Is A.I. overhyped?

[mooseolly]

Outside of big tech, consumer staples are losing revenue. The wider US economy is in trouble. Verizon just announced a 15% cut to their entire workforce.

There are also reports that big tech is not being honest with their earnings reports, a lot of hidden debt not reflected in their ER. They are borrowing massively using SPV, to fuel their data center boom and those debt is not shown to shareholders.

I already said around 85% of SP500 beat estimates. It is a big number when historical average about 78%.

The reports from Michael Burry?
Well he has just given up close his asset management business:

 

[focus1974]

Boss asked me switch some ops to AI.Shuddered to think someday will become the movie I,Robot ..where they really take over ..

Not hype if you can automate things with AI.

you donno, then it is hype.

 

[Mechafanboy]

I already said around 85% of SP500 beat estimates. It is a big number when historical average about 78%.

The reports from Michael Burry?
Well he has just given up close his asset management business:

Yet US consumer sentiment is the lowest in years. US consumer data is weak.... Red flags everywhere.

 

[FORSTA]

bubble will burst soon.
like 2008.

 

[focus1974]

Look at the big techs financials. Double digit growth % yoy revenue. In fact around 85% of SP500 beat earnings.

Yes.. even if there is a Bubble.... or panic about capex spending on AI infrastructure without corresponding demand...

I think the existing giants will be the final victor as they are the only one with existing huge cashflow from ops to fund AI infrastructure as well as the one who have direct access to customer AI demands. So they can scale correspondingly without much pain.

So my existing portfolio has the cloud providers with other moats in their respective spaces
1) Google
2) Microsoft
3) Amazon

Software side.
1) Palantir

Consumer hardware side
1) Apple (now understand their core is brand and user experience and their own ecosystem, not AI - outsource to Google - no need build).

Hardware side (hardware usually go thru cyclical cycle but this time it's a megatrend)
1) TSMC
2) AVGO (this one I don't have)
3) AMD
4) NVDA
5) MU

 

[mooseolly]

Yet US consumer sentiment is the lowest in years. US consumer data is weak.... Red flags everywhere.

That’s because of more jobless due to automation. If I get retrenched I will keep my wallet tight.

 

[mooseolly]

Yes.. even if there is a Bubble.... or panic about capex spending on AI infrastructure without corresponding demand...

I think the existing giants will be the final victor as they are the only one with existing huge cashflow from ops to fund AI infrastructure as well as the one who have direct access to customer AI demands. So they can scale correspondingly without much pain.

So my existing portfolio has the cloud providers with other moats in their respective spaces
1) Google
2) Microsoft
3) Amazon

Software side.
1) Palantir

Consumer hardware side
1) Apple (now understand their core is brand and user experience and their own ecosystem, not AI - outsource to Google - no need build).

Hardware side (hardware usually go thru cyclical cycle but this time it's a megatrend)
1) TSMC
2) AVGO (this one I don't have)
3) AMD
4) NVDA
5) MU

When majority of the people expecting a bubble, it won’t happen, at least for now.

I notice your list do not have Meta. It is the only undervalued Mag7 now due to the fear of repeating Metaverse spending. However to me that is a growth capex. Meaning to say they are building and increasing their hardwares. If it comes to the stage that they have surplus capacity they can do like Amazon, Microsoft & Google to outsource the capacity to others. That would be a new stream of income for the company.

 

[Mechafanboy]

That’s because of more jobless due to automation. If I get retrenched I will keep my wallet tight.

US household debt is also at record highs. US consumers are driven to the brink...

There is a reason why DJT is polling so poorly when it comes to the economy.

 

[Hello_Kitty]

Korea made an AI movie liao. The script was bad but the effects were typical Hollywood.

don’t look nice

 

[audiovideo]

many tiong US market YouTubers believe GPU is a piece of graphic card

 

[focus1974]

When majority of the people expecting a bubble, it won’t happen, at least for now.

I notice your list do not have Meta. It is the only undervalued Mag7 now due to the fear of repeating Metaverse spending. However to me that is a growth capex. Meaning to say they are building and increasing their hardwares. If it comes to the stage that they have surplus capacity they can do like Amazon, Microsoft & Google to outsource the capacity to others. That would be a new stream on income for the company.

True. I actually bought META when it was really undervalued around $160... but got out during the Trump Tarriff crash.

So now hard to get back in. hehe.
Though, I think their execution has not been there. Really like a llama. Now pay king's ransom to hire AI people to play catchup
Though they have ops cashflow to fund, but their capex would be more massive than the other 3 because they have existing cloud infrastructures and working AI models (for Google and Microsoft). AWS Cloud is there but their AI is missing, though i would opine their robotics automation thru AI is the best of the rest. Don't think anyone can boast of the size of robotics automation in logistics like them.

 

[mooseolly]

True. I actually bought META when it was really undervalued around $160... but got out during the Trump Tarriff crash.

So now hard to get back in. hehe.
Though, I think their execution has not been there. Really like a llama. Now pay king's ransom to hire AI people to play catchup
Though they have ops cashflow to fund, but their capex would be more massive than the other 3 because they have existing cloud infrastructures and working AI models (for Google and Microsoft). AWS Cloud is there but their AI is missing, though i would opine their robotics automation thru AI is the best of the rest. Don't think anyone can boast of the size of robotics automation in logistics like them.

Yea Amazon robotic automation is one of the best if not the best in the industry.

 

[Evcats]

 

[mooseolly]

US household debt is also at record highs. US consumers are driven to the brink...

There is a reason why DJT is polling so poorly when it comes to the economy.

Side effects of AI & automation. These people have to adapt or get phased out. The same thing is happening in China.

 

[Shinning_Wolf]

It’s going to burst by the end of next year and peasants are going to be one paying the price.

 

[Mechafanboy]

Automation shouldn't be confused with LLM / AI... Automation follows a fixed set of programmed instructions...

LLM / current AI is a token predictor/generator. Not the same.

 

[mooseolly]

 

[NintendoSwitch]

True. I actually bought META when it was really undervalued around $160... but got out during the Trump Tarriff crash.

So now hard to get back in. hehe.
Though, I think their execution has not been there. Really like a llama. Now pay king's ransom to hire AI people to play catchup
Though they have ops cashflow to fund, but their capex would be more massive than the other 3 because they have existing cloud infrastructures and working AI models (for Google and Microsoft). AWS Cloud is there but their AI is missing, though i would opine their robotics automation thru AI is the best of the rest. Don't think anyone can boast of the size of robotics automation in logistics like them.

Zuck created ownself problem. Pay new hires moar then old hires, so current AI head leaving.
https://nypost.com/2025/11/11/busin...nn-lecun-to-depart-launch-own-startup-report/

 

[JivBunny]

Boss asked me switch some ops to AI.Shuddered to think someday will become the movie I,Robot ..where they really take over ..

it is not cheap and not so easy unless you already have some tech pple or IT process.

Currently a lot of the AI is more at LLM or automation of some mundane minutes taking or slides generation. I call it level 1 of AI.

AI needs to couple with large data plus a lot of process changes.

 

[JivBunny]

Cos need to invest in robotics, maintenance and training the AI and human to manage it. For security guards, it will be good like Jewel Changi Certis that kind to perform remote monitoring of premises (means no more prowling). But how well it performs we won't know. Maybe one day Certis fed up they go back human patrol

Security robots are not what i call AI.

I class them as tech enablers to augment productivity.