1
- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Coaching Changes
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Nebius - NBIS - AI Infrastructure Company
Posted on 9/1/26 at 1:02 pm to NotStupid
Posted on 9/1/26 at 1:02 pm to NotStupid
quote:What does this have to do with NBIS other than constant winning?
ChatGPT, Claude, Gemini…can’t do this. They can’t make the end user experience free while also monetizing and making it profitable. And that doesn’t take into account the open source free ones that will always be competing against them.
Posted on 9/1/26 at 1:06 pm to NotStupid
quote:You're simply failing to imagine the reach of physical ai, in addition to healthcare and finance as Jax stated.
So in the end what will these products really do? I think they will be niche B2B products. Just another iteration of SAAS. On a much smaller scale than what it’s TAM is being billed as today. SAAS has clearly been disrupted due to AI/LLMs replicating what they do. But I don’t see them doing anything outside the SAAS space.
Take your Ed boyfriend and compare what he says to Gavin Baker's position. Have them duke it out in your favorite LLM.
Posted on 9/1/26 at 4:32 pm to bayoubengals88
quote:
What does this have to do with NBIS other than constant winning?
NBIS biggest customers are META and Google.
Do you expect META and Google to continue to spend enormous amounts of money on LLM products (Gemini and whatever model Zuck has tried to build), if they can’t monetize them? And for how long do you think they will do this? Do you expect Meta/Google to both grow their own revenues substantially to support their ability to allocate capital toward NBIS? Will their investors approve of these expenditures?
NBIS isn’t contracted to OpenAI. But if OpenAI were to fail, that would have big impacts on all the Data Center builders because at that point there would be excess supply on the market. Does that concern you?
OpenAI/Anthropic have ordered 70% of the data center build. 2 companies. Both very unprofitable. When only 2 companies represent 70% of the demand in an industry that is a very unstable industry and why it’s being shorted
Posted on 9/1/26 at 5:01 pm to bayoubengals88
quote:
You're simply failing to imagine the reach of physical ai, in addition to healthcare and finance as Jax stated.
Since this is the NBIS thread, I’ll stick with them.
How are META and Google (the two largest NBIS customers) going to create a physical AI product that will be ran thru the NBIS data center. What’s the product? And what do you imagine it doing?
Posted on 9/1/26 at 7:54 pm to NotStupid
quote:
NBIS biggest customers are META and Google
Roughly 50% of revenue now and declining, which is good.
quote:
Do you expect META and Google to continue to spend enormous amounts of money on LLM products (Gemini and whatever model Zuck has tried to build), if they can’t monetize them? And for how long do you think they will do this? Do you expect Meta/Google to both grow their own revenues substantially to support their ability to allocate capital toward NBIS? Will their investors approve of these expenditures?
I suspect that by the time the contract is over, Nebius will already be capacity constrained by the Higgsfields and Revoluts of the world, in addition to big pharma, finance, and governments.
The hyperscaler deals were always a funding source. Simply a means to an end. Nothing more.
quote:This is a legitimate question, but I don’t think it’s likely that they fail, and if it appears they will, they will be back stopped.
NBIS isn’t contracted to OpenAI. But if OpenAI were to fail, that would have big impacts on all the Data Center builders because at that point there would be excess supply on the market. Does that concern you?
It’s just the reality.
Nebius won’t be unscathed, but they’ll make it through and be an even better long term investment if that were to happen. When people talk about Neoclouds folding and getting acquired, they’re not talking about Nebius Group.
They’re talking about the likes of KEEL and CIFR.
Posted on 9/1/26 at 8:02 pm to NotStupid
quote:Surely you know that one of Nebius Group’s fully owned subsidiaries is Avride, a US based autonomous vehicle and robot delivery company.
How are META and Google (the two largest NBIS customers) going to create a physical AI product that will be ran thru the NBIS data center. What’s the product? And what do you imagine it doing?
NBIS has been actively seeking Robotics/Physical AI startups that they think will blossom into the likes of Higgsfield AI. This is the main business as the exponential growth of select customers continually fuels Nebius revenues.
Loading Twitter/X Embed...
If tweet fails to load, click here.Posted on 9/1/26 at 8:02 pm to NotStupid
quote:
How are META and Google (the two largest NBIS customers) going to create a physical AI product that will be ran thru the NBIS data center. What’s the product? And what do you imagine it doing?
Unlike CRWV and other Neoclouds, Nebius is planning for the hyperscalers to be about 10% of their business by 2031. Already, the hyperscalers buy less than half of Nebius's capacity, as opposed to 80% of other neoclouds.
Thats the next phase of AI. Do you think Meta and Microsoft are going to be the big spenders forever, and when they stop buying, AI will wither on the vine?
