Eldodi

With the v2 spec, MCP became a lot closer to APIs by becoming stateless. And there is now a push to use HTTP verbs more extensively to improve caching even better in v3. MCP is converging into APIs but wit great auth and auditing

A shame agents will never read this, just like they almost never read llms.txt or try to get the .md version of your html pages!

Is it really contrarian though? Cohere, Aleph Alpha and many other model companies went that route and are doing well

MCP usage is at an all-time peak - 17M daily (!) downloads of the official SDKs.

We're using MCP for virtually all our SaaS and infra providers (CRM, ATS, banking and accounting, Notion, Linear, all dev tools and hosting platforms) it's become a hard requirement in our buy bechmark. No MCP = NOGO

The advantages are:

- baked-in auth: no need to configure anything, it just works with our employees accounts

- one click install with Codex Plugins and Claude Connectors

APIs and WebMCP don't have the same purpose:

- WebMCP allows agents to assist humans on their own interface ie the web UI.

- APIs bypass the human interface completely

And I'm not sure most of the work today is done on legacy desktop apps. We're in 2026, not 2016

it's not a question of model size, it's a question of your model API engineering. The initial problem was that the all tools schemas needed to be loaded in the model's initial system prompt, and changing the tools list in the middle of a session was busting the cache. Any model can now overcome that with a Tool Search tool implementation

you can actually invoke an MCP App by name in both Codex and Claude using the @ handle

MCP context bloat is a solved problem since at least February. Both OpenAI and Anthropic support Client-side MCP tool search, which makes MCP context loading as efficient as the Skills one (progressive discovery).

Code mode is great but not needed anymore in 95% of use-cases.

- CLI tools don't have standardized auth, nor do they have granular tool permissions. MCP does

- CLI tools need to access your secrets in your .env. With MCP the model never sees the secrets and refresh tokens.

- CLI tools don't have auditing and telemetry. MCP provides all this for free.

- CLI tools imply full bash access, a nightmare in terms of security and very costly if you have to spin up a machine for every request for billions of users.

- etc

Models benchmarks start to get saturated again!

Dolby is only the most recent case, Sisvel consorsium actually bills licences per device:

Consumer Display Device: EUR 0.32

Consumer Non-Display Device: EUR 0.11

(source here: https://www.sisvel.com/licensing-programmes/audio-and-video-...)

How is AV2 expected to avoid the patent-pool issues AV1 ran into?

AV1 was designed as royalty-free, but Sisvel’s pool and the recent Dolby/Snap proved the contrary.

https://accessadvance.com/2026/03/24/access-advance-licensor...

I was at the event, and was impressed by the attendance, all the leaders from the major european listed companies were there.

Also interesting to note the number of partners they invited. Going from Microsoft, Accenture and EY to startups like alpic.ai or lingo.dev . Seems like they are ramping up their M&A game too

The protocol becoming stateless is a huge step forward!

Some might argue that Adwords got so successful because ads competed like search results, on bid AND relevance, not just bid.

If your ads inventory is big enough, ads can actually be a better answer to your intent than organic content, because the companies behind the ads have a much stronger incentive to satisfy your need.

It will be interesting how hidden those ads will be compared to current Search experience or what OpenAI is already doing.

It's a lot easier to mislead a user with an AI generated ad that with a Search result IMHO, I'm betting on a huige backlash if they don't make it VERY clear that ads are ads.

Crazy this company will IPO for >1B with such bad financials! That said, Starlink seems to be a real cash machine, not as good as ads but enough to support AI bets.

2025:

- Revenue: $18.7B, up from $14.0B in 2024

- Operating loss: -$2.6B

- Net loss: -$4.9B

- Adjusted EBITDA: $6.6B

- Operating cash flow: $6.8B

- Capex: $20.7B

Segment breakdown:

- Starlink / Connectivity: $11.4B revenue, $4.4B operating income, $7.2B adj. EBITDA

- Space / launch: $4.1B revenue, -$657M operating loss

- AI / xAI / X: $3.2B revenue, -$6.4B operating loss

Starlink metrics:

- Subscribers: 8.9M at end-2025, 10.3M by Mar 31 2026

- ARPU: $99/month in 2023, $81 in 2025, $66 in Q1 2026

Balance sheet as of Mar 31 2026:

- Cash: $15.9B

- Marketable securities: $7.8B

- Total assets: $102.1B

- Total liabilities: $60.5B

- Debt / finance leases: about $30.3B

I guess they forgot to run Mythos on their own code!

Looks closer to a Codex on a mac mini

Is this report AI generated? It feels so aí sloppy to read, even if the information inside is insightful, I just can't take it seriously

It will be interesting if Mistral succeeds to keep up with US and Chinese labs in terms of Models, or if they just become an integration company of chinese OSS frontier labs, like more and more of their competitors

Most of MCP servers and Apps are way under-designed today. A lot of MCP B2B servers still wrap legacy APIs, and most MCP Apps try to reproduce a website experience instead of trying to reinvent the experience from scratch.

It feels like we're in the early mobile years where companies have not figured out what to do with this new technology. I hope the Uber and Candy Crushes of the AI era will land in 2026! (well maybe not candy crush, but some IA native games would be nice)

Anthropic is really good at releasing features that are almost the same but not exactly the same as other features they released the week before

"MCP is less discoverable than a CLI" -> not true anymore with Tool_search. The progressive discovery and context bloat issue of MCP was a MCP Client implementation issue, not a MCP issue.

"MCPs needs to be wrapped to be composed." -> Also not true anymore, Claude Code or Cowork can chain MCP calls, and any agent using bash can also do it with mcpc

"MCPs needs to implement stateful behavior, shell + cli gives it to you for free." -> having a shell+cli running seems like a lot more work than adding a sessionId into an MCP server. And Oauth is a lot simpler to implement with MCP than with a CLI.

MCP's biggest value today is that it's very easy to use for non-tech users. And a lot of developers seem to forget than most people are not tech and CLI power users