Why does SPACEX do the things it does?

[fatpiggy]

Xai was upside down before the merger. These folks are leaving probably due to pressure to produce results as grok falls behind competition. The people they are hiring right now include legal and video designers. They are clearly going in a different direction: less science more content.
The important thing that im seeing is: Models across companies are peaking. Meaning the training and releasing of new ones that actually score LESS than their previous versions. This is a catastrophe for those super ai dreams. Hang on to your butts as tgis realization spreads.

It came out they left due to getting laid off. Elon restructured Xai under SpaceX and let these people go to to overlap. They all left on good terms, and said positive things about Xai, Elon and SpaceX. If i had to guess, I bet they get rehired in a different position.

So you can throw all of your speculation in the garbage.

The science is providing the content. The science leads to innovation like image generation. Whole movies will be rendered by AI, by the end of this year or maybe next at the latest. I hear you will interact with the movie in the future. You will get to choose what path the movie takes.

Did you see where Xai is going to bypass coding? I thought that was pretty cool. Instead of going from english to code to compiler to result .... they are just going to go english to result. Skip those other steps. Coding will be obsolete in 3-5 years or less.

We are in for an exciting future.

And to your last point, that goes exactly against what one of the people that left Xai proclaimed. This is a LONG article, but i think it's right up your alley. It contradicts your last claim and i have to say i think you are wrong. From one of the engineers that left.

Link below. Check it out, think you will like it.

Something Big Is Happening — matt shumer

 

[baltimorened]

Xai was upside down before the merger. These folks are leaving probably due to pressure to produce results as grok falls behind competition. The people they are hiring right now include legal and video designers. They are clearly going in a different direction: less science more content.
The important thing that im seeing is: Models across companies are peaking. Meaning the training and releasing of new ones that actually score LESS than their previous versions. This is a catastrophe for those super ai dreams. Hang on to your butts as tgis realization spreads.

being pressured to produce results...sounds like the work experience of the majority of americans

 

[firegiver]

being pressured to produce results...sounds like the work experience of the majority of americans

The technology is cool, the actual revenue generation ... is flimsy. This is like the dot Com bubble all over again.

 

[firegiver]

It came out they left due to getting laid off. Elon restructured Xai under SpaceX and let these people go to to overlap. They all left on good terms, and said positive things about Xai, Elon and SpaceX. If i had to guess, I bet they get rehired in a different position.

So you can throw all of your speculation in the garbage.

The science is providing the content. The science leads to innovation like image generation. Whole movies will be rendered by AI, by the end of this year or maybe next at the latest. I hear you will interact with the movie in the future. You will get to choose what path the movie takes.

Did you see where Xai is going to bypass coding? I thought that was pretty cool. Instead of going from english to code to compiler to result .... they are just going to go english to result. Skip those other steps. Coding will be obsolete in 3-5 years or less.

We are in for an exciting future.

And to your last point, that goes exactly against what one of the people that left Xai proclaimed. This is a LONG article, but i think it's right up your alley. It contradicts your last claim and i have to say i think you are wrong. From one of the engineers that left.

Link below. Check it out, think you will like it.

Something Big Is Happening — matt shumer

How do you use AI today to generate revenue?

 

[fatpiggy]

How do you use AI today to generate revenue?

Multiple trading firms use AI.

 

[firegiver]

Multiple trading firms use AI.

Sorry, I was asking about you. Not 'trading firms "use" AI'.

The rub is this: Its a cool tech, getting it to help you generate rev. Thats the real talent in todays world and what EVERYONE IS TRYING TO DO.

Do you use claudebot? Do you do MCP's, Do you use RAG?
If you don't understand these things, then I encourage you to learn about them and understand whats happening at the actual factual implementation level. If you are just some speculator who pretends to understand technology, then thats ok. If you are speculating on these companies you should, if you are serious, be looking at revenue generated... not hype claims.

 

[fatpiggy]

Sorry, I was asking about you. Not 'trading firms "use" AI'.

The rub is this: Its a cool tech, getting it to help you generate rev. Thats the real talent in todays world and what EVERYONE IS TRYING TO DO.

