them in minutes, then producing a recommendation of it all. This happens now. Sure you want an analyst to greenlight it, but AI just let 1 analyst be as effective as a team.

In how many fields is that example useful tho?

 

[ulrich]

ChatGPT and any artificial intelligence is statistical in nature.
It can only create something by analyzing thousands of inputs and averaging them.

What does that mean?… it means that it can never create state of the art or highly expert work because it has not enough input to analyze.

Take for example social media content… millions of people are churning input for the machine to analyze everyday… it can easily see the pattern.

Then think of business advice , for example, where common knowledge tends to be wrong… ChatGPT spits useless generic advice because 99% of the content out there is like that.

The same goes for dating, ask ChatGPT anything about dating and you get watered down feminist influenced mainstream advice.
Any advice in this site is 10X better and maybe ChatGPT can read it but its input get lost along the thousand other vanilla sites it reads.

The technology is impressive, but always keep in mind that you are just getting an average of what is out there.
The more input the AI can get, the better responses it can deliver.

It will never replace a true expert for it has no way to identify him.
And some fields just don’t have enough quality info for the tool to see the pattern.

 

[Skjöldr]

Also "Artificial intelligence" is a meme. It has all been coded and written by humans. The "algorithms" or whatever. I'm no expert, but it's basically just written by humans.

 

[mist]

when i get my ai big titty goth girlfriend i'm not sharing it with any of you nonbelievers

 

[TheEcho]

I don't get how people read what rr had 1 guy do, and with how I'm about to flatline the basis of thousands of jobs (it's happening anyways, so I might as well get a cut) and think things are going to be fine. The steps I need to take are very straight forward, with only a couple sticking points (the avenue is obvious, the solution may take a day or 2 of research).

 

[Train]

Just to give another reference point, my industry is FAR from being fully automated. AI is essentially a gimmick in it. I don't see AI fully automating my industry in the next five years. Way too many inefficiencies that can't be reliably automated (yet).

 

[TheEcho]

Keep in mind, the more effective (profitable) AI proves to be in other industries, the more energy will flow into automating the "hard cases" because there's huge money there.

It's funny, I automated RuneScape skilling through "advanced" macros for years because I enjoyed the challenge of making an ironclad cycle for a series of tasks. Of course the simple crafting skills where you sit at a bank, but also things like hunting and doing rune runs, with zero allowance for it to get off track and several safeguards to correct deviation. On there, I got several accounts banned for it, but in real life, it may be my key to big income.

 

[rr2021 aka DEVENCI]

holy shiet.

i cringe so hard reading this thread.

why do so many people feel the need to contribute their "opinion" to a subject they obviously know nothing about.

It will never replace a true expert for it has no way to identify him.

omg dude.

please stop.

what planet do you live on?

are people really this clueless?

AI can already pass medical licensing exams and the bar exam.

the best chess and go player in the world has been an AI for years.

ai cannot replace a true expert?

lol what?

hahahahhaha

omg.

 

[Train]

holy shiet.

i cringe so hard reading this thread.

why do so many people feel the need to contribute their "opinion" to a subject they obviously know nothing about.

My brother, cringe is you contributing your opinion about game while having a 0.1% approach-to-lay ratio

 

[rr2021 aka DEVENCI]

Just to give another reference point, my industry is FAR from being fully automated. AI is essentially a gimmick in it. I don't see AI fully automating my industry in the next five years. Way too many inefficiencies that can't be reliably automated (yet).

lol.

this is hilariously and incredibly wrong.

i cannot believe some of the stuff being posted in this thread.

am i in the twilight zone?

this would be like someone in 1997 saying "my industry will not be impacted by the internet in the next 5 years"

it's like that but 100x worse/more inaccurate

i really hope the average human is not as misinformed or this subject as most people who have posted on this thread seem to be

if the average human is this clueless on the subject then the human race is probably in big trouble

 

[rr2021 aka DEVENCI]

My brother, cringe is you contributing your opinion about game while having a 0.1% approach-to-lay ratio

please quote a post where i give my opinion on game.

my game related posts are almost always me asking a question or saying "this program helped me.

as far as me directly telling someone what to do, i guarantee you will not find a single post like that.

if im wrong then quote the post for all to see.

also, i have an "approached to lay ratio" NOT approach to lay ratio

girls approach me.

 

[ulrich]

I don't get how people read what rr had 1 guy do, and with how I'm about to flatline the basis of thousands of jobs (it's happening anyways, so I might as well get a cut) and think things are going to be fine. The steps I need to take are very straight forward, with only a couple sticking points (the avenue is obvious, the solution may take a day or 2 of research).

Well, that has been said a thousand of times before…

Who knows? Maybe this time, it is true… I’m just placing my bet on “it will turn out not as impressive as everyone thinks”