If the AI bubble like the Internet, in what year are we now?

[james22]

Early, late?

 

Not a bubble?

 

Not comparable?

[brobro777]

I've read this forum for a while and got good ideas (St. Joe thread) so I figure I would contribute

 

I do not think this AI thing will not go much further from here. In fact I think the upward run in tech in general that began in 2003 was finished as of last year and this AI thing will not lead to the tech sector making new highs

 

I will probably turn out to be completely wrong and NVDA will hit $1 Tril in market cap in 2 months 

[james22]

I believe we are still pretty early.

 

But I found it easier to believe that than to actually shift the 10% I did recently into tech (mostly because optimism looks dumb).

 

Re-reading a couple articles I've saved over time helped.

 

Perhaps a couple broader lessons I've learned: first, to try to do a better job of overcoming my natural aversion to ideas that are popular. One of the mental blocks for me was - "if everyone loves this so much, how can it still be a great value?" Going forward, I'm trying to do a better job of focusing more on an idea's merits and tuning out who likes or doesn't like it.

 

Second, and similarly, I've realized that it's internally inconsistent to A) believe that I have no edge in interpreting near-term price action, and B) prefer buying securities that have fallen over the near to medium-term past than securities that have skyrocketed. As with the previous item, I'm now trying to focus more on the "signal" - i.e. the idea's fundamental merits - and less on the "noise" - i.e. whether it's recently up, down, or sideways.

 

https://seekingalpha.com/article/4229722-small-cap-bear-next

 

Winning in the investing game isn’t simply about having true prior beliefs about the world. It’s also about efficiently updating those beliefs in response to feedback from reality. 

 

https://www.philosophicaleconomics.com/2017/09/pmbtdo/

 

We could be in the midst of the first fear-based investment bubble in American history, with the masses buying in not out of avarice, but from a mentality of abject terror. Robots, software and automation, owned by Capital, are notching new victories over Labor at an ever accelerating rate. It’s gone parabolic in recent years – every industry, every region of the country, and all over the world. It’s thrilling to be a part of if you’re an owner of the robots, the software and the automation. If you’re a part of the capital side of that equation.

If you’re on the other side, however – the losing side – it’s a horror movie in slow motion.

The only way out? Invest in your own destruction. In this context, the FANG stocks are not a gimmick or a fad, they’re a f***ing life raft.

 

Market commentators rhetorically ask aloud what multiple should investors pay to own the technology giants. That’s the wrong question when people feel like they’re drowning.

 

What multiple would you pay to survive? Grab a raft. 

 

https://thereformedbroker.com/2017/10/16/just-own-the-damn-robots/

 

There's a very good reason why growth is killing value, because reversion of the mean is not working for a lot of value industries. That's a classic value technique: You buy a retailer or another cyclical stock that's down on its luck and figure things will turn around. Historically, that's how it worked. But, of course, if Amazon or Google is eating your lunch, that's not how it's working today, and that money just might not be coming back.

 

https://www.morningstar.com/articles/929002/kinnel-3-closing-thoughts-from-the-conference

 

 

To be a good equity investor, I think you have to be an optimist.

 

https://www.oaktreecapital.com/docs/default-source/memos/something-of-value.pdf

[Spooky]

AI is the new thing for the market to get fixated on. Run away from fads. My personal view is that this will probably lead to an increase in productivity across many companies / industries and the best play is probably just a total world index since it is likely impossible to pick the winners and losers at this stage.

[Sweet]

There is an article on Marketwatch today of Bill Gates claiming that AI will mean we won’t use a search engine again i.e. Google, or Amazon again.

 

https://www.marketwatch.com/story/bill-gates-lays-ai-tombstone-on-amazon-and-google-61814cdd?mod=home-page

 

I think he is wrong on both counts.

 

I think some of the predictions of what AI will replace right now, as Gates predicted above, will be quite wide of the mark.

 

I think AI will have an impact that nobody can quite predict.

 

I don’t see a broad market bubble as a result of AI right now, so if one were to develop we would be in the very early innings.

[Luca]

As of today, Nvidia has a marketcap of 760b, trades at somewhat 160x earnings if we deduct SBC, 28x revenue. In order to make a 10% CAGR it has to be at a 2 T marketcap in a decade.

 

Lets look at growth assumptions, current earnings are at 7b. In order to be valued at 2T in 10 years, i would expect at least 100b in Cashflows in 2033. 