Do you use claudebot? Do you do MCP's, Do you use RAG?
If you don't understand these things, then I encourage you to learn about them and understand whats happening at the actual factual implementation level. If you are just some speculator who pretends to understand technology, then thats ok. If you are speculating on these companies you should, if you are serious, be looking at revenue generated... not hype claims.

I was in the first class of electronic options traders. It used to be the old dudes in the pit flashing hand signals like in the movie Trading Places. Thats how I started in the year 2000.

It just so happened that the internet exploded when we were in college and I had good computer skills, much better than your average trader. For a few years I was able to run circles around people electronically but the competition is much tougher theses days.

I’ve recently trained AI on 10 years of market data (data is expensive!) . I understand how it works. Trading groups like Jane street, Susquehanna, Goldman, et al all use AI in their trading.

AI is just pattern recognition. Guess what trading is? Same thing. You recognize the pattern before everyone else and then you have to beat them to the trade. It’s incredibly competitive. Honestly, you are kind of a dick, so you would probably be good at it.

Anyway, just letting you know my history is much more than just a speculator. Speculating, high speed trading (computer science), portfolio management (statistics, economics), are all part of the job.

I won’t question your qualifications. You know what you are talking about and I appreciate your feedback even though you are kind of a dick.

 

[firegiver]

I was in the first class of electronic options traders. It used to be the old dudes in the pit flashing hand signals like in the movie Trading Places. Thats how I started in the year 2000.

It just so happened that the internet exploded when we were in college and I had good computer skills, much better than your average trader. For a few years I was able to run circles around people electronically but the competition is much tougher theses days.

I’ve recently trained AI on 10 years of market data (data is expensive!) . I understand how it works. Trading groups like Jane street, Susquehanna, Goldman, et al all use AI in their trading.

AI is just pattern recognition. Guess what trading is? Same thing. You recognize the pattern before everyone else and then you have to beat them to the trade. It’s incredibly competitive. Honestly, you are kind of a dick, so you would probably be good at it.

Anyway, just letting you know my history is much more than just a speculator. Speculating, high speed trading (computer science), portfolio management (statistics, economics), are all part of the job.

I won’t question your qualifications. You know what you are talking about and I appreciate your feedback even though you are kind of a dick.

Yah did you work in Charleston at automated trading desk? Before citi bought them? You might want to understand i have a lot of knowledge, first hand about what you are talking about.

 

[fatpiggy]

Yah did you work in Charleston at automated trading desk? Before citi bought them? You might want to understand i have a lot of knowledge, first hand about what you are talking about.

No I lived in Chicago and worked on the trading floor 15 years. When the servers got co-located in Secaucus and Aurora I was able to move out of the frozen Windy City. Downtown Chicago is no place to raise a family.

ATD was big on stock trading from what I remember. Back in the early 2000’s options weren’t electronic yet, or the transition from pit to electric was just beginning. I saw electronic trading move from seconds a trade, down to milliseconds, to microseconds and now down to nanoseconds seconds. Trading on the millisecond to microseconds level requires a lot of understanding of computers and physics.

I’ve never doubted you know what you are talking about. I can tell from your posts you k ow what you are talking about. I think it’s funny you doubt me. Maybe you are trying to warn me on SpaceX? I invested when it was valued at $45B, so I think I’ll be ok. (My connection is an old roomate and good friend with the SpaceX CFO. Look when I first heard about space data centers, it was over a year ago. I made a post about it at the time ).

A large reason I post on the board so often is because trading is a lot of psychology and hearing everyone’s feedback helps me get a feel for what moves the markets. I stare at the screen all day and thus what allows me to post a lot.

What about you? What is your line of computer work?

 

[firegiver]

No I lived in Chicago and worked on the trading floor 15 years. When the servers got co-located in Secaucus and Aurora I was able to move out of the frozen Windy City. Downtown Chicago is no place to raise a family.

ATD was big on stock trading from what I remember. Back in the early 2000’s options weren’t electronic yet, or the transition from pit to electric was just beginning. I saw electronic trading move from seconds a trade, down to milliseconds, to microseconds and now down to nanoseconds seconds. Trading on the millisecond to microseconds level requires a lot of understanding of computers and physics.

