Matt Wolfe

AI News: Microsoft Finally Reveals They're Plan!

⏱ 30 min video · 4 min read5 Jun 2026Worth watching

TL;DR

Matt Wolfe recaps a packed week of AI announcements from Microsoft Build and Nvidia Computex, covering 7 new Microsoft AI models, autonomous agent products, and a wave of third-party model releases. The video also includes an exclusive sit-down interview with Microsoft AI CEO Mustafa Suleyman on self-sufficiency, healthcare AI, and the ethics of model training.

Key points

Microsoft unveiled 7 in-house AI models including MAI Transcribe 1.5 (fastest and most accurate transcription model), MAI Image 2.5 (ranked #2 in image editing), and MAI Voice 2 (15-language speech model), signaling a strategic push toward independence from OpenAI.

Microsoft Scout is an always-on autonomous agent powered by OpenAI's open-source technology, integrated at the OS level across Windows, Teams, Outlook, OneDrive, and SharePoint.

GitHub Copilot launched a new standalone app offering a Codex-like IDE experience but with the ability to choose any model from any provider — though new plan signups were temporarily paused at time of recording.

Nvidia announced the RTX Spark chip, combining GPU and CPU with up to 128GB unified memory for running large local LLMs, previewed in the new Microsoft Surface Laptop Ultra.

Notable third-party releases included Nvidia Llama Nemotron 3 Ultra (550B open-weight model), Ideogram 4.0 (open-weight image model with native background transparency), Kling Imagine 1.5 (video with dialogue), and Miso One (open-source voice model described as highly emotive and near-indistinguishable from human).

Actionable insights

→

If you do vibe coding in Codex, keep an eye on the new GitHub Copilot standalone app — it offers the same experience but lets you swap in any model provider, which could be significantly more flexible once signups reopen.

→

For image editing tasks, MAI Image 2.5 is currently ranked #2 globally and is worth testing; Ideogram 4.0 is the top open-weight text-to-image model with native background transparency, making it useful if you need fine-tunable, self-hostable image generation.

→

Miso One is an open-source voice model claimed to be the most emotive available — if realistic AI voiceover for content creation matters to you, this is worth evaluating before paying for proprietary alternatives.

→

For enterprises, Microsoft Scout and Project Solara (AI agents in physical devices like badges and desktop units) signal Microsoft is betting heavily on ambient, always-on workplace AI — relevant for anyone planning IT or workflow strategy.

→

Mustafa Suleyman predicts that within 2-3 years, AI will deliver Mayo Clinic-level healthcare advice to everyone via a pocket assistant — a strong signal that health-focused AI products are an imminent priority for Microsoft.

Notable quotes

“This is about taking first steps towards true self-sufficiency in AI, and that means that we have to have the capacity to build our own models from scratch and show that we can achieve the absolute frontier.”

“I really think that we are close to medical super intelligence. I think 2 to 3 years and it will be possible to get access to the absolute best healthcare in the world which currently you can only get if you go to Mayo.”

“I'll drink from the fire hose all week. I will read all the news. I will try to keep up with it and then turn around and just bring you what I think you need to know.”

Worth watching?

✅

Worth watching the full video?

Watch if you want the Mustafa Suleyman interview clips and the live demos of MAI Image 2.5 and Runway ALF 2.0 — the key announcements and context are all captured here if you just need the facts.

Topics

AI & Tech Microsoft

Explore more summaries on these topics →

Saved you some time? The creator still deserves a like.

Watch on YouTube →

More like this