 

If Nvidia grows earnings at 30% annualized for 10 years, we will be at 96b in earnings in a decade. So right around that 2T Marketcap at 20x earnings, excluding SBC and excluding possible buybacks or dividends. 

 

So we buy into 30% of annualized growth for a decade for a 10-15% CAGR. 

 

And the downside? 

 

Approximately -65% for a decade of 15% earnings growth.

 

Approximately -80% growth for a decade of 10% earnings growth.

 

 

 

 

 

I think from current valuations we will get a -30% to -50% in 10 years 

 

 

No thank you!

Edited May 24, 2023 by Luca

[dealraker]

In my view the reaction of an over-valued market will be for the most over valued and well-known stocks to get even more over valued.  That's preciscly what happened with the dot.com bubble of the late 1990's, the big cap tech stocks soared to even new heights in 2000.  It turns out this bunch may have been fully valued literally years earlier.

 

So for years now those believing in Berkshire on the other Berkshire board I follow have in unison chanted loudly, "Berkshire is under valued while the S and P is wildly over valued."  At the same time this discussion is happening those in the discussion have made it well known their picks outside of Berkshire and it is always an assortment - but not all of course - of Amazon, Alphabet, Meta, Microsoft, Nvidia, and Tesla...the very core and dominating bunch of stocks these value investors claim is in aggregate over valued.

 

It isn't surprising what's happening today with AI or whatever (I don't think investors in these stocks needed an excuse- they were going to get run up regardless) and it isn't surprising as to why Berkshire hasn't kept up with the S & P.   To me it seems the big cap stocks of today in the S & P are more durable than those of the early 2000's.  But of course I felt the exact same way about the big cap tech stocks back then.

 

Time will tell.  But one thing's for sure and that is the stocks racing ahead today are  precisely the stocks you'd expect to race ahead.   It is all investors are interested in for the most part.  Everybody can retire on the same few stocks is the name of the game.  Here come the studies that "this is the only bunch that's created value in the last x years" and whatnot.

[james22]

37 minutes ago, Luca said:

As of today, Nvidia has a marketcap of 760b, trades at somewhat 160x earnings if we deduct SBC, 28x revenue. In order to make a 10% CAGR it has to be at a 2 T marketcap in a decade.

 

Only evidence of how early we are:

 

[james22]

https://archive.ph/9h450

[brobro777]

On 5/23/2023 at 1:04 PM, brobro777 said:

I've read this forum for a while and got good ideas (St. Joe thread) so I figure I would contribute

 

I do not think this AI thing will not go much further from here. In fact I think the upward run in tech in general that began in 2003 was finished as of last year and this AI thing will not lead to the tech sector making new highs

 

I will probably turn out to be completely wrong and NVDA will hit $1 Tril in market cap in 2 months 

 

 

Haha I thought it would take 2 months for NVDA $1Tril, it'll take 2 days! haha

[james22]

 

[james22]

 

[Gamecock-YT]

On a ship that has 10 channels, one of which is CNBC, every time I flip it on they’re talking AI. Guess it’s like when bitcoin/SPACs/EVs hype took off. Don’t know why people even watch that shit. 

[Guest]

On 5/23/2023 at 4:04 PM, brobro777 said:

I've read this forum for a while and got good ideas (St. Joe thread) so I figure I would contribute

 

I do not think this AI thing will not go much further from here. In fact I think the upward run in tech in general that began in 2003 was finished as of last year and this AI thing will not lead to the tech sector making new highs

 

I will probably turn out to be completely wrong and NVDA will hit $1 Tril in market cap in 2 months 

 

Did you mean 2 months or 2 weeks? 

[brobro777]

9 minutes ago, stahleyp said:

 

Did you mean 2 months or 2 weeks? 

 

It wasn't even 2 weeks! It was like 2 days!

 

Shows how much I know! Haha

 

[james22]

On 6/2/2023 at 3:40 AM, Gamecock-YT said:

On a ship that has 10 channels, one of which is CNBC, every time I flip it on they’re talking AI. Guess it’s like when bitcoin/SPACs/EVs hype took off. Don’t know why people even watch that shit. 

 

1. Siegel's Siegel, whatever the medium.

 

2. Being aware of the current hype is a reason to watch.

[Gregmal]

Let’s hope this goes on long enough for Cerebras to IPO. Then we can call it a day. 

