socrateslee

> We can know more than we can tell.

The paradox arises when humans develop SaaS, apps, workflows, etc. The question is "will AI simply do things or know things more than we can tell?"

SEO became a standardized industry discipline pretty much after Google dominated the market and stabilized the rules of engagement. The search engine giants evole their ranking algorithms mostly to fight exploiting of their mechanims, the core principles of SEO don't change much.

In contrast, LLMs are still constantly evolving. Not only the models, the way of agents doing things, the tools/skills agent shall use, the whole system may be completely different next month. GEO is merely an attempt to ride the LLM hype，clumsily stitching together unrelated concepts.

Gemini 3 on Gemini CLI (free version) would meet quota limit for about 3-4 messages, but it will take much longer time since it responses pretty slow.

in https://android-developers.googleblog.com/2025/09/lets-talk-... install an app via adb is not affected, seems that means the wrapper software of adb will prosper.

Just to think about the blocking of installing apps from unverified source on Android by Google, isn't this an exclusive contracts upon users?

It's like double "double the dose"

This article will show you how to call a MCP server in the shell, without mcp dev or any third party tools, only with echo > or copying/pasting JSONRPC message directly.

For the OAuth part, the access_token is all an MCP server needs. So users could do an OAuth Authorization like in the settings or by the chatbot, and let MCP servers handle the storage of the access_token.

For remote MCP servers, storing access_token is a very common practice. For MCP servers hosted locally, how to deal with a bunch of secret keys is a problem.

https://crt.sh can find your subdomain only when it doesn't have a wildcard certificate(*.somedomain.com)

> In the late 20th century, democracies usually outperformed dictatorships, because they were far better at processing information. We tend to think about the conflict between democracy and dictatorship as a conflict between two different ethical systems, but it is actually a conflict between two different data-processing systems. Democracy distributes the power to process information and make decisions among many people and institutions, whereas dictatorship concentrates information and power in one place. Given 20th-century technology, it was inefficient to concentrate too much information and power in one place.

> However, artificial intelligence may soon swing the pendulum in the opposite direction. AI makes it possible to process enormous amounts of information centrally. In fact, it might make centralized systems far more efficient than diffuse systems, because machine learning works better when the machine has more information to analyze. If you disregard all privacy concerns and concentrate all the information relating to a billion people in one database, you’ll wind up with much better algorithms than if you respect individual privacy and have in your database only partial information on a million people.

May be the case is that filtering user by browser UA is no longer a feasible solution(new browsers and alike are growing), and neither running javascript(headless chrome everywhere).

For local physical store, geo-location is a naturally filter for customers as long as beaming a person from a spaceship to earth is not invented. For web, a equally effective solution is very hard to find.

The VCs thinked that the scaling law may build a moat for very large computing powers like Google and OpenAI, and VCs rushed into the area the foundation models.

The emerging of DeepSeek V3 and DeepSeek R1 somehow make the moat vulnerable with their much smaller training cost. The fact verified the claim of Google.

I agree that LLM could do most of the moderation work. You could use a multi-modal LLM for image moderation.

AI code copilot like cursor provide a immersive context than most of other AI products.

or saying that most of the Google engineers are using tools like copilot, and they use the copilot just as everyone else.

When we say “the contextvars is somehow like thread local, but mostly for coroutines”, we think that coroutines are just another types of threads. But actually we are not right. The coroutines are not threads. A thread has its own life cycle and its own stack space, so a thread local storage actually belongs to a thread. But a coroutine is just a function that can be paused and resumed, scheduled by a loop. So the context of a coroutine belongs not to the coroutine itself. When in a chain of coroutines, like the calling chain of Tasks, we need to be aware that the consistency of context are manually kept by the users, not automatically kept by the system.

Anyway, just note that Google do encourage RSS with the Google Discover feature

https://developers.google.com/search/docs/appearance/google-...

ChatGPT only swallows data or content but bring back no traffic to content creators. Maybe this is how Google is different from ChatGPT. But what will become of Google if the experimental generative AI's answer replace all the SERPs? An closed web where no search engine no AI can actually enter?

I have been thinking about two questions for a long time. First, why the market share of IE is so resilient to die? Second, why did IE lose to Chrome from the position of market domination while Windows has always been dominating the desktop OS market? Now I realized that the answers to the two questions are the same — the mechanism of the software update.