12 ms·

Have you considered the possibility that 1 million similar companies worth $100 each will appear? I know nothing about product, but think how many people could

Have you considered the possibility that 1 million similar companies worth $100 each will appear? I know nothing about product, but think how many people could have done it with the same tools as you.

gametorch 1y ago
I think you have a super good point. Cost of anything that is purely driven by code is falling like a rock now. I don't have an answer to this question. I don't have an answer to the moat question. Best guess is some sort of "network/platform value" upon which you deliver the commoditized output of a model.
- msgodel 1y ago
  Code was always cheap, anyone could write it and it's really not hard to learn. I don't think LLMs change that.
  - alganet 1y ago
    > Code was always cheap It is more correct to say that cheap code always had a large supply compared to expensive code, and LLMs haven't changed that. Expensive code, though, is expensive. LLMs also haven't changed that. Expensive code accurately and efficiently encapsulates knowledge. There is no horizon for humans getting rid of this need.
- alganet 1y ago
  I know nothing about moat or network or anything like that. This is basic stuff. People have money. People use money to buy things. Money is limited. It can move, and shift, but not grow. People earn X and will spend a portion of that on stuff on different things. If there's a lot of similar competitors of similar cost and quality, they'll share the money pool (what people are willing to spend on your kind of thing). AI is, therefore, irrelevant. It drives your costs down, but it also drives the cost down for everyone. Any competitor can use it to achieve similar results. If you base your production on it, you get no edge. What can you do that is unique and you can offer to your potential customers?
byoung2 1y ago
An app like an AI resume customizer based on job description will quickly devolve into a million clones. Something with a network and a first-mover advantage would have some staying power even if people clone it. For example an app where you post a job description and it sets up an AI voice agent as a virtual employee ready to be onboarded. once you have thousands of of virtual employees with company email addresses and attending zoom meetings, it will be hard for clones to break into that vertical
- alganet 1y ago
  Quality and low cost are the hardest things to clone. It's really not about features. Ideally, you want a lot of clones, but all of them more expensive or inferior quality than you. That's something that can beat any first-mover advantage.

gametorch 1y ago

I think you have a super good point. Cost of anything that is purely driven by code is falling like a rock now. I don't have an answer to this question. I don't have an answer to the moat question. Best guess is some sort of "network/platform value" upon which you deliver the commoditized output of a model.

msgodel 1y ago
Code was always cheap, anyone could write it and it's really not hard to learn. I don't think LLMs change that.
- alganet 1y ago
  > Code was always cheap It is more correct to say that cheap code always had a large supply compared to expensive code, and LLMs haven't changed that. Expensive code, though, is expensive. LLMs also haven't changed that. Expensive code accurately and efficiently encapsulates knowledge. There is no horizon for humans getting rid of this need.
alganet 1y ago
I know nothing about moat or network or anything like that. This is basic stuff. People have money. People use money to buy things. Money is limited. It can move, and shift, but not grow. People earn X and will spend a portion of that on stuff on different things. If there's a lot of similar competitors of similar cost and quality, they'll share the money pool (what people are willing to spend on your kind of thing). AI is, therefore, irrelevant. It drives your costs down, but it also drives the cost down for everyone. Any competitor can use it to achieve similar results. If you base your production on it, you get no edge. What can you do that is unique and you can offer to your potential customers?

byoung2 1y ago

An app like an AI resume customizer based on job description will quickly devolve into a million clones. Something with a network and a first-mover advantage would have some staying power even if people clone it. For example an app where you post a job description and it sets up an AI voice agent as a virtual employee ready to be onboarded. once you have thousands of of virtual employees with company email addresses and attending zoom meetings, it will be hard for clones to break into that vertical

alganet 1y ago
Quality and low cost are the hardest things to clone. It's really not about features. Ideally, you want a lot of clones, but all of them more expensive or inferior quality than you. That's something that can beat any first-mover advantage.