[John Hjorth]

I a firm believer in it. It seems to be able to fix everything, it even appears to be able to fix the development pipelines in the Swedish Real Estate Sector for the contractors, when the Swedish corporate real estate companies have problems with access to [further] liquidity in the banks or the bond market. [J/K]

 

[Voted 1995.]

 

Edited June 3, 2023 by John Hjorth

[tnp20]

From an investment perspective here is the state of affairs in the AI space.

 

Its still early but looks promising longer term as new developments are occurring rapidly both in breadth and depth .....this a revolution in the making....its real. There will no doubt be a 2001 moment as with dot.com....and the idea is to identify who will lift off during the inevitable bubble...

 

Best Play: Specialized semiconductor players - huge parallel compute needed, and efficient compute needed for inferencing

(i) NVIDA - this is where its at...only game in town. Grossly overvlaued but could be driven higher by momo.

(ii) AMD - about 1.5-2 years behind but have everything it takes to catch up - reasonably valued/may be slightly over valued. This would be my choice on pull backs.

(iii) INTC - they dropped the ball - ship is going to take a while to turn around and they are loosing ground - they need to quantum leap to catch up and not likely in next 3-5 years.

(iv) SKhynix/Samsung/Micron - besides GPU chips, high bandwidth memory is needed for AI chips in huge demand .... each has issues so not compelling case by itself. not sure how to invest in SKHynix. Samsung is too diversified. Micron may be behind in High band width memory.

 

Second Best:  Hyperscalers - need huge compute to both train and then inference (run the model)

(i) Microsoft - by association with OpenAi and they have an ability to imbed AI in all their products and charge more.  - valuation seems a bit rich but not unreasonable.

(ii) Google - Leader in AI space ...they have a lot of stuff in the lab that needs to get productized but will get there, they also make their own AI chips - good value here - this would be my choice on pull backs.

(iii) Amazon - later starter but this is a marathon and with their huge customer base they have a lock on customers locked to AWS - good value here - this would be my choice on pull backs.

 

 

Third tier: Data Management companies since data is what gets fed into the models

(i) Snowflake - low conviction but may get there - seems over valued but Buffet or his lieutenants bought its during IPO at higher prices.

(iv) Oracle  - new one popped up on my radar - they bought a lot of the new high end NVDIA gpus and building a cloud with it and with DB customer lock seems a natural fit and beneficiary long term - good value here but conviction low as I havent been following Oracle story much.

 

 

Unicorns: Pre-ipo companies that are coming on the radar but you cant invest in them yet

(i) Data Bricks

(ii) Anthropic

(iii) Scale AI

(iv) Hugging Face

(v) Cohere

(vi) Tenstorrent

...many others ...I dont follow these closely as its too early for these guys from an investment perspective...

 

Edited July 20, 2023 by tnp20

[Saluki]

Bezos had an interesting take on the internet many years ago.  He analogized it to electricity.  When people got electricity into their houses, they only used it for light. Then they figured out other things they could do with it like a washing machine, then a radio, then a TV.   So the internet, he said, was only in it's early innings.  It's entirely possible that AI will continue to improve and do things that we can't even envision now.  But like the internet, there are winners and losers.  Amazon won, and arguably consumers, because by putting all the retailers online, it drove down margins and wiped out a lot of retail. So early retailers thought it was great until everyone else showed up.  If you are a copyrighter or graphic designer this is scary. When Apple started using multiple fonts and automatic Kerning, the work of a lot of low level typesetters and graphic designers disappeared overnight.  Instagram filters make it easier to take pictures so it's harder to be a photographer, but easier to be an influencer. 

 

Like the early internet, it may be hard to pick the winners.  Where is Pets.com and where is Netscape?  But like the gold rush, the people selling the picks and shovels will probably do well.  So maybe not ChatGPT, but the servers (AWS, GOOG, MSFT) where the data is stored, or the chips needed to make AI work by crunching the numbers. The problem is that everyone knows this.  So is it better to buy now and put it your coffee can portfolio or wait till the next tech crash like in 2000, when AMZN was selling for single digits per share? 

