| ▲ | 2001zhaozhao 7 hours ago |
| This couldn't be more true. Companies will be driven by AI harnesses (as defined by this article) that automate decision-making and prioritization. In the medium term, may the company with the best harness win. In the longer term, the downstream impact is massive commoditization of software and invalidation of most existing moats. Data moats are gone if you can simulate the data with AI. Even platform effects can be sidestepped if AI replaces one side of the platform. In addition, while right now agile startups have the advantage, at some point the balance will start tilting towards whoever has the most tokens (OR perhaps durable moats will trump even near-infinite tokens; we will have to see). Startups have a limited time window to have whatever impact in the world they are hoping to have, or to build a moat that won't be disrupted by AI, but there are few of them left in the world. The upside is that when there is a lot of commoditization, then the consumer benefits. |
|
| ▲ | vanuatu 7 hours ago | parent | next [-] |
| imo there are a few real moats left assuming no superintelligent RSI scenario - capital, as money is scarce - network effects, as human attention is scarce - relationships, as human attention is scarce - research talent, assuming there exists some field(s) that AI is unable to surpass the best researchers - proprietary data and sensors, as systems of record and action are scarce (training data, on the other hand, can maybe we simulated, but I'm pretty bearish in general on the idea of fully simulated data) |
| |
| ▲ | ForHackernews 6 hours ago | parent [-] | | Trust. I trust the Debian maintainers. I don't trust OpenAI. If AI levels out most other distinctions, maybe in the end we choose to give money to people we trust. |
|
|
| ▲ | ForHackernews 7 hours ago | parent | prev | next [-] |
| > Data moats are gone if you can simulate the data with AI. This is a hilarious premise if you work in a domain where it matters even a little bit whether the data is correct or not. |
| |
| ▲ | epistasis 7 hours ago | parent | next [-] | | I work in science, the data is becoming far bigger of a moat than it ever was before because of this. Every company can now apply the latest and greatest analysis. Data generation is where the cost is. It's where the time was spent, time that can never ever be retrieved at any cost. AI won't solve biology, make a pathogenic virus, etc, without tons and tons of data, of both types we know and types we have not yet figured out how to generate. Perhaps the area where AI has the most to help bio is in figuring out novel measurement technology. But it's not going to be able to reason or deep-net its way to figuring out systems for which we can't even measure the parts. | | |
| ▲ | 2001zhaozhao 7 hours ago | parent [-] | | Yep private scientific knowledge is a massive new moat that you can pull if you simply invest in scientific discovery methods in general and throw enough resources and tokens at the problem. I think there are already startups specifically trying to do this. The main obstacle is whether you're actually able to pull significantly ahead of (AI-enabled) public science to make a difference, but I guess the math works out if you're sufficiently AGI-pilled. I agree this is a kind of data moat, but it's also arguably distinct enough to be its own thing. |
| |
| ▲ | 2001zhaozhao 7 hours ago | parent | prev | next [-] | | You are assuming that getting the AI to generate correct (or accurate, representative) data will be very difficult. I would agree, but I think it will become possible in the long run. (Edit: alternatively you just use AI to get rid of the need for data to solve a problem, like Jev did for traditional classification models) I think current incentives definitely go against any efforts to build this. It's very hard to build this and be rewarded for it by, say, investors or your boss, because you can't really prove that your system is non-sloppy while your competitor's is (even if being non-sloppy is all that matters), because by definition your novel results are not verifiable or else the model labs will have already trained it into their model. But the same is true for high-quality AI systems in general. In general, I think AI model advancements will make the systems easier and easier to build until some small guy accountable to no one but themselvs can build it, and then it will actually be built. | |
| ▲ | theturtletalks 7 hours ago | parent | prev [-] | | There’s 2 arguments. 1. He’s talking about training new models and at one point, having data was valuable. Now synthetic data is being used to train models 2. Companies like SalesForce who’s moat is having all your customer data so you’ll be locked in. You could extract it but you’d have to clean it and then change it to your new schema. With LLMs, you can do that in minutes and even use SalesForces MCP or API to get all your data and leave. It’s exactly why companies like Figma are gate keeping their MCP. They know that swapping their MCP with Paper’s or any new one is easy. The moats are evaporating as we speak. Distribution is one of the smaller ones left, but the personal software trend might eat that too. | | |
| ▲ | 2001zhaozhao 7 hours ago | parent [-] | | > Distribution is one of the smaller ones left, but the personal software trend might eat that too Being a platform for personal software is gonna be valuable, but it needs a lot of trust. (I have a nonprofit idea around this right now) Btw, I think distribution might temporarily become less important (because with better AI you can actually pull so far ahead of competitors quality-wise and therefore succeed despite a distribution drawback), but long run it actually becomes more important because of AI persuasion and commodification? If you are the super app then, well, you are the super app |
|
|
|
| ▲ | what 2 hours ago | parent | prev [-] |
| Maybe stop to think why they are selling you tokens and the promise of being able to recreate and outcompete every business instead of just doing it themselves. |
