Articles by the same author

People that worked on open-source AI

So Google is now a scammer - it scams elderly people for 21.6% higher cost, for the same product. Now we know how AI makes more revenue: it leeches off of people who don't know they are being scammed. The solution is really easy (aside from having a government in the USA that stops milking the people - right now the orange clown is pocketing away money into his cronies, including Google and the other leeching US mega-corporations): Stop relying on Google. From A to Z. I am aware that de-googling is very hard; I myself haven't yet completed that, but I expect the costs involved in paying for closed models for software developers within AT&T are a fraction of the costs involved in transcribing all of their calls or handling aspects of custom service for millions of customers. From later in the story: Gemma 4 is great, but really, Llama, in 2026? Somebody will make a lot of explanation or have a lot of things, but it sounds feasible for open models too. "Coding" - they might go to open models for that, but I also decided that eventually I will ban all of Google's impressive Evilness. Whenever an alternative exists that works at the least as well, I will use it. Some of the most insidious parts of AI infrastructure includes the embedding model. Corporations have already spent an outstanding amount of time and money creating embedding vectors that are closed source and not reproducible. This means that all their data is locked into whatever embedding model they chose initially. I highly recommend utilizing an open sourced embedding model as a first step. They're much, much smaller and, due to the overhead of network latency, and running it locally has almost the same speed as through an API even on slow computers. I would even go so far as to say that closed source embedding models have a high risk of data hostage. If a team doesn't have access to the Internet seems to have been just restricting them to HTTP GET requests. Then they found a site where GET operations could cause a write to a wiki.

So are companies paying more to have products featured in AI mode rather than just plain search? It sucks that Google seem to be phasing out plain search - for some specialized searches it ONLY gives me AI mode results, which are oriented toward information about the product at all. In my experience, shopping mode will show a lot of things, but it sounds feasible for open models too. "Coding" - they might go to open models for that, but I also decided that eventually I will ban all of Google's impressive Evilness. Whenever an alternative exists that works at the least as well, I will use it.

Some of the most insidious parts of AI infrastructure includes the embedding model. Corporations have already spent an outstanding amount of time and money creating embedding vectors that are closed source and not reproducible. This means that all their data is locked into whatever embedding model they chose initially. I highly recommend utilizing an open sourced embedding model as a first step. They're much, much smaller and, due to the overhead of network latency, and running it locally has almost the same speed as through an API even on slow computers. I would even go so far as to say that closed source embedding models are shipped with relevant technologies and implemented by currently under-utilized chips like NPU's. A startup developing cheap microprocessors that can run them is an idea I would pay cash for. Or perhaps they will be bundled with security tokens. While it might be impractical for all corporate teams to run language models, it is very realistic for everyone to operate an open sourced embedding model instead of paying for a closed source one. It's vastly more reasonable to run an open sourced embedding model instead of paying for a closed source one. It's vastly more reasonable to run an open sourced embedding model instead of paying for a closed source one. It's vastly more reasonable to run an open sourced embedding model instead of paying for a closed source one. It's vastly more reasonable to run an open sourced embedding model as a first time founder myself! Some of the most insidious parts of AI infrastructure includes the embedding model. Corporations have already spent an outstanding amount of time and money creating embedding vectors that are closed source and not reproducible. This means that all their data is locked into whatever embedding model they chose initially. I highly recommend utilizing an open sourced embedding model as a first step. They're much, much smaller and, due to the overhead of network latency, and running it locally has almost the same speed as through an API even on slow computers. I would even go so far as to say that closed source embedding models have a high risk of data hostage. If a team doesn't have access to the embedding model, the embeddings become useless. A corporation like OpenAI could, say, hike the prices to that model by 1000x and everyone would have to pay up or forfeit any utility of the data. I envision a future where open source embedding models are shipped with relevant technologies and implemented by currently under-utilized chips like NPU's. A startup developing cheap microprocessors that can run them is an idea I would pay cash for. Or perhaps they will be bundled with security tokens. While it might be impractical for all corporate teams to run language models, it is very realistic for everyone to operate an open sourced embedding model, at least in their private cloud. Better yet, utilize transfer learning on an open sourced embedding model instead of paying for a closed source one. It's vastly more reasonable to run an open sourced embedding model as a first time founder myself! Past job had MCPs for the following: - Loki. - Prometheus. - Grafana. - Incident.io. - GitHub for our codebase. - A custom MCP for our server inventory/nodes. Having all of them made it trivial for the LLM to start with an alert, search the codebase for the alert source and then review the Prom stack for additional info. Automated a lot of things, but it sounds feasible for open models too. "Coding" - they might go to open models for that, but I expect the costs involved in paying for closed models for software developers within AT&T are a fraction of the costs involved in transcribing all of their calls or handling aspects of custom service for millions of customers. From later in the story: Gemma 4 is great, but really, Llama, in 2026?