They're So Desperate Now...

Summary of Technological Concepts / Product Analysis

The speaker (Mudaher) discusses a rumored internal Microsoft initiative called “Windows K2”, described as Microsoft’s response to rising competition—particularly from Linux and Valve’s Steam/SteamOS ecosystem. The central claim is that Windows feels bloated, largely due to AI-driven features and web-based / Electron-style app components. K2 is positioned as a strategy to improve performance, reliability, and resource usage.

---

1) Why Apps Consume Too Many Resources (WebView / Electron Model)

• Examples mentioned: WhatsApp and Discord.
• Explanation: many “apps” are essentially browser engines wrapped inside native shells (e.g., Chrome WebView / Electron).
• Result: launching them spawns multiple Chrome-like processes, which is described as inefficient and contributes to Windows feeling sluggish and consuming more memory/CPU.

---

2) K2’s Proposed Direction: More Native Apps, Less “Bloat”

K2 is described as aiming to reduce default overhead by shifting away from heavy wrapper-style applications. The speaker claims Microsoft wants to:

• stop “bullying” or pushing Copilot everywhere
• build more native applications (not just webview/Electron wrappers)
• optimize startup and performance
• reduce overall OS/application overhead

---

3) Debloating AI Features (Especially “Copilot Everywhere” UX)

The speaker highlights user backlash against Microsoft’s Copilot/AI presence across the OS UI—even in basic tools like Notepad.

• AI shouldn’t fully disappear, but default exposure should be reduced
• features should be opt-in rather than enabled by default
• installation should support a basic vs full style experience (similar to how some Linux distributions let users choose feature sets)

---

4) Gaming as a Major Priority; Targeting Steam/SteamOS Performance

K2 is presented as prioritizing gaming compatibility, since Windows currently tends to lead.

A key claim:

• Microsoft wants Windows performance to reach “parity with SteamOS” on handheld-class hardware.

Conceptual benchmark discussion:

• Linux/SteamOS on similar handheld hardware allegedly shows higher average FPS in some examples (e.g., “Returnal” cited)
• Windows results may vary depending on drivers/settings

---

5) Handheld Gaming PCs and UX Improvements

The speaker mentions Windows adding a full-screen Xbox interface (“Xbox mode”) to compete with Steam Deck/SteamOS controller UI.

• SteamOS UI is praised as cleaner for non-technical users because it makes launching games easier.
• The speaker also references AI-assisted upscaling conceptually (referred to as auto SR) as a potential visual improvement for handhelds.

---

6) Requirements for Success: Control, Offline Friendliness, Lighter Installs

The speaker argues K2 must deliver user-facing changes such as:

• No mandatory 100% internet connectivity for local use
• support for local Windows accounts
• the ability to install software normally
• a snappy baseline gamer-friendly install
• feature selection at install time (basic vs full, with bloat/AI made opt-in)

---

7) Concerns About Trust and Execution

The speaker expresses low faith in Microsoft due to recent history of making Windows heavier, especially since Windows 11.

He argues Microsoft must provide:

• a real, lighter ISO
• measurable improvements
• not just promises

---

8) Bigger Strategic Context: Why Competition Matters

The speaker frames Linux’s mainstream traction as coming from:

• compatibility layers (e.g., Proton on Steam Deck)
• more user openness/control

He also notes Linux still has gaps, especially:

• online anti-cheat behavior
• modding friction

So, Microsoft’s best path is argued to be reducing those “excuses” by addressing performance and bloat.

---

9) Battery / Power Consumption on Handhelds

Final point: installing Windows on handhelds like Steam Deck / Legion Go reportedly increases baseline power consumption, reducing battery life.

• If Microsoft doesn’t fix this, handheld adoption could suffer.

---

Main Speakers / Sources

• Main speaker: Mudaher (referenced as “Me Mudahar”)
• Referenced external sources (as context/benchmarks, not directly quoted as primary data providers):
  • Valve / Steam Deck / SteamOS
  • Steam / Proton
  • “Benchmarking research” mentioned generically (no specific named study)