Can Your Coworking Space Dream at Night?

Recently, we came across Google's Long Horizon harness, an AI implementation described as being able to "dream."

During a dedicated process, the AI reviews earlier conversations, connects information that appeared at different times, and reorganises what it has learned into a more useful memory. A detail mentioned on Monday can acquire new meaning when another piece of information appears on Wednesday.

The term "dreaming" is playful. The management capability behind it is significant.

It raises a useful question for coworking operators: what would happen if a coworking business could remember its own experience well enough to reason from it?

Some theories of dreaming suggest that it helps the brain process experience and consolidate memory. Information gathered during the day is reviewed, reorganised and integrated with what is already known.

A similar process could give a coworking business a more coherent understanding of itself.

A coworking business produces a detailed record of itself every day.

Booking platforms record demand. Access control shows how buildings are used. Accounting software holds payment history. CRM records capture parts of the sales process. Contracts contain commercial commitments. Emails, support tickets, meeting notes and team conversations contain explanations that never reach the formal records.

Each tool remembers a fragment.

The full history of the business remains distributed across databases, documents, conversations and people. Founders and experienced team members often provide the missing connections themselves.

They remember why a discount was approved, why a room was repositioned, which client raised an issue before renewal, why an event format was discontinued, or which assumptions supported an expansion decision.

As the business grows, this informal memory becomes harder to maintain. Teams change. Locations multiply. Decisions accumulate. The reasoning behind earlier choices becomes separated from the results that followed.

This fragmentation came up repeatedly during Coworking Tech Week 2026. Across conversations about AI, data and coworking technology, the same requirement kept returning: useful AI needs reliable context, clearly organised information and a defined role inside the business.

Most coworking companies already have enough raw information to improve decision-making. The challenge is preserving the relationships between that information.

A management platform can show that meeting room revenue declined. It may hold no record of the pricing discussion six months earlier, the equipment complaints that followed, the competitor that opened nearby, or the sales conversations in which prospects questioned the product.

A CRM can show that an opportunity was lost. It may contain little about the prospect's internal timeline, the objections raised during the tour, the promises made by the team, or the way similar companies eventually bought elsewhere.

The missing layer is a durable business memory: a structured history of facts, conversations, assumptions, decisions and outcomes that can be examined together.

AI already has a practical role inside coworking businesses.

It can answer common member questions, organise meeting notes, prepare invoice reminders, qualify enquiries, draft communications and route support requests.

These applications matter because administrative work absorbs a significant amount of team capacity.

During Spacebring’s Coworking Tech Week session, Helga Moreno presented findings from their research involving 200 coworking owners, 60% of whom said they felt overwhelmed by repetitive manual work. The resulting 90-day AI plan for coworking operators focused on immediate, practical use cases, including member support, community communication, unpaid invoices, room performance and daily reporting.

The Hamlet session examined the same pressure from the perspective of community teams. James Brouard explained that community managers are hired to work with people, yet much of their day is absorbed by bookings, billing, repeated questions, follow-up and internal coordination. The practical recommendation was to identify what AI should take off coworking teams’ plates so teams can spend more time on work that requires judgement, trust and human attention.

These are sensible starting points. They reduce administrative load and improve response speed.

They also generate new material for the organisation's memory.

Every summarised meeting, resolved request, qualified lead and approved exception adds another piece of context. The greater opportunity appears when the AI can retain that context, connect it across time and bring it into future decisions.

Most management decisions depend on how the history of the business is interpreted against its current financial position and the market around it.

Dashboards and monthly reports provide an important view of performance. A well-structured AI memory adds the context behind that performance: what the team knew at the time, which assumptions shaped its thinking, what decision was made, which external conditions influenced it, and what happened afterwards.

Over time, this creates a decision record that can be revisited and applied to new situations. Earlier pricing changes, product choices, expansion plans and commercial exceptions become part of a growing body of evidence rather than isolated moments in the company’s history.

The quality of the next decision improves because the AI can compare present conditions with previous choices and outcomes. It can identify which assumptions proved reliable, which signals were overlooked, and which circumstances produced stronger or weaker results.

This was also one of the important conclusions from the This was also one of the important conclusions from the Coworking Tech Week discussion with Nexudus CEO Carlos Almansa. AI can identify earlier retention and commercial signals when it has access to bookings, feedback, support, access activity and member behaviour. Almansa compared AI agents to new employees: they need onboarding, context and clear boundaries before they can contribute reliably. The wider conversation is covered in Where AI Actually Helps Coworking Operators Today.

The same principle applies to strategic work. The usefulness of the advice depends on the quality, structure and continuity of the memory behind it. An AI that can examine the history of the business, place it within current market conditions and connect decisions with their outcomes gives management a richer basis for judgement.

During the working day, AI may help the team answer questions, prepare documents, summarise meetings and complete defined tasks.

A dream cycle is dedicated to the memory itself.

At scheduled intervals, the AI reviews new material and improves the structure of the business memory. It reconciles records referring to the same member, company, location or issue, then connects conversations with contracts, decisions with outcomes, and recent signals with earlier events.

