After messing around with ChatGPT...

[padlock]

I believe AI tech is going to usher in unpreceded productivity growth. I feel the same way I did when I first found out about the internet back in 1992.

What's the best way to invest in this area? I heard Microsoft has invested heavily in OpenAI. Would they be my best bet?

 

[MichaelB]

I believe AI tech is going to usher in unpreceded productivity growth. I feel the same way I did when I first found out about the internet back in 1992.

What's the best way to invest in this area? I heard Microsoft has invested heavily in OpenAI. Would they be my best bet?

OpenAI is set up as a non-profit, so it’s unlikely to list directly, at least for now. It looks like Microsoft will have privileged access, so that’s probably the best way to go.

Another option is to look for a competitive product. Alphabet and Meta will probably want to have their own AI engines.

 

[CaptTom]

I'm going to sit this one out for a while. See my post in the other ChatGPT thread. Right now I'd call it a solution in search of a problem. Admittedly, a very powerful solution, and there will be applications. Just not sure yet what they might be. I can tell you it's not great at many of the things which have been touted.

 

[Lsbcal]

I believe AI tech is going to usher in unpreceded productivity growth. I feel the same way I did when I first found out about the internet back in 1992.

What's the best way to invest in this area? I heard Microsoft has invested heavily in OpenAI. Would they be my best bet?

What did you do on ChatGPT that impressed you?

 

[padlock]

What did you do on ChatGPT that impressed you?

I got it to write several reasonably non-trivial sample software programs involving both Python code and SQL queries.

As a former software developer, I can easily see practical applications for it in the field.

 

[Lsbcal]

I got it to write several reasonably non-trivial sample software programs involving both Python code and SQL queries.

As a former software developer, I can easily see practical applications for it in the field.

Sounds like a nice application but it is very specialized. Not akin to a general purpose thing like the internet which you mentioned in the OP.

I would like to know about more use cases before I got excited enough to invest in it. Tech is fascinating but it is hard to pick pure case winners. It probably fits nicely into large company needs but will it move the needle enough?

For example, with vaccine development there was Pfizer which did OK but Moderna skyrocketed in 2020-21. Long term, who knows.

 

[Fermion]

I, for one, welcome our chatbot overlord

 

[padlock]

Sounds like a nice application but it is very specialized. Not akin to a general purpose thing like the internet which you mentioned in the OP.

I would like to know about more use cases before I got excited enough to invest in it. Tech is fascinating but it is hard to pick pure case winners. It probably fits nicely into large company needs but will it move the needle enough?

For example, with vaccine development there was Pfizer which did OK but Moderna skyrocketed in 2020-21. Long term, who knows.

I've also heard from colleagues who used it to help write blog posts and some marketing material. In any case, it certainly isn't my intent to try to convince anyone that it's going to be revolutionary tech. I believe it is and was just wondering what the best way to gain some investment exposure in it might be.

 

[Lsbcal]

I've also heard from colleagues who used it to help write blog posts and some marketing material. In any case, it certainly isn't my intent to try to convince anyone that it's going to be revolutionary tech. I believe it is and was just wondering what the best way to gain some investment exposure in it might be.

Padlock, I'm just curious and hence the questions. Thanks.

 

[padlock]

Padlock, I'm just curious and hence the questions. Thanks.

No worries Lsbcal. Just trying to clarify my motivation for the post.

 

[MichaelB]

<mod note> This is a thread about investing in ChatGPT. Posts about using the tool itself were moved to an ongoing discussion about that, here https://www.early-retirement.org/forums/f54/chatgpt-116205.html

 

[gromit]

OpenAI is set up as a non-profit, so it’s unlikely to list directly, at least for now. It looks like Microsoft will have privileged access, so that’s probably the best way to go.

Another option is to look for a competitive product. Alphabet and Meta will probably want to have their own AI engines.

+1. The first to market, or even the first to have a large share of a new market is not necessarily the long term winner. For those of us old enough to remember, Wordperfect was a word processing application that at one point had a 90%+ share of the US market. Then along came Microsoft Word.

 

[GravitySucks]

+1. The first to market, or even the first to have a large share of a new market is not necessarily the long term winner. For those of us old enough to remember, Wordperfect was a word processing application that at one point had a 90%+ share of the US market. Then along came Microsoft Word.

Old enough to remember WordPerfect? You to young for ElectricPencil?
Sorry, gotta get back to writing my Lotus123 macros

 

[MichaelB]

+1. The first to market, or even the first to have a large share of a new market is not necessarily the long term winner. For those of us old enough to remember, Wordperfect was a word processing application that at one point had a 90%+ share of the US market. Then along came Microsoft Word.

Good example. VisiCalc and Lotus123 are other similar examples. Google has an event Feb 8 and appears to be preparing its own AI ChatBot announcement.

