devil1432

Why would you use AI to format your files? It literally takes 1 second with shortcut.

Meanwhile, an illegal pirate version from torrents works flawlessly after installation and double-clicking on the exe file.

People focus too much on discs and cartridges, but too little on licensing and drm. The disc is meaningless if you cannot run it after drm servers died. Also, very often the game on the disc is unplayable because of lack of day-1 patch. Meanwhile you can burn games you bought from GOG on a cd and play it 10 years later without internet connection. So the physical vs digital discussion is completely irrelevant to the problem.

Looks like AI reached it's peak already. Every new architecture is just a new wrapper that has nothing to do with ml.

From my experience, the biggest flaw of mcp is lack of control over your system prompt.

Prompts need to be tailored for specific model (duh). Tool definitions are de facto part of your system prompt. By injecting tool definition from MCP servers into your prompt, you are basically adding prompt that was (likely) not tailored for your specific model. That causes drop in quality of answers.

Other issue is that adding one new feature via MCP can introduce regression into your system and cause other MCP features to work incorrectly (and I am not talking about malicious prompt injection by MCP provider).

Real life example: tools from MCP server for ms outlook requires certain date format for filtering emails. That worked fine in our prod. Until we added another MCP server (built in-house in another department of our company) that required dates in different format. After adding these tools, now our agent started making mistakes: putting dates in outlook format in our internal mcp tool and vice versa - putting dates in our internal format into outlook tool. There is no obvious way to fix it other than separating these prompts. And that goes against standard MCP architecture and consumes way more tokens to cycle through multiple prompts and agents.

If that's already the issue with two MCP servers, then this architecture clearly does not scale.

I wonder if these github failures are just systematic incompetence or MS cutting budget on purpose to promote its own cicd tools

I remember when I was 12yo, I used to create a lot of small fan pages, blogs and forums (with total of 10 users if I was lucky). They were always indexed by Google with no problem. I miss these simpler times. Nowadays, Internet is no longer democratic.