For example, self driving cars are going to be a thing. Those cars will use data center to train and keep their on board AI systems. That is a physical AI application. Ford, GM, Toyota, etc, will all have their own AI systems, and Nebius will be an option for them.
Nebius has positioned itself as having the easiest and most efficient software package available, and that makes a difference when you have small to medium companies trying to use AI for sales, design, logistics, etc. Their AI has security and regulatory and SOX compliance already baked in.
Small companies can use Nebius tokens without having to dedicate a lot of resources to buy software, storage, and system architecture expertise. This is good for all of us, because otherwise, large companies will completely take over sectors of our economy, because they will be able to everything so much more efficiently. I'm reminded of the scene in Demolition man where Sylvester Stallone learns that all restaurants are "Taco Bell".
Nebius is positioning for the enterprise phase where everyone else jumps into the AI arena.
Posted on 9/1/26 at 8:03 pm to bayoubengals88
Posted on 9/1/26 at 8:06 pm to bayoubengals88
Posted on 9/1/26 at 8:57 pm to bayoubengals88
Grubhub already uses robot campus delivery They have been around for years. They didn’t need to be trained by AI models either.
Posted on 9/1/26 at 9:04 pm to NotStupid
Damn. That’s your takeaway from our responses?
And I was hoping for some real discussion…
And I was hoping for some real discussion…
Posted on 9/2/26 at 8:00 am to NotStupid
You realize that Grubhub now uses AI now, right? And that when they weren't using AI, they were not autonomous - they had someone with a joy stick driving it remotely, right?
The whole point of autonomous vehicles is to have AI drive them instead of humans.
Companies are going to invent physical products that require AI. It won't be Meta/icrosoft/Google that develops those products. It will be GM, GE, Boeing, Caterpillar, Samsung, and eventually, Bob's bicycle shop.
Companies that don't have the resources to buy the hardware and hire the expertise to build their own agentic AI will use Nebius to develop it.
Once the LLMs are on autopilot, the enterprise models will utilize a combination of LLMs and their own AI models to help improve their business. The will train theory own AI based own their data and business model - not a shared data set (can't have the Colonel's original recipe available on Grok).
Once businesses realize how they can use AI for a competitive advantage, the demand will increase. Meta, Microsoft, AND Nebius will be building data centers to meet that demand.
We're in the 2nd or 3rd inning of this game.
The whole point of autonomous vehicles is to have AI drive them instead of humans.
Companies are going to invent physical products that require AI. It won't be Meta/icrosoft/Google that develops those products. It will be GM, GE, Boeing, Caterpillar, Samsung, and eventually, Bob's bicycle shop.
Companies that don't have the resources to buy the hardware and hire the expertise to build their own agentic AI will use Nebius to develop it.
Once the LLMs are on autopilot, the enterprise models will utilize a combination of LLMs and their own AI models to help improve their business. The will train theory own AI based own their data and business model - not a shared data set (can't have the Colonel's original recipe available on Grok).
Once businesses realize how they can use AI for a competitive advantage, the demand will increase. Meta, Microsoft, AND Nebius will be building data centers to meet that demand.
We're in the 2nd or 3rd inning of this game.
Posted on 9/2/26 at 3:59 pm to Jax-Tiger
So your investment thesis is that “someone will invent a product that needs AI”… but you don’t know what this product is. We’ve clearly moved on from Chatbots (bc open source makes them a no go), so now it’s imaginary physical products. Great. And that’s why companies like NBIS are being shorted.
The AI industry is like a car that takes $25 a gallon gas to power it. If the cost to power the technology far outweigh any efficiency gains from its use, then you are left with a ton of poor investments. Even if the techology can do some cool things.
The AI industry is like a car that takes $25 a gallon gas to power it. If the cost to power the technology far outweigh any efficiency gains from its use, then you are left with a ton of poor investments. Even if the techology can do some cool things.
Posted on 9/2/26 at 4:24 pm to NotStupid
quote:
So your investment thesis is that “someone will invent a product that needs AI”… but you don’t know what this product is.
It's happening every day...
Caterpillar and FieldAI Advance AI-Powered Industrial Innovation
News provided by
Caterpillar Inc.
Sep 02, 2026, 09:15 ET
Leveraging robotics, digital twins and NVIDIA technologies, the collaboration helps create safer, smarter and more productive operations
IRVING, Texas, Sept. 2, 2026 /PRNewswire/ -- As industries face growing labor shortages and increasing productivity demands, Caterpillar Inc. (NYSE: CAT) is collaborating with FieldAI to advance physical AI, autonomy and robotics that help jobsites and factories operate safer and more efficiently. By turning real-time observations into actionable insights, physical AI helps improve decision-making, enhance safety and maximize the effectiveness of every worker and machine.
