Tech strategist exploring AI sys, executive intelligence, and the value of professional experience beyond a CV. Building MimirOne — The Professional Continuum.
A CV tells you where someone has been. Rarely how they think, solve problems, or exercise judgment.
As AI reshapes work, professional experience deserves more than a list of achievements.
That’s why we’re building MimirOne.
The Professional Continuum.
https://t.co/gMrxd88mxQ
A CV tells people what you did. A professional record should also help explain the context, reasoning, decisions and lessons behind it.
That's the thinking behind the Professional Continuum. https://t.co/gMrxd88mxQ
AI can take over tasks. But some of those tasks were also how people learned to recognise exceptions, challenge assumptions and develop judgment.
If we automate the work, where will the next generation develop the expertise?
A CV records roles and results. It rarely preserves the reasoning behind a decision, the context of a trade-off, or the lesson that changed how you work. MimirOne helps build a Professional Continuum that grows with you. Start with one Opportunity: https://t.co/gMrxd88Uno
AI can improve output while displacing the practice through which people build skill. Diener et al.’s 2026 model makes the trade-off explicit: how do we preserve the expertise needed to oversee AI?
https://t.co/SW3xNnei8c
That's a fair distinction. An ontological graph can represent the relationships, circumstances, constraints and dependencies surrounding a decision far better than a collection of disconnected documents.
Where I see a complementary challenge is preserving how an experienced professional interpreted those circumstances: which signals mattered, what alternatives were considered, what assumptions were challenged and why one course of action was chosen.
I agree that AI is more likely to accelerate skilled professionals than replace them in the near term. The question is how we preserve and make their accumulated judgment more accessible without reducing it to a set of generic rules.
AI can improve today’s output and still weaken tomorrow’s expertise if people stop practising. A new November preprint models this tension. The workforce question: which skills must people keep exercising, even when AI can do the task? https://t.co/SW3xNnei8c
I don’t think it needs to become time tracking. I’d start with a small sample of AI-assisted tasks and capture three outcomes: accepted, materially corrected, or redone. Add one reason for the correction—accuracy, missing context, tone or relevance—and use broad effort bands rather than minutes.
That should expose the rework pattern without creating another reporting burden. The challenge is making the measurement almost invisible.
@thoughtworks Resilience is also a knowledge problem. When essential systems fail, can the organisation find the people, decisions, escalation paths and workarounds needed to operate? Critical operational know-how should not exist only in a few people's memory.
@johnroodepic That distinction matters. A summary can preserve the topic while losing the qualifier that changes the decision. For any durable professional record, I'd want three things kept distinct: what was said, what was inferred, and what was actually decided—with provenance.
@aviral_kumar2@XuIvan78278@QuYuxiao@ianwu97 Interesting actor/writer framing. One question beyond next-task reward: how does the system distinguish a reusable lesson from a local exception? For professional knowledge and KM, provenance and the conditions under which a decision worked can be as important as the outcome.
Important question. A meeting transcript can contain confidential contributions from people outside the host organisation. Participants should know before the call what is captured, where it goes, who can access it, and whether it will be reused. Consent should not be assumed simply because someone joined the meeting.
AI literacy is more than tool fluency. It is framing the task, testing the output and retaining the learning. A 297-worker experiment found goal-setting and task decomposition improved AI-assisted writing—even after scaffolds were removed. https://t.co/YHw15Zfb1z
A CV records where you worked and what you did.
It rarely captures why you chose one path, what failed, or how your judgment changed.
That accumulated experience deserves a better record.
MimirOne: https://t.co/gMrxd88mxQ
@AndrewYNg AI engineering will matter, but the ability to frame the problem, select relevant context and evaluate the result is just as important. Technical fluency builds the agent; professional judgment determines whether its work is useful.
@Snowflake@kargo The speed from request to production is impressive. The organizational question is what the team learns during that acceleration: which assumptions changed, which decisions mattered, and what should be retained for the next workflow.
@IBM I'd nominate contextual judgment: knowing which evidence matters, what is missing, when to challenge an answer, and how to take responsibility for the outcome. AI literacy matters, but judgment gives it direction.