1. OpenAI Closes $40B Round — The Biggest Private Raise in History

OpenAI has closed a $40 billion funding round led by SoftBank, pushing its valuation to $300 billion. This is the largest private tech fundraise ever recorded. The capital will be used to scale compute infrastructure, build out data centers, and advance the next generation of AI models.

So what? When OpenAI raises $40B in one round, it signals that the AI arms race has moved from a sprint into a full-scale industrial buildout. Whoever controls the most compute wins the next era of AI.→ Read more

2. Spud — OpenAI's Next Base Model Is Weeks Away: OpenAI President Greg Brockman confirmed Spud — the company's next flagship model — is a "big model" representing nearly two years of research and a new base for all future ChatGPT models.

• Not a minor update — an entirely new foundation model

• Designed to better infer user intent without needing perfect prompts

• Altman told staff it will "accelerate the economy" — expected mid-to-late April

• Internal team now calling their work "AGI Deployment"

Spud isn't just a model launch — it's OpenAI's bet that AI is now ready to run the economy.

3. Google Open-Sources Gemma 4 — A Powerful Model Anyone Can Run

Google DeepMind released Gemma 4, its latest open-weight model series. The new models range from 1B to 27B parameters and outperform many proprietary competitors on key benchmarks. Gemma 4 is freely available for developers to download, fine-tune, and deploy.

So what? Google giving away a best-in-class model for free is a strategic move to lock developers into its AI ecosystem. Open-source is no longer the underdog approach — it's a competitive weapon.→ Read more

4. AI Captures 81% of All Q1 2026 VC Funding — $297B in One Quarter

Venture capital data for Q1 2026 shows AI startups attracted $297 billion — representing 81% of all global VC investment in the quarter. The concentration of capital into AI is accelerating, with mega-rounds becoming the norm rather than the exception.

So what? When 81 cents of every VC dollar goes to AI, it's not a trend — it's a structural shift. Non-AI startups are being squeezed out of funding conversations entirely, while AI infrastructure and foundation models dominate investor portfolios.→ Read more

Gemma 4 on a Phone: Developers are already running Gemma 4's E4B model on Android devices. The speed benchmarks are stunning — sub-100ms responses on mid-range hardware. This is what local AI looks like when it's truly competitive.

Why it matters: When frontier-class reasoning runs on the device in your pocket, the cloud dependency for AI starts to break down. Privacy, latency, and cost all change.

Builder Moves

If you’re building with AI right now, this week’s moves matter:

Our take on this week: Everyone is celebrating the $122B round. But we are watching something else.

When 4 companies absorb 64% of all global VC in one quarter, that’s not a healthy ecosystem — that’s a consolidation. The same dynamic that gave us 3 cloud giants and 2 social media monopolies is now playing out in AI, but 10x faster.

The real opportunity isn’t in being the next OpenAI. It’s in building on top of these models — and tools like Gemma 4 (free, open, runs locally) make that more accessible than ever. The frontier is expensive. The applications layer is where the next generation of businesses will be built.

Build in the layer, not at the frontier.

Imagine you’re sitting in a WeRide robotaxi in Dubai. No driver. Fully autonomous. You’re gliding through Jumeirah, the AI navigating flawlessly.

The question we’d want to ask isn’t ‘is this safe’ — the data says it is. The question is: what happens to the 4 million taxi and rideshare drivers globally when this scales from Dubai to Dallas, from Shanghai to Seattle?