Hybrid agentic AI powerhouse, or human-AI slop soup?
September 22, 2026 · OSASI Commentary
Agentic AI is sold as doing the job faster. A consultant who builds hybrid organisations makes the better case: quality must remain the key focus, and not the inverse. Humans and AI working together in ways that neither could manage alone.
How a brief is made
A member reads an article or listens to a podcast and sends the snippets, key thoughts or concerns tied to a passage or a timestamp. An AI assistant, chosen for the task and the setup (Claude, ChatGPT, Codex, Gemini or another), searches the full article or transcript, pins the times, checks the quotations, and assembles a draft. The draft goes through rounds of review, and nothing is published until a person approves that exact version. The cues on each post start the player at the passage discussed. Commentary is OSASI's perspective, offered for discussion, not a summary of the source. Each brief is co-authored by a person and an AI assistant. AI assistant for this brief: Claude (Anthropic).
Source podcast · Value Driven Data Science
This episode prompted the commentary below. It is one perspective, offered for discussion.
Value Driven Data Science · September 3, 2026
Episode 121: What Hybrid Agentic AI Organisations Mean for Data Science
Listening highlights · Dr Genevieve Hayes with Victor Coimbra, partner and CTO at Artefact. About thirty minutes.
Tap a time above to start the player there. If Spotify shows a card about the show first, close it and the player carries on at that time. Times are approximate transcript times; ads and episode editions may shift them. This plays the full episode, not an extracted clip.
Open episode in Spotify · Episode page · Timestamped transcript
Who is talking
Dr Genevieve Hayes hosts Value Driven Data Science, a show for data professionals who want to become strategic experts rather than order-takers. Her guest is Victor Coimbra, a partner and chief technology officer at Artefact, which she introduces as the world's largest pure-play AI consulting firm; he co-founded its Latin American operations and was named to the Forbes 30 Under 30 Brazil list in 2024.
Hayes opens with the cycle everyone in a data profession has lived through. First the chatbots and the promise of humans and AI working together. Then the question of whether human workers were needed at all, and the first wave of redundancies. Then the discovery that agents are not cheap and that experienced people see things AI cannot, the quiet rehiring, and a return to the starting point with far better technology: "a hybrid agentic organization, where humans provide the judgment, creativity, and context while AI agents provide the speed, memory, and execution power". Coimbra has helped global enterprises make that shift, which is why his account is worth the half hour.
The hybrid organisation
Coimbra builds what he calls hybrid agentic organisations for large companies: workplaces where people and AI agents share the work, the way factory floors once learned to share space with machines. He set out how to do it well, and, more usefully for anyone being sold agentic AI, how it goes wrong.
His first move is to reject the sales pitch. The gain, he says, is not speed.
"Right now the real productivity gains is not on I'm able to do my job 30 percent faster. That's not where the value actually lives. The value lives when I have a shared cognitive capacity that I can expand, because now I have more workers, and I can raise the quality of my work."
The hybrid organisation exists "to produce better quality in terms of products and service for societies as a whole". Efficiency is the by-product. Quality is the point.
A science team again
The second theme is what the arrangement does to the people. Coimbra's data science teams, relieved of code they no longer need to write, are becoming scientists again: forming hypotheses, running experiments, reading papers. At Artefact "we had these different types of models that we like to compete. Like LLM is one type of model, but we have fine-tuned models, we have visual models." One executive's realisation stuck with him: "my data science team is a science team", not an analytics team that produces whatever model the executive wanted to see.
"I think the principles of the scientific process is what actually is driving people to this career right now." For appraisers the parallel is exact. The analysis is the work; the report writing and the data wrangling are the parts a machine can carry.
The paradox that leaders create
Hayes brings the reality check. AI creates time to think, and at the same time every knowledge worker is now expected to deliver faster. Having the time is not the same as being allowed to use it.
Coimbra's answer is that the paradox is self-inflicted. It appears "when you assign the wrong task to the wrong type of archetype". He cites the MIT finding that most AI use cases failed, "not because the technology's bad, but because we use it in the wrong way", and gives the everyday example: a presentation is a human thing, a storyline someone must think through; hand the whole of it to a model and the result is what leaders have started calling AI slop.
So his rule for where pressure belongs:
"You need to put pressure in terms of the agent needs to produce stuff faster, but you cannot put time and pressure in something that's creative by nature, because this gonna become AI slop."
Split the tasks properly and, in his experience, quality rises on its own, because the person is using the time to think and the machine is supplying the speed.
The locomotive
Positioning yourself
Asked what a data scientist should do now to stay indispensable, Coimbra names three things and none of them is a tool. Master scientific thinking, "because that's where you're gonna shine". Truly study the problem you are asked to solve, whether it is marketing, finance or supply chain. And be able to "communicate, debate, and defend your ideas in a structured way", the way an academic defends a hypothesis. Coding, he adds, is nice, but these are the skills that matter.
What this means for OSASI
Every line of that advice reads as if written for a valuation professional. The value of an appraiser was never typing speed. It is the analysis, the study of the problem, and the ability to defend a conclusion in front of a client, a reviewer or a court. Those are exactly the parts Coimbra says to keep for the human and protect from time pressure.
The pieces to hand to a machine are also clear: assembling the data, checking it, drafting the parts of a report that are description rather than judgment, and running more analysis than one person had time for. That is where speed is welcome, and it is what OSASI's tools are built to provide, so that the time saved goes back into the work only a professional can do.
The warning applies with equal force. A profession that lets the sales locomotive set the pace, and uses AI mainly to say yes faster, has learned nothing from 2008. Better, not faster, is the standard, and it is the one that survives scrutiny.