It's beginning to look a lot like Christmas. 1999.

Prism

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

Linton

Prism said:

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

Yes but the real benefits of new tech generally go to the users, not the original developers.  Most of those go bust or are taken over early in their life.  Very few are still in existance when the technology matures. Consider Commodore, DEC, Sun, HP and the now forgotten heros of the late 1990's tech boom.

NVIDIA at the moment is a niche monopoly supplier who were lucky because their high speed graphics parallel processing hardware happened to be a very good fit for AI.  It wont take long for others to catch up.  In 5 years time the market leaders could be Chinese.

In any case will AI really be a significant part of the world economy or simply another business tool?  It may have massive effects on global society but that does not guarantee massive  profits compared with more traditional sectors once the technology becomes routine.

itwasntme001

Prism said:

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

I understand it will help solve problems in some areas and improve efficiency.  But how big can it actually get?

So far, for the consumer, its about generative AI, which TBH is pretty impressive.  But I can't imagine this being anywhere close to being an "iPhone" moment.

I think the real benefits are in cost reduction or maintaining margins for big tech, which is why they have been buying these chips.  Many of these companies have huge amounts of data to be processed.  They need these chips to give better insights, so that they can maintain their customer base.  It is not necessarily going to improve profitability, except perhaps for Microsoft.

Then there will the edge cases where AI helps with certain tasks but I doubt wide scale improvement in productivity and efficiency resulting in an economic boom.

But the above is still speculation on my part, so what am I missing?

Arthurian

As we see in the Post Office scandal, (and I wouldn't be surprised at HMRC,) many large corporations and government departments are run on software which is built on old software which is built on even older software, with spaghetti-style work-arounds all the way along. Everyone is too busy batting away problems at work on a day to day basis to be able to sift through coding from the 1970s. I hope AI could quickly run through these systems and not only correct bad mistakes, but streamline whole systems to make companies and departments much more efficient. 

Hoenir

itwasntme001 said:

Prism said:

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

I think the real benefits are in cost reduction or maintaining margins for big tech, which is why they have been buying these chips.  Many of these companies have huge amounts of data to be processed.  They need these chips to give better insights, so that they can maintain their customer base.  It is not necessarily going to improve profitability, except perhaps for Microsoft.

"Tech" companies previously were capital light. Microsoft software for example once created could be rolled out endlessly at high margin. ( Move the IP to Costa Rica and hey presto. Minimal tax to pay as well ! ). Now that's changing. Capital expenditure is running into many billions. To maintain and enhance EPS the companies are changing their accounting policies for the depreciation of "computer" equipment quite aggressively.  Extending the economic life. As is always the case Finance Directors love to paint pictures. The market for the products they create needs to be there or future financial results are going to disappoint. Microsoft itself has gone from extremely conservative to the most aggressive. 

ZeroSum

https://m.youtube.com/watch?v=qLDMv2POHiE&pp=ygUMcGVuc2lvbmNyYWZ0

Prism

Linton said:

Prism said:

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

Yes but the real benefits of new tech generally go to the users, not the original developers.  Most of those go bust or are taken over early in their life.  Very few are still in existance when the technology matures. Consider Commodore, DEC, Sun, HP and the now forgotten heros of the late 1990's tech boom.

NVIDIA at the moment is a niche monopoly supplier who were lucky because their high speed graphics parallel processing hardware happened to be a very good fit for AI.  It wont take long for others to catch up.  In 5 years time the market leaders could be Chinese.

In any case will AI really be a significant part of the world economy or simply another business tool?  It may have massive effects on global society but that does not guarantee massive  profits compared with more traditional sectors once the technology becomes routine.

I meant in a more general sense than an investing opportunity as such. Which is probably a good thing, as aside from a few surges here and there in value of companies like NVIDIA, I would prefer to see a slow and steady business as normal kind of progression.

No idea who or if there will be company that comes out of nowhere to take this by the scruff of the neck. I feel at the moment that the existing big tech companies have got so much spare cash and their fingers on the pulse of change that its likely to still be them. Disrupting an incumbent is only possible when they themselves are not the disrupter. 

Prism

itwasntme001 said:

Prism said:

It doesn't worry me at the moment and I am beginning to believe that this new surge in AI may really be the next big thing. It seems to solve problems and at the same time is making actual money for those companies involved. And don't personally hold any NVIDIA in any of my funds.

I understand it will help solve problems in some areas and improve efficiency.  But how big can it actually get?

So far, for the consumer, its about generative AI, which TBH is pretty impressive.  But I can't imagine this being anywhere close to being an "iPhone" moment.

I think the real benefits are in cost reduction or maintaining margins for big tech, which is why they have been buying these chips.  Many of these companies have huge amounts of data to be processed.  They need these chips to give better insights, so that they can maintain their customer base.  It is not necessarily going to improve profitability, except perhaps for Microsoft.

Then there will the edge cases where AI helps with certain tasks but I doubt wide scale improvement in productivity and efficiency resulting in an economic boom.

But the above is still speculation on my part, so what am I missing?

To me it feels a little like the start of the internet than the iPhone moment. With the iPhone, it was an example of Apple taking an existing design (Blackberries and HP mainly) and making it much better. They kept tight hold of those improvements and there wasn't much sharing going on. 

This feels like everyone is participting. The core ideas are open source and pretty much anyone can have a go at developing one of these LLM based generative AIs. At work we have fed a chunk of our materials into a custom AI engine based on ChatGPT and turned it into a very decent Q&A system. Something that would have taken so much time before, and been a bit rubbish and out of date, has become very easy. Fundamentally it is no different than 'search' but much easier to use. 

We have been playing around with Microsoft CoPilot and even in its early form it can do things in seconds that would have taken much longer. I get it do write a fair bit of my content nowadays, reducing hours into seconds. It can also do things that would otherwise have been impossible for an average person. Reverse engineering a piece of malware to determine its effect or something as simple as generating a useful piece of art.

As for the future, who knows, but I think there is a place for all of this in the world of VR. The latest Open AI demos of 3D content generation, simply from text, are pretty amazing. One of the issues in imagining VR is where does all of the content come from. Lets say you want a training simulator on some form of commercial plane, then somebody previously would have to build that from the spec data. An AI could do it on demand. VR worlds could be full of auto generated content so that simulations/games/hanging out was different and unique each time.

m_c_s

Apple bought out 32 different AI startups during the 2023. Google snapped up 21, Meta 18 and Microsoft 17. These are the acquisitions that we know about. Apple in particular often makes quiet acquisitions without any announcement. CEO Tim Cook said back in 2021 that the company bought a startup every 3 to 4 weeks on average, and in 2022 he said it was every 2 to 3 weeks so they seem to be accelerating acquisitions.

Hoenir

m_c_s said:

Apple bought out 32 different AI startups during the 2023. Google snapped up 21, Meta 18 and Microsoft 17. 

How many of these concepts will ever see the light of day as being commercially viable ?  What have Apple and Google spent on driving self driving vehicles so far? With the days of cheap equity and low borrowing costs over. Burning cash for no return isn't such an attractive business proposition.