In partnership with
TL;DR
Cursor 3.1 turns vibe coding into a real engineering environment
NVIDIA open-sources AI models that make quantum computers faster and more accurate
Adobe's new AI assistant edits across all Creative Cloud apps from a single chat interface
Meta is set to out-earn Google in ads for the first time ever
Build, sell, & follow up automatically with this repeatable product workflow
Open Agents lets you build your own AI coding factory, like Stripe and Spotify
Become An AI Expert In Just 5 Minutes
If you’re a decision maker at your company, you need to be on the bleeding edge of, well, everything. But before you go signing up for seminars, conferences, lunch ‘n learns, and all that jazz, just know there’s a far better (and simpler) way: Subscribing to The Deep View.
This daily newsletter condenses everything you need to know about the latest and greatest AI developments into a 5-minute read. Squeeze it into your morning coffee break and before you know it, you’ll be an expert too.
Subscribe right here. It’s totally free, wildly informative, and trusted by 600,000+ readers at Google, Meta, Microsoft, and beyond.
CURSOR SHIPS MULTI-AGENT WORKFLOWS IN CURSOR 3.1
Cursor now supports a tiled layout, letting you run multiple agents in parallel and compare outputs in real time.
Voice input is more reliable, with full clip recording and a new Ctrl+M hold-to-talk shortcut.
You can select which branch a cloud agent runs on before launching, cutting down on mistakes.
Diff-to-file navigation now drops you directly into the exact line with full editor control.
Workspace search includes include and exclude filters for tighter codebase scoping.
Performance upgrades include an 87% reduction in dropped frames when streaming large edits.
In other words, Cursor is optimizing for what happens after the prototype works.
Cursor built its reputation on speed. Prompt something, get code, ship fast. But these updates are not about speed. They are about structure.
Running multiple agents at once. Choosing branches before execution. Navigating diffs like a traditional IDE. These only matter when projects get complex, teams get involved, and mistakes get expensive.
The early wave of AI tools made it easy to start. The next wave is about staying in control as things scale. Multi-agent setups, tighter feedback loops, and better visibility into changes all point in the same direction. Builders are no longer just generating code. They are coordinating systems.
Cursor is evolving from a prompt box into an environment.
OUR TAKE
Now that most of us have creation down, we need to focus on coordination.
Anyone can spin up an agent. Fewer people can run several at once, guide them across branches, validate outputs, and turn that into something stable.
Cursor is betting that the future is not one agent doing everything. It is multiple agents working in parallel, with a human acting as the orchestrator.
That is a very different skill set than vibe coding.
QUICK HITS
NVIDIA just open-sourced AI models built specifically to fix quantum computing's biggest bottlenecks. Called Ising, the models handle quantum error correction and processor calibration — cutting calibration time from days to hours, and running error correction 2.5x faster and 3x more accurately than the current standard.
Adobe just launched Firefly AI Assistant. Describe what you want, and it handles the edits across Photoshop, Premiere, Lightroom, and more. The conversational interface runs complex multi-step workflows on your behalf, learns your preferences over time, and will eventually work inside third-party tools like Claude too. No launch date yet, but it's coming to Firefly soon.
Meta is on track to beat Google in global ad revenue for the first time ever, projected to pull in $243B vs. Google's $240B by end of 2026. Meta's AI-powered Advantage+ ad suite is the main driver, growing at 24% this year vs. Google's 12%. The search giant still dominates in other areas, but in pure ad dollars, Meta is about to take the crown.
One brand built 30+ landing pages through Viktor without a single developer.
Each page mapped to a specific ad group. All deployed within hours. Viktor wrote the code and shipped every one from a Slack message.
That same team has Viktor monitoring ad accounts across the portfolio and posting performance briefs before the day starts. One colleague. Always on. Across every account.
5,700+ teams. 3,000+ integrations.
AI PRODUCT WORKFLOW THAT MAKES MONEY
Most people build the product. Almost nobody builds the system around it.
Why build this? Everyone's using AI to create digital products fast. Prompt packs, template bundles, mini-guides. The problem is they stop at the landing page and wonder why nothing sells. The real issue isn't the product. It's that there's no follow-up. Most people don't buy the first time. They click, get distracted, and disappear. This workflow fixes that.
Steps:
Create a simple digital product with AI. Pick one specific problem for one specific person. A prompt pack, a template bundle, a content system. Use ChatGPT or Claude to draft it fast, then go in and make it feel human. Don't spend weeks on this.
Build a one-page landing page. Use Onepage or any simple builder. You need a headline, a short product description, what's included, and a buy button. Keep the price low friction: $9, $19, $29.
Set up a welcome flow. In Omnisend, build a basic email sequence for anyone who opts in but doesn't buy right away. Control the timing, customize the emails, and let the system do the follow-up for you.
Set up an abandoned checkout flow. Someone starts to check out and bails. Without this, they're gone. With it, they get a follow-up automatically (this is the step most people skip entirely).
Set up a post-purchase flow. After someone buys, follow up. Upsell, ask for feedback, bring them back. One purchase can turn into more if you have the right sequence in place.
Drive traffic into the system. Short-form content, a newsletter, a free lead magnet. Whatever fits your audience. The workflow is modular, so you can swap the product or niche and the structure still holds.
What you'll end up with: You walk away with a full system, not just a product. Someone sees your content, clicks through, and enters an automated flow that follows up, recovers abandoned checkouts, and converts over time.
Full tutorial:
TOOL OF THE DAY
The open-source blueprint for building your own AI coding factory.
Off-the-shelf coding agents choke on big monorepos. They don't know your codebase, your workflows, or your stack. Open Agents fixes that. Every agent gets a full cloud sandbox, automatic git commits, and durable workflows that survive restarts and pick up exactly where they left off.
Stripe, Ramp, and Spotify are already building versions of this internally. Now it's open source. The moat isn't the code you ship anymore. It's the system that ships it.
