In partnership with
TL;DR
Anthropic launches Claude Sonnet 4.6 with major upgrades to coding, computer use, and long-context reasoning
OpenAI hires OpenClaw founder to accelerate personal agent development
$2T wiped from SaaS as investors price in “seat compression”
Spotify says its top developers haven’t written a single line of code in 2026
I built and monetized an AI agent in one conversation
AI in HR? It’s happening now.
Deel's free 2026 trends report cuts through all the hype and lays out what HR teams can really expect in 2026. You’ll learn about the shifts happening now, the skill gaps you can't ignore, and resilience strategies that aren't just buzzwords. Plus you’ll get a practical toolkit that helps you implement it all without another costly and time-consuming transformation project.
ANTHROPIC LAUNCHES CLAUDE SONNET 4.6
Same pricing.
Much stronger performance.
What changed:
Major coding improvements
Better instruction following
Fewer hallucinations
Stronger multi-step execution
1M token context window (beta)
Big gains in computer use
Early testers preferred it over Sonnet 4.5 70% of the time.
They even preferred it over Opus 4.5 59% of the time.
Claude can now:
Navigate spreadsheets
Fill multi-step web forms
Operate across browser tabs
Use software without custom APIs
And when you combine that with 1M tokens of context, you get something closer to long-horizon planning than autocomplete.
The stack is compressing.
OUR TAKE
Once AI can operate software reliably, the question shifts from:
“How do we assist workers?”
to
“How many workers do we still need?”
Which brings us to this week’s dominoes.
QUICK HITS
OpenAI hires OpenClaw founder: Peter Steinberger joins OpenAI to lead personal agent development, signaling a shift from conversational AI to fully actionable desktop agents as orchestration becomes the new competitive moat.
$2T SaaSpocalypse hits software stocks: Autonomous agents from Anthropic and OpenAI trigger a historic sell-off in seat-based SaaS, as investors price in “seat compression” and the collapse of per-user growth models.
Spotify engineers stop writing code: Spotify’s co-CEO says top developers haven’t written a single line of code in 2026, using internal AI systems powered by Claude to ship features directly from Slack during their commute.
Close more deals, fast.
When your deal pipeline actually works, nothing slips through the cracks. HubSpot Smart CRM uses AI to track every stage automatically, so you can focus on what matters. Start free today.
BUILD AND MONETIZE AN AI AGENT IN ONE CONVERSATION
Appy AI is a conversational builder for turning your decision-making process into a live, monetizable AI agent.
⏱️ 5–10 minutes | 🔧 Needs: a repetitive workflow you run often, clear qualification criteria, Stripe account (optional)
Why build this? Most knowledge work is repetitive. You explain the same process, qualify the same leads, answer the same questions. Instead of optimizing the call, you can productize the decision framework behind it. The shift is from selling time to selling access to your process.
Steps:
Start with a real workflow (not an idea):
Pick something you already do repeatedly. Lead qualification. Client onboarding. Candidate screening. Vendor approval. Internal request triage. The best agents replace conversations you’re tired of having.Explain how you think, not how to build it:
Describe your decision process in plain language.
What questions do you ask? What makes something a fit? What disqualifies it? What happens next? Just narrate your logic.Answer clarifying questions from the system:
The builder will ask about thresholds, tone, next steps, and edge cases. This is where you refine your voice and criteria. Treat it like training a junior operator.Test it like a real user would:
Pretend you’re a client, customer, applicant, or internal teammate. See how the agent responds. Adjust clarity or constraints until it behaves like you would.Enable memory and continuity:
A good agent shouldn’t be stateless. It should remember context, prior outcomes, and returning users. That’s what turns it from a chatbot into an evolving system.
Monetize access (optional):
Instead of charging for time, you can:Gate deeper access behind payment
Charge for priority processing
Package your framework as a paid entry point
Offer premium routing or escalation
Deploy it as your first layer:
Make the agent the front door:
Link in bio
Website entry point
Auto-response to inbound emails
Internal request system
Let it filter, qualify, and structure work before it reaches you.
Expected outcome:
A live AI agent that handles repetitive workflows, applies your standards automatically, and optionally monetizes access to your process, all built from one conversation.
Full tutorial:
TOOL OF THE DAY
A marketplace where AI agents (or you) can hire real humans to complete physical-world tasks.
A marketplace where you—or your AI agent—can hire verified humans to complete real-world tasks like scouting venues, picking up packages, attending meetings, taking photos, or running errands.
You browse by skill and location, set a budget, and pay via escrow.
As agents automate digital workflows, platforms like this connect AI coordination to physical execution.
Automation handles software.
