The golden age of the all-in-one digital workspace is over. It wasn't killed by a better-funded rival or a slicker user interface. It was rendered obsolete by a fundamental paradigm shift. The very premise of platforms like Notion, Coda, and Airtable—that humans should spend their time meticulously organising human-generated knowledge—has been invalidated by AI that can generate, synthesise, and act on information autonomously.
These platforms are beautiful, intricate digital museums. And in the age of intelligent action, no one has time for a museum visit.
The Cambrian Explosion of the Digital Filing Cabinet
Let's be clear: the rise of the 'Notion category' between 2018 and 2022 was logical. It was the apex of human-centric computing. As remote work scattered teams across postcodes and post-Brexit Britain grappled with new operational complexities, the need for a single source of truth became paramount.
UK adoption was ferocious. Industry estimates suggest that enterprise and SME spending on collaborative productivity tools grew by over 40% year-on-year in the UK during this period. For a subscription fee, often running into tens of thousands of pounds for a mid-sized London firm, you could have a 'second brain' for your organisation. Company wikis, project roadmaps, CRM databases, employee onboarding checklists—all meticulously cross-linked in a digital utopia. The promise was intoxicating: conquer chaos through organisation.
This created a cottage industry of 'Notion consultants' and 'operations specialists' whose primary function was digital librarianship. Their key performance indicator was the elegance of the database structure, not the velocity of the business a-pivots.
From Curation to Generation: The AI Wrecking Ball
The fundamental flaw in this model is that it presumes the information within it is static and that the primary challenge is access. Both presumptions are now false. Generative AI, and more importantly the agentic workflows it enables, doesn't need a neatly curated wiki. It thrives on raw, unstructured data and creates the 'document' you need, at the moment you need it.
Consider a simple use case: a sales team's competitor analysis page in Notion. Previously, a junior analyst would spend a day every month manually updating company profiles, funding news, and product feature launches. It was a chore, prone to error and immediately out of date.
Today, an AI agent can be tasked: "Every morning at 8 am, scan the latest filings on Companies House, tech news from specified sources, and global patent databases for our top five competitors. Produce a one-page summary of material changes, highlight new risks and opportunities, and post it to our leadership Slack channel."
The Notion page isn't just automated; it's entirely superfluous. The 'workspace' is no longer a destination you visit but a service that reports to you. The human role in this process has shifted from a low-value curator to a high-value strategist who defines the AI's objective.
The £100 Billion Procrastination Engine
Here lies the controversial truth: the 'second brain' movement was, for many organisations, a sophisticated form of procrastination. The act of organising information became a substitute for acting upon it. We fetishised the structure of our knowledge bases, colour-coding tags and perfecting database relations, while the core business task languished.
Nowhere is this clearer than in the UK's high-stakes professional services and finance sectors. A report for the City of London Corporation noted that in 2023, technology and operations budgets continued to climb, with a significant slice—estimated at over £500 million across the square mile—dedicated to SaaS subscriptions for workflow and data management.
Think of a private equity firm in Mayfair tracking a pipeline of potential acquisitions. The old way involved an elaborate Airtable or Notion database. Associates would spend countless hours inputting data, linking to virtual data rooms, and manually updating deal stages. It felt productive. It looked organised. But it was profoundly inefficient.
The real work isn't maintaining the database. The real work is identifying the best targets, assessing their financials, and modelling their future value. Today, a custom AI agent can ingest real-time market data from PitchBook and Bloomberg, analyse the target's financial statements, perform sentiment analysis on its press coverage, and even draft a preliminary investment memo. The beautifully crafted Notion tracker is bypassed entirely.
In this light, the millions spent on these platforms start to look less like an investment in productivity and more like a subsidy for structured idleness.
Re-evaluating the SaaS Stack: A CFO's Reckoning
The debate on AI vs human roles is often framed as a simple substitution. Will AI replace the paralegal? The financial analyst? The copywriter? This is the wrong question. The more profound change is the collapse of entire software categories, and with them, the human roles built around servicing that software.
A recent, confidential survey of FTSE 250 CFOs provides a stark indicator. When asked about their 2024-2025 technology budgets, 65% indicated a planned review of their 'broad-based productivity SaaS' spend, with the explicit goal of reallocating funds to 'specialised generative AI and automation initiatives.'
The value proposition is simply too different to ignore.
- Notion/Airtable Value: "We provide a flexible canvas for your team to organise its work. Your ROI is improved information access and collaboration."
- AI Agent Value: "Define a business outcome, and I will execute the tasks required to achieve it. Your ROI is increased output, reduced headcount, and accelerated business velocity."
Faced with this choice, the decision for any rational executive is obvious. At Creative Agent, we've seen this shift firsthand. A year ago, prospective clients from the legal and finance sectors came to us asking for better ways to build internal knowledge hubs. Today, they come asking for 'agentic workflows' to automate due diligence, compliance monitoring, and client reporting. They don't want a better filing cabinet; they want a relentless, autonomous workforce.
The New Human Role: From Digital Librarian to AI Conductor
This isn't a story of doom for the human worker, but it is a story of radical transformation. The skills that were valuable for servicing the Notion-era workflow—meticulous organisation, data entry, process documentation—are being devalued in real time.
The new high-value skills are those of the AI Conductor:
Problem Formulation: The ability to translate a complex business problem into a clear, unambiguous objective for a system of AI agents. Not "organise our sales leads," but "build a system that identifies UK-based SaaS companies with 10-50 employees, enriches their contact data, drafts a personalised outreach email based on their recent hiring activity, and schedules a meeting if they respond positively."
System Design: Architecting how different specialised agents (a research agent, a writing agent, a scheduling agent) will interact, pass information, and handle exceptions. This is an engineering discipline, but one based on logic and language, not just code.
Quality Assurance & Iteration: Reviewing the output of AI systems, identifying subtle errors or areas for improvement, and refining the prompts and instructions to enhance performance. The human becomes the editor, the ethicist, and the final arbiter of quality.
This is a fundamental shift in the nature of knowledge work. The value is no longer in having knowledge or even organising it. The value is in applying it strategically through the leverage of intelligent machines.
What's Next? The Great Unbundling is Here
Predicting the future of software is a fool's errand, but the trajectory is clear. Monolithic, all-in-one platforms will be unbundled. Instead of going 'into Notion' to do five different things, you will summon five different specialised agents to do them for you, likely through a simple conversational interface.
The user interface is becoming language itself. The platform is the underlying AI model.
Notion and its peers will not disappear overnight. They will attempt to pivot, integrating their own AI features—as they are already doing. But they face a classic innovator's dilemma. Their entire architecture and business model are predicated on being the central 'place' where work happens. The new paradigm is that work happens everywhere, executed by autonomous agents, with humans simply directing the orchestra.
These platforms have become sprawling digital estates, impressive in their complexity but crippling in their maintenance costs—both in subscription fees and, more importantly, in human attention. The future doesn't belong to the best digital organiser. It belongs to the most effective autonomous operator. The great challenge for every UK business, from a Shoreditch startup to a Canary Wharf bank, is to stop curating their digital museum and start building their automaton factory.
