SEPT ’26: HUMANS STILL VERY MUCH HAVE JOBS. AI isn't even close to meeting the standard required of the most junior professionals in banking, advisory, and private equity today. ATLAS Finance agents must navigate complexity like a human does: "Johnny you just got staffed - check your email." Human: 100%. Best AI: 12%. Agents have access to data rooms, email, chat, calendars, docs, notes, Excel, etc. Our real finance professionals pushed the realism frontier further by building dozens of coworker and client personas that introduce ambiguity, (controlled) contradictions, and real-time updates that must be adjudicated to successfully complete the client-ready deliverable. We gave 11 frontier models 100 expert-level tasks that each take humans 15–30 hours. On Wall Street there's a saying, "If it's 95% right, it's 100% wrong." Autonomous knowledge work outside software still has a very long way to go.

AI is bringing massive change, and with that comes uncertainty, particularly about jobs. Our work at Handshake is to help more people take part in this new economy, to use AI to help push forward human knowledge, productivity and creativity. We’re already seeing people use AI to change the trajectories of their careers and lives in dramatically new ways: to earn real $$ from what they know, more easily learn skills that open new doors and spend more time doing work they care about, whether as an employee or entrepreneur. We have lots to do to make that work for billions of people, but I’m optimistic. And I’m incredibly grateful to everyone on the Handshake team who comes to work every day believing we can do that. Thank you to @TIME for the recognition. #TIME100AI#

Handshake AI CSO @Sbhaiwala03 reveals how they build simulated white-collar work environments to train frontier AI agents: "An environment consists of a few elements. The first is the software tools the agent needs access to. If I'm an investment banker, I have access to Excel, SEC EDGAR where all the financial reports are, and PowerPoint. Agents need the same access." "We build simulated versions because of infrastructure and access constraints. Labs are querying that tool thousands of times in a given second." "The second thing you need is data that populates the environment. If I'm building a financial model, I need the data room with the messy files from the company I'm evaluating." "The final piece is tasks and verifiers. What are the requests of the agent that would reflect what a real investment banker might ask. We try to find tasks complex and realistic enough that the models fail. If the model can't perform, we know it needs more training data." "What we send to labs is essentially a Docker file, a full containerized task. It has the tools the model has access to, the full data that populates the world, and the task and verifier. We'll send tens of thousands of these to a given lab for just one domain." @joinHandshake @GarrettLord

Handshake CEO @GarrettLord explains why AI agents exploded in software engineering but not the rest of white-collar work: "In software engineering, the average engineer is two, three, four times more productive. Software engineering agents are able to operate for multiple days at a time, automating hundreds of hours of work." "CEOs are not seeing that level of automation and ROI in other functions beyond software engineering. Software engineering is accelerated in such a massive way because everything was written down in GitHub. And it's extremely clear with what the right answer is." "When you have enough verifiable rewards and you scale up enough compute against verifiable rewards, you get what happened in software engineering, which is agents able to operate and do the job of humans." @Sbhaiwala03 @joinHandshake

Handshake CEO @GarrettLord reveals the white-collar automation gap inside the world's biggest banks: "We're not seeing that same level of acceleration in new business onboarding for a top three bank in the world. They have 65,000 people in their back office functions. It takes them 32 days to onboard a new business into their systems." "These are tasks and practices inside the companies that they believe could be automated and putting the human in a much more accelerated position. But they're not able to do that by sharing skill files back and forth in Claude Cowork, for example." @Sbhaiwala03 @joinHandshake

My favorite way to explain a compounding AI advantage, from my Cloud Wars chat: A Waymo hits a scenario it can't handle. A human takes over. That drive gets captured and pumped into training. The next day the car does it on its own. The future is running that loop across knowledge work!

The benchmarking landscape is evolving: 1. Domain-specific evals built by the companies that know the workflows best 2. Agent environments extended beyond a container into sandboxed infrastructure (data-eng-bench has two versions: local DuckDB and remote Snowflake) 3. Public evals released by companies to prove the effectiveness of their agent product on the workflows their customers care about 4. Benchmarks maintained as living software, with the community submitting trajectories to the leaderboard and proposing new tasks

Frontier data is a research problem. For example, some of the biggest academic CUA papers are solely focused on data synthesis for SFT (OpenCUA, from last August) and RL (CUA-Gym, from this June). Good data requires diligent researchers and engineers. By next year, I'd expect data talent to look increasingly similar to the current research scientist talent. Already hear small stories of data talent poaching wars and such.