[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]  Kamahsutra [@TheKamaHsutra](/creator/twitter/TheKamaHsutra) on x 4137 followers Created: 2025-07-20 16:07:29 UTC It is apparently a touchy subject when it comes to $IREN being able to produce software stack that can rival top tier CSP providers such as $CRWV and $NBIS. The naysayers say it can't be done because it will take literally thousands of software engineers/developers and years to accomplish. But is that really the case? In the age of AI assisted coding, could $IREN develop a software stack that is capable of competing with top tier CSPs in a relatively short time and a much lower headcount? That is a question I asked myself several weeks ago, and asked ChatGPT to model out a timeline and headcount for such an endeavor. The results were very surprising to me since the timeframe and headcount was far lower than I had expected. In recent conversations, I had brought up this subject speculating that this is indeed possible and not just fantasy. On X discussions and post with other people in the software development there seems to be a divide on how safe a software moat is in the age of AI assisted coding, or in the not distant future, agentic AI assisted coding. People such as @seacliffstweetr would agree that software as a moat is shaky as we move forward in AI development, while others who profess to having large scale software development such as @Mabiverse123 scoff at the idea. I am not a software developer by trade, nor have I ever been part of a large scale software development project. I was only curious on what such an endeavor would require. I recently ran this topic through the current X best LLMs to model out this endeavor. I am simply a messenger of their output and I have no biases on how accurate the output data is. This is the prompt that I used: If I am IREN and need to create a software stack that can rival that of CRWV. How long would it take, and what is the headcount needed. Also assume that the coding will be assisted using a platform such as Poolside AI. I don't think I phrased my prompt to be biased and sway it to a desired outcome of my choice. It is a very generic and neutral question. Here are the key parts of each of the LLMs in deep research mode, with the full output in links below. Again, I'm a not a software developer, and the outputs are from the LLMs while being asked a neutral question regarding the timeline and headcount need for this project. If you disagree with this output, it would be better to argue with the respective LLMs than me. 😉 GROK X (would have preferred GROK 4, but should be suffice) GEMINI XXX PRO (scroll to second prompt as this was done in a the same conversation as another question, however, this should not create any bias for this question) ChatGPT 4o I have used the upper bounds of what has been quoted for timeline in months and headcount in number of people to be on the conservative side of these models. The numbers are below: It is suffice to say that these numbers are simply shocking to most people who have an idea of software development times. I have not been involved in a software development project, however, I have worked in large organizations that do have large software internal software development teams and have rough idea on how labour intensive these projects were. With that said, we are entering the age of AI, and agentic AI shortly. Can we really expect such drastic reduction in software development cycles at such reduced headcounts? I do not know. What I think we do be is open minded to the possibilities of what AI can do for us in terms of productivity. Just X years ago, had I attempted to research this kind of topic on my own without AI, I would have been completely lost without much hope. Yet, today AI would give me a much thoroughly researched output in less than XX minutes each. What kind of wonders can AI do for software development? Stay open minded to the possibilities of the future. 😉  XXXXXX engagements  **Related Topics** [iren](/topic/iren) [coins ai](/topic/coins-ai) [stack](/topic/stack) [$iren](/topic/$iren) [$crwv](/topic/$crwv) [stocks technology](/topic/stocks-technology) [$nbis](/topic/$nbis) [Post Link](https://x.com/TheKamaHsutra/status/1946965194250293256)
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
Kamahsutra @TheKamaHsutra on x 4137 followers
Created: 2025-07-20 16:07:29 UTC
It is apparently a touchy subject when it comes to $IREN being able to produce software stack that can rival top tier CSP providers such as $CRWV and $NBIS. The naysayers say it can't be done because it will take literally thousands of software engineers/developers and years to accomplish. But is that really the case?
In the age of AI assisted coding, could $IREN develop a software stack that is capable of competing with top tier CSPs in a relatively short time and a much lower headcount? That is a question I asked myself several weeks ago, and asked ChatGPT to model out a timeline and headcount for such an endeavor. The results were very surprising to me since the timeframe and headcount was far lower than I had expected.
In recent conversations, I had brought up this subject speculating that this is indeed possible and not just fantasy. On X discussions and post with other people in the software development there seems to be a divide on how safe a software moat is in the age of AI assisted coding, or in the not distant future, agentic AI assisted coding. People such as @seacliffstweetr would agree that software as a moat is shaky as we move forward in AI development, while others who profess to having large scale software development such as @Mabiverse123 scoff at the idea.
I am not a software developer by trade, nor have I ever been part of a large scale software development project.
I was only curious on what such an endeavor would require. I recently ran this topic through the current X best LLMs to model out this endeavor. I am simply a messenger of their output and I have no biases on how accurate the output data is.
This is the prompt that I used:
If I am IREN and need to create a software stack that can rival that of CRWV. How long would it take, and what is the headcount needed. Also assume that the coding will be assisted using a platform such as Poolside AI.
I don't think I phrased my prompt to be biased and sway it to a desired outcome of my choice. It is a very generic and neutral question. Here are the key parts of each of the LLMs in deep research mode, with the full output in links below. Again, I'm a not a software developer, and the outputs are from the LLMs while being asked a neutral question regarding the timeline and headcount need for this project. If you disagree with this output, it would be better to argue with the respective LLMs than me. 😉
GROK X (would have preferred GROK 4, but should be suffice)
GEMINI XXX PRO
(scroll to second prompt as this was done in a the same conversation as another question, however, this should not create any bias for this question)
ChatGPT 4o
I have used the upper bounds of what has been quoted for timeline in months and headcount in number of people to be on the conservative side of these models. The numbers are below:
It is suffice to say that these numbers are simply shocking to most people who have an idea of software development times. I have not been involved in a software development project, however, I have worked in large organizations that do have large software internal software development teams and have rough idea on how labour intensive these projects were. With that said, we are entering the age of AI, and agentic AI shortly.
Can we really expect such drastic reduction in software development cycles at such reduced headcounts? I do not know. What I think we do be is open minded to the possibilities of what AI can do for us in terms of productivity. Just X years ago, had I attempted to research this kind of topic on my own without AI, I would have been completely lost without much hope. Yet, today AI would give me a much thoroughly researched output in less than XX minutes each. What kind of wonders can AI do for software development? Stay open minded to the possibilities of the future. 😉
XXXXXX engagements
Related Topics iren coins ai stack $iren $crwv stocks technology $nbis