WritingAI21 LabsAI21 Labspublished Jun 25, 2026seen 4w

Ai All The Work

Open original ↗

Captured source

source ↗
published Jun 25, 2026seen 4wcaptured 4whttp 200method plain

“AI all the work”: Inside Ori Goshen’s RAISE Summit keynote | AI21

Skip to Main Menu

Skip to Main Content

Skip to Footer

Back to Blog

-->

Back to Blog

TL;DR:

In his keynote at the RAISE Summit 2025, AI21 Labs co-founder and co-CEO Ori Goshen explains what it means to “AI all the work.” He breaks down why real work depends on multi-step tasks—and why LLMs alone aren’t consistent enough for enterprise use. That’s where AI21’s Maestro comes in: a system built to plan, execute, and validate complex workflows with transparency and trust.

See his full talk below.

In his talk from the RAISE Summit in Paris, Ori lays out a clear path to “AI all the work.” Work, he says, is just six billion tasks waiting to be understood, broken into steps, and executed correctly. But while large language models often get single steps right—about 90% of the time—real-world tasks require multiple steps. And when you string steps together, those small errors compound. That’s why prompting alone isn’t enough for enterprises. What’s needed is something as flexible as prompting, but as robust as a hand-crafted workflow.

That’s the idea behind Maestro . It helps developers build enterprise-grade knowledge agents that solve tasks by generating multiple plans, selecting the optimal one, and validating every step. It’s model-agnostic, allows control over budget and requirements, and returns outputs with a report card and execution graph that shows what the agent did and why. We built Maestro to move beyond single-turn reasoning and give enterprises a system they can trust to…

Discover more

Jun 25, 2026

Token spend isn’t going down. You need more than naive routing to manage it

Labs in Front

-->

Jun 24, 2026

Tipping the scales: Merging weak agents into a state-of-the-art deep researcher

Labs in Front

-->

Jun 4, 2026

First scale, then enrich: How the right execution strategy helped us reach state-of-the-art on SWE-rebench

Notability

notability 3.0/10

Routine blog post from AI21 Labs.