I’ve never doubted you know what you are talking about. I can tell from your posts you k ow what you are talking about. I think it’s funny you doubt me. Maybe you are trying to warn me on SpaceX? I invested when it was valued at $45B, so I think I’ll be ok. (My connection is an old roomate and good friend with the SpaceX CFO. Look when I first heard about space data centers, it was over a year ago. I made a post about it at the time ).

A large reason I post on the board so often is because trading is a lot of psychology and hearing everyone’s feedback helps me get a feel for what moves the markets. I stare at the screen all day and thus what allows me to post a lot.

What about you? What is your line of computer work?

Data science, software engineering is my trade today, but now I manage several teams at a company. I was with ATD starting in 2005. Was there when we installed a server on wall st. Was crazy times. Then 2009 happened.

 

[fatpiggy]

Data science, software engineering is my trade today, but now I manage several teams at a company. I was with ATD starting in 2005. Was there when we installed a server on wall st. Was crazy times. Then 2009 happened.

ATD was a little before my time. I remember reading about them when i was living in Chicago. But as I recall they were strictly equities and not derivatives. It's kind of funny, when i moved to CHS i ended up right over by those old ATD buildings in Mt Pleasant. I live right over in that area now.

For many years i worked side by side with software engineers daily building my own high speed trading software. Unfortunately, they were based in Russia (When I started it was a Swedish company, but over the years it got sold to Russians). When the Russia / Ukraine war broke out our relationship really soured. I now use SpiderRock Trading software instead of my custom-built software. I probably have 5% of your skills, but know enough to be dangerous.

Are you a part of the data center build outs? I've noticed CAT stock is on a tear, tells me the build out claims are probably true. I just think growth is going to explode with all these companies making these huge investments. META who would normally spend $50B on R&D upped it to $200B. Same with most of the big names. They see something and I don't think they are going to miss. Of course ANYTHING is possible, it's just my gut feeling that has developed after following markets for 30 years. At least the next few years should be really interesting.

What is your main concern? Just the valuations of these companies?

 

[tigres88]

Data science, software engineering is my trade today, but now I manage several teams at a company. I was with ATD starting in 2005. Was there when we installed a server on wall st. Was crazy times. Then 2009 happened.

It'll be a while before the everyday user understands what "MCP" and "RAG" means. I'm customer facing but our entire SaaS product is turning into a Generative AI solution- it'll be interesting to see how it goes. A few of my customers really want us to switch from being API first to an MCP solution, which makes sense to me, but then it opens up the ability (depending on what we send through the MCP server) for them to build a solution like ours easier.

I do think some companies are finding ways to generate revenue through Gen AI, but for the most part its been about speeding up deliverables, synthesizing information, and making processes faster (including some code work). To actually affect the bottom line of Enterprise customers (that's only who I service) I think it will take AI being able to action on data/information rather than just synthesizing, aggregating/etc.

 

[firegiver]

It'll be a while before the everyday user understands what "MCP" and "RAG" means. I'm customer facing but our entire SaaS product is turning into a Generative AI solution- it'll be interesting to see how it goes. A few of my customers really want us to switch from being API first to an MCP solution, which makes sense to me, but then it opens up the ability (depending on what we send through the MCP server) for them to build a solution like ours easier.

I do think some companies are finding ways to generate revenue through Gen AI, but for the most part its been about speeding up deliverables, synthesizing information, and making processes faster (including some code work). To actually affect the bottom line of Enterprise customers (that's only who I service) I think it will take AI being able to action on data/information rather than just synthesizing, aggregating/etc.

same

I'd add, I think C suites are just now discovering the real challenges with this technology. Compunding engineering knowledge is AMAZING, but it can quickly get waaay out of control and hallucinate and infect all your value streams and bring things to a hault. Jobs are now appearing in the market that are just: Unfuck the AI slop.
Great tech, but its lightening in a bottle and needs a lot more maturing of processes and guardrails to harness it in a meaningful way.

 

[firegiver]

ATD was a little before my time. I remember reading about them when i was living in Chicago. But as I recall they were strictly equities and not derivatives. It's kind of funny, when i moved to CHS i ended up right over by those old ATD buildings in Mt Pleasant. I live right over in that area now.