Digital twin technology delivers enhanced visibility into facility operations, helping teams identify risks earlier and optimize performance.
Caterpillar is harnessing physical AI to help create safer jobsites and enable faster, more informed decision-making for customers.
"The future of our industries will be shaped by how effectively we combine human expertise with AI-powered machines," said Jaime Mineart, Caterpillar chief technology officer. "Collaborations like this help accelerate the journey. We're focused on applying AI where it matters most, helping customers improve safety, increase productivity and get more value from every operation."
Posted on 9/2/26 at 4:27 pm to NotStupid
quote:
. We’ve clearly moved on from Chatbots (bc open source makes them a no go), so now it’s imaginary physical products.
There's no way you're serious right now. quote:LINK
RoboForce builds Robo-Labor for dirty, dangerous and repetitive industrial work across solar, data centers, shipping, mining, manufacturing and logistics. With Nebius AI Cloud, NVIDIA Blackwell infrastructure, scalable storage and expert support, RoboForce reduced AI setup pipeline time by 70% and cut iteration cycles from months to days.
Posted on 9/2/26 at 4:29 pm to NotStupid
quote:Do you seriously not grasp the current and expanding market for EVs, drones, and automated industrial and agricultural heavy equipment?
We’ve clearly moved on from Chatbots (bc open source makes them a no go), so now it’s imaginary physical products.
Posted on 9/2/26 at 5:44 pm to bayoubengals88
quote:
Do you seriously not grasp the current and expanding market for EVs, drones, and automated industrial and agricultural heavy equipment?
No I don’t. I’m going to enjoy watching you lose all your money. Along with the rest of the 60,000,000 short shares in NBIS.
Markets work if they are efficient.
Or if it’s the govt that doesn’t care if they lose money. It’s how TSLA kept an absurd valuation essentially govt subsidies. But TSLA is one company. The govt can’t create environmental subsidies for AI. If anything AI has the exact opposite effect. So government spending isn’t the backstop or too big to fail for the AI complex
Right now, the cost of the techology (and all the various inputs into it), are far higher than the output the technology produces.
For example, let’s say Uber’s engineers like using Anthropic’s AI tools for $100. And at $100, it helps Uber’s bottomline. But, it costs Anthropic $400 to power these tools. Something has to give. Either the customer has to pay more, or Anthropic’s needs to get far more efficient. There’s no market equilibrium. We are almost 4 years since ChatGPT launched, none of this has happened. It won’t happen either. That’s why none of these things that you are imagining are going to scale, and in NBIS you are holding shares in the biggest risk of the entire chain, the data center builder.
Who pays the Data Center builders? Companies that are either highly unprofitable (Anthropic’s/OpenAI) or hyperscalers that have now ran out of FCF and are tapping debt markets. That’s the signal to get out.
If you are bullish on AI, go buy a power company.
Posted on 9/2/26 at 5:52 pm to NotStupid
quote:I’m failing to understand why you would enjoy watching a middle class American on a message board lose money. It seems a bit weird. Like, at the very least, aren’t there better things to do?
I’m going to enjoy watching you lose all your money.
Posted on 9/2/26 at 5:54 pm to NotStupid
Notstupid is so much smarter than Goldman and JP Morgan. I better sell
Posted on 9/2/26 at 6:06 pm to NotStupid
Generally speaking, when people make shite up to serve their argument, it’s not a sign of being not stupid.
On cost-to-serve, Anthropic spent about $0.71 in compute per dollar of revenue in Q1 2026 and a projected $0.56 in Q2. Not $4.
Blended gross margin went from negative 94% in 2024 to the mid-60% range, with the API business exceeding 80%. The labs lose money on a net basis because they pour everything into training frontier models, not because serving a query costs 4x its price.
The training expenses are smart business that will pay off later.
Bottom line, you think you’re smarter than the billionaires
quote:
For example, let’s say Uber’s engineers like using Anthropic’s AI tools for $100. And at $100, it helps Uber’s bottomline. But, it costs Anthropic $400 to power these tools. Something has to give. Either the customer has to pay more, or Anthropic’s needs to get far more efficient.
On cost-to-serve, Anthropic spent about $0.71 in compute per dollar of revenue in Q1 2026 and a projected $0.56 in Q2. Not $4.
Blended gross margin went from negative 94% in 2024 to the mid-60% range, with the API business exceeding 80%. The labs lose money on a net basis because they pour everything into training frontier models, not because serving a query costs 4x its price.
The training expenses are smart business that will pay off later.
Bottom line, you think you’re smarter than the billionaires
Page 477 of 499
Popular
Back to top
Follow TigerDroppings for LSU Football News