[brobro777]

15 minutes ago, Saluki said:

Bezos had an interesting take on the internet many years ago.  He analogized it to electricity.  When people got electricity into their houses, they only used it for light. Then they figured out other things they could do with it like a washing machine, then a radio, then a TV.   So the internet, he said, was only in it's early innings.  It's entirely possible that AI will continue to improve and do things that we can't even envision now.  But like the internet, there are winners and losers.  Amazon won, and arguably consumers, because by putting all the retailers online, it drove down margins and wiped out a lot of retail. So early retailers thought it was great until everyone else showed up.  If you are a copyrighter or graphic designer this is scary. When Apple started using multiple fonts and automatic Kerning, the work of a lot of low level typesetters and graphic designers disappeared overnight.  Instagram filters make it easier to take pictures so it's harder to be a photographer, but easier to be an influencer. 

 

Like the early internet, it may be hard to pick the winners.  Where is Pets.com and where is Netscape?  But like the gold rush, the people selling the picks and shovels will probably do well.  So maybe not ChatGPT, but the servers (AWS, GOOG, MSFT) where the data is stored, or the chips needed to make AI work by crunching the numbers. The problem is that everyone knows this.  So is it better to buy now and put it your coffee can portfolio or wait till the next tech crash like in 2000, when AMZN was selling for single digits per share? 

 

That seems to be a whole lot of things to know now, on top of all the things to know in the future as the variables constantly change. I'd understand making a bet if interest rates were 0% or these techs were trading at 10PE like MSFT in 2011, but now with risk free rate at 5%? Hoo boy...

 

The fact that people are rushing headfirst into these things with a huge list of uncertainty to pop NVDA $200Bil in one day makes me think we're at year 2000! 

 

I wouldn't be surprised if I turn out to be completely wrong though, haha

 

[Spekulatius]

if you like to listen to an AI bull in podcast format:

https://podcasts.apple.com/us/podcast/tech-trader-the-outlook-for-q2-tech-earnings/id1523729084?i=1000621565711

[Gmthebeau]

AI is wildly overhyped.  You can tell this because there are virtually no practical applications for it.  Every company issues press releases and powerpoint presentations.  There is no there there.  NVDA reached pretty much same P/S ratio that CSCO did at its peak in 2000.  I suspect they may be able to push it up a bit more but it's mostly done.  Elon Musk will never have full self driving cars its all smoke and mirrors.

Edited July 20, 2023 by Gmthebeau

[tnp20]

30 minutes ago, Gmthebeau said:

AI is wildly overhyped.  You can tell this because there are virtually no practical applications for it.  Every company issues press releases and powerpoint presentations

 

You are both right and wrong.

 

Hype has gotten ahead of reality where everyone is touting use of AI.

 

The reality is moving forward with real AI applications getting implemented and used everyday - the most glaring example is not ChatGPT but something called co-pilot. Its used in coding as well as office suite of products. Others have things that are similar. Many , many other applications now being used and its using the new AI (so called transformer or foundation model, not the old AI...raw neural networks you had to configure yourself).

 

Its coming...but S curve applies...first slowly and then all of a sudden. We are far from 2000-2001 moment. The mother of all bubbles may be coming ...

 

There are 3 opportunities here......

(i) trade into a bubble

(ii) short the heck out of the top (Jessy moment)

(iii) and when babies get thrown out of the bath water, pick up the amazons and the microsofts of AI for long term hold...

Edited July 20, 2023 by tnp20

[Gmthebeau]

7 minutes ago, tnp20 said:

 

You are both right and wrong.

 

Hype has gotten ahead of reality where everyone is touting use of AI.

 

The reality is moving forward with real AI applications getting implemented and used everyday - the most glaring example is not ChatGPT but something called co-pilot. Its used in coding as well as office suite of products. Others have things that are similar. Many , many other applications now being used and its using the new AI (so called transformer or foundation model, not the old AI...raw neural networks you had to configure yourself).

 

Its coming...but S curve applies...first slowly and then all of a sudden. We are far from 2000-2001 moment. The mother of all bubbles may be coming ...

 

There are 3 opportunities here......trade into a bubble, short the heck out of the top (Jessy moment), when babies get thrown out of the bath water, pick up the amazons and the microsofts of AI for long term hold...

 

There will be some applications but it's all incremental basically, nothing game changing.  It won't be anywhere near as big as the internet.

 

Prior to the AI frenzy, we heard how the Metaverse was going to be the next big thing.  It hasn't caught on, so they moved on to AI - basically to get a rally started.

 

I would bet the XLU (utilities) vastly outperforms the AIQ (AI ETF) over the next 20 years.