 

[Masquernom]

Visicalc and Lotus123 are interesting examples. Visicalc was developed by a college professor. After developing the code, he went to a patent attorney and asked to get a patent on the software. The attorney correctly informed him that software could not be patented. The professor then release the spreadsheet software unprotected. That allowed Lotus and eventually Microsoft and others to issue their own version of spreadsheet software without paying royalties.

What the original patent lawyer failed to inform the professor is that while software can not be patented, it CAN be copywrited. That was a very expensive oversight for the professor.

The professor originally came up with the idea based on stories from students returning from co-ops with a particular company. That company had a long room with blackboards all down the wall. The co-ops were charged with collecting financial information about the company and entering the info into squares on the blackboards. Each square included calculations or directions with what to do the information.

 

[copyright1997reloaded]

Visicalc and Lotus123 are interesting examples. Visicalc was developed by a college professor. After developing the code, he went to a patent attorney and asked to get a patent on the software. The attorney correctly informed him that software could not be patented. The professor then release the spreadsheet software unprotected. That allowed Lotus and eventually Microsoft and others to issue their own version of spreadsheet software without paying royalties.

What the original patent lawyer failed to inform the professor is that while software can not be patented, it CAN be copywrited. That was a very expensive oversight for the professor.

The professor originally came up with the idea based on stories from students returning from co-ops with a particular company. That company had a long room with blackboards all down the wall. The co-ops were charged with collecting financial information about the company and entering the info into squares on the blackboards. Each square included calculations or directions with what to do the information.

While that ("Software cannot be patented") was generally true in 1978/79, software was patented prior to Visicalc coming into existence. The 1981 "Diamond v. Diehr" case opened up more territory for software patents. While mathematical formulas could not be patented, novel uses could be patented.

I remember those days (somewhat) well as a young pup as I am co-author of one of the early software patents (filed in 1983 on work already done - we had to rush to get it written up given it was ready to be released because once the product was commercially available it would no longer be patentable.)

p.s. I use VisiCalc as an example in an intro to CS class as an example of a single application that essentially "made" a platform successful (i.e. a "killer" app).

 

[NanoSour]

I believe AI tech is going to usher in unpreceded productivity growth. I feel the same way I did when I first found out about the internet back in 1992.

What's the best way to invest in this area? I heard Microsoft has invested heavily in OpenAI. Would they be my best bet?

Maybe you could just ask ChatGPT what the best way to invest in its technology. I would what be curious what it recommends.

 

[JoeWras]

"Dear ChatGPT. Is Google going to smoke you?"

I don't have an account. I can't try.

EDIT: got an account, got my answer. In short: no, we don't compete.

I got an interesting answer on another query which I'll take to the non-investing thread.

As a language model created by OpenAI, I don't compete with companies like Google. OpenAI and Google have different goals and approach AI research and development in different ways. While Google is a leading technology company that offers a variety of products and services, OpenAI is a research organization focused on advancing AI in a responsible and safe manner. Both organizations play important roles in shaping the future of AI and are working towards their own unique objectives.

 

[Alex The Great]

The problem is any AI heavily depend on electrical grid when going through training. While it need so much power, human accomplish may be smaller task but with much less resources. I do see it as a major limiting factor.

 

[mickeyd]

The problem is that there are no requirements that we 78 yos write a term paper.

 

[aja8888]

The problem is that there are no requirements that we 78 yos write a term paper.

You got that right! At 79, all I need to do is feed the dog and take out the trash. I don't need Chat GPT for any of that.

 

[Time2]

I, for one, welcome our chatbot overlord

I, for one, welcome our chatbot overlord /s/
Fixed that for, me!

 

[MichaelB]

You got that right! At 79, all I need to do is feed the dog and take out the trash. I don't need Chat GPT for any of that.

Maybe ChatGPT will decipher and translate your dogs barking and efforts to communicate with you. “aja8888, I don’t like that stuff you serve me. The food you eat smells so much better. Can I have that instead ? ”

 

[aja8888]

Maybe ChatGPT will decipher and translate your dogs barking and efforts to communicate with you. “aja8888, I don’t like that stuff you serve me. The food you eat smells so much better. Can I have that instead ? ”

Yes, another good reason to keep that Bot away from here!

 

[Christine]

I believe AI tech is going to usher in unpreceded productivity growth. I feel the same way I did when I first found out about the internet back in 1992.

What's the best way to invest in this area? I heard Microsoft has invested heavily in OpenAI. Would they be my best bet?

Way too early to know which would be the best bet.


Personally I am invested in Tesla and they are working on self driving cars and humanoid robots. And on creating the next generation of computing platforms (dojo) needed to train all these AI systems.

Having future server parks with advanced AI computers you can let customers rent ala Amazon AWS and Microsoft Azure do with general old fashioned computing now might be one way to invest here.