For many years i worked side by side with software engineers daily building my own high speed trading software. Unfortunately, they were based in Russia (When I started it was a Swedish company, but over the years it got sold to Russians). When the Russia / Ukraine war broke out our relationship really soured. I now use SpiderRock Trading software instead of my custom-built software. I probably have 5% of your skills, but know enough to be dangerous.

Are you a part of the data center build outs? I've noticed CAT stock is on a tear, tells me the build out claims are probably true. I just think growth is going to explode with all these companies making these huge investments. META who would normally spend $50B on R&D upped it to $200B. Same with most of the big names. They see something and I don't think they are going to miss. Of course ANYTHING is possible, it's just my gut feeling that has developed after following markets for 30 years. At least the next few years should be really interesting.

What is your main concern? Just the valuations of these companies?

Yes the valuation of TSLA and SpaceX are pretty crazy. Objectively, TSLA's PE ratio is abysmal.
Also, as a long time investor in TSLA, and I've stated this several times, they are not delivering on a lot of the promises that caused investors to flock to them, and when I read and hear from their CEO, he just keeps jumping to the next thing instead of delivering. Its been 10 years since I've seen TSLA have a true success. JMO

 

[firegiver]

Also, alarm bells should be going off based on sales and their goals: 10m a year by 2030.
Sales have stagnate at 1.6 per year, while overall ev sales continue to go up for all product makers. Total global ec sales are now 20 million as of 2025.
Tsla has achieved... 2% of their goal with 4 years left on their goals.
Also, look up dojo... their super computer ... it was to be used to train their AI models. Well that was quietly abandoned.
His recent habit of selling his companies to himself, is objectively a huge red flag to all of this.

 

[fatpiggy]

Yes the valuation of TSLA and SpaceX are pretty crazy. Objectively, TSLA's PE ratio is abysmal.
Also, as a long time investor in TSLA, and I've stated this several times, they are not delivering on a lot of the promises that caused investors to flock to them, and when I read and hear from their CEO, he just keeps jumping to the next thing instead of delivering. Its been 10 years since I've seen TSLA have a true success. JMO

No argument from me on TSLA valuation. I sold all my shares. Will get back in at some point.

 

[fatpiggy]

P/ E is a little deceptive in my opinion o these high growth tech companies. It requires you to look at earnings in the rear view mirror. I think fwd P/e is a better benchmark. Tesla still trading on a fwd p/e of about 250 I believe. Thats crazy high.

Take NVDA. It’s trading closer to a 30 fwd p/e I believe and Amazon around 20!

Elon is the master of the dangling carrot. He often misses timelines, but he does deliver great products. His fans hold onto the stock no matter the valuation and he is a master of creating hype. The stock is expensive for a reason, they are well diversified startup company with tremendous growth.

I dont disagree with Firegiver the Tesla stock is expensive. But it’s also hard to find good entry points. I don’t blame Elon for this. People should blame the lack of competition. Elon is just out working everyone imo.

 

[fatpiggy]

Detailed breakdown of the math and physics required for orbital data centers. Elon said it was a pretty good estimate.

Interesting that instead of trying to cool the equipment in space, they are just trying to engineer the chips to run at a higher temperature so they don’t have to cool them at all.

 

[fatpiggy]

Happily will bet you on that one

Where do I collect?

SpaceX Readies IPO Paperwork For March, Targeting $1.75 Trillion Valuation<!-- --> | ZeroHedge

ZeroHedge - On a long enough timeline, the survival rate for everyone drops to zero

www.zerohedge.com

 

[fatpiggy]

Bezos getting in the game. Good for SpaceX to have some competition.

 

[fatpiggy]

Secondary offering, but yes. He’s doing this because he needs to inflate the valuations because he has so much leverage across all his businesses. In particular he is falling behind on Xai which is generating very little revenue and he needs to keep borrowing. The circular financing of AI is child’s play compared to his portfolio. Teslas bubble popping would literally break him.

The banks will happily accommodate this because they want his dealflow so keep him in good graces. Other investors will do a bad deal because they want access to his dealflow. Space x will be a massive IPO if it ever goes public, but good luck getting a $800B valuation at IPO.

It may close over $3T the first day of trading.