This process strengthens the collective memory of the space. It creates a clearer record of what the business knew, what it decided, what followed and which conditions shaped the result.

With that memory in place, the AI can analyse internal performance through a wider decision framework. It can compare the company’s history with current market conditions, identify where its position is strengthening or weakening, and show how internal choices relate to external change.

This is where the metaphor of dreaming becomes useful. The AI reorganises accumulated experience and returns it in a form that can guide the next decision.

Over longer periods, it can examine how the company behaves,and propose ajustments, encourage decisions and and propose adjustments eherr most needed.

Improving the space’s memory becomes part of improving the business.

An AI memory is only as dependable as the information it can access.

The Coworking Tech Week session with Koho.ai highlighted the commercial consequences of poor data management. Oliver Easton-Hughes explained how missing renewal dates, notice periods, break clauses, discounts, contract terms and customer-value information can create direct revenue risk.

The same applies to the wider memory of the organisation. When important decisions, promises and exceptions are poorly recorded, the business loses the ability to learn from them.

The article on why data health becomes revenue infrastructure shows how contract information, engagement, payments, sentiment and customer value affect renewal, pricing and account decisions.

AI expands the importance of that discipline.

A field that was once useful for reporting may now shape a recommendation. A missing decision note may prevent the AI from understanding why a commercial exception was approved. An inconsistent company name may split one account's history across several records.

Building an effective memory therefore begins with organisation.

Members, companies, leads and partners need consistent identities across tools. Contracts, renewals, discounts and notice periods need reliable fields. Commitments and decisions need owners. Expected outcomes need to be recorded. Later results need to be connected with the choices that produced them.

The quality of the memory determines the quality of the reasoning.

A coworking space can only dream with the information it has been taught to remember.

Most business reporting is organised around categories such as occupancy, revenue, leads, renewals, bookings, support requests and outstanding invoices.

Persistent AI memory could organise the same information around the decisions management needs to make.

A conventional report records what changed. A decision brief adds the history behind that change, connects it with earlier actions and external conditions, and explains which factors may require attention.

If occupancy declines, for example, the AI could examine how the change relates to pricing, product quality, customer behaviour, sales performance and local competition. It could then outline possible responses, supported by the relevant history, expected impact, commercial risk and unresolved questions.

The role of AI is to prepare the ground for better judgement by bringing together the context that is usually dispersed across reports, platforms, conversations and individual memory.

Founders and operators still decide which risks to accept, which relationships matter, which opportunities fit the company and what kind of business they want to build. The quality of that judgement improves when the relevant history is organised around the decision at hand.

Memory supports more than diagnosis.

Once the AI understands the current business and its history, it can explore what the company could become.

It can combine internal evidence with market analysis, local demand, workplace trends, competitor positioning, neighbourhood development and changes in the customer base.

This creates a different type of management conversation.

The AI may find that repeated pricing exceptions are evidence of an undefined product.

It may identify a cluster of companies from the same sector that use the space in similar ways and share unmet needs.

It may connect declining demand for one office type with increased requests for project rooms, hybrid-team access or shorter commitments.

It may recognise that an underperforming area of the building could support a product that has already appeared repeatedly in sales conversations.

The dream process can turn these observations into proposals. These proposals should come with evidence, assumptions and uncertainties.

A founder should be able to see which signals produced the idea, which parts are supported by internal history, which depend on external market information and what would need to be tested.

The aim is to create credible directions for consideration.

AI cannot reconstruct context that the business never captured.

A valuable memory therefore requires discipline in how the organisation records decisions.

Important decisions should leave behind a compact record:

• What decision was made?
• Who owned it?
• What evidence was considered?
• Which assumptions shaped the choice?
• What outcome was expected?
• When should the result be reviewed?
• What eventually happened?

The same principle applies to pricing exceptions, contract concessions, product changes, staffing choices, marketing experiments and member commitments.

This does not require another heavy administrative process.

A short decision note, a recorded meeting or a structured approval can provide enough material for the AI to preserve the reasoning. The AI can prepare the first version, identify missing information and ask the responsible person to confirm it.

Over time, the organisation builds evidence about its own judgement.

It learns which assumptions are usually accurate, where forecasts tend to be optimistic, which experiments deserve wider adoption and where the same mistakes keep returning under different names.

The company gains a clearer record of how it thinks.

Internal history provides only part of the context required for a strong decision.

A coworking business operates within a changing market. Employer behaviour, office supply, transport patterns, local development, interest rates, sector growth and competitor offers all influence performance.

A mature AI memory should connect internal developments with external conditions.

If tour volume falls, the AI should examine whether the change appears in one location, one product or the wider market.

If demand for larger offices grows, it should compare this with local hiring activity and changes in conventional lease availability.

If members begin requesting shorter terms, it should investigate whether the pattern reflects account-specific uncertainty or a broader change in business confidence.

This helps management distinguish between an internal problem and a market movement.

It also gives the business a way to recognise opportunities earlier.

Several weak signals may acquire strategic meaning when combined: enquiries from the same industry, repeated requests for a particular service, growth among existing companies in that sector, and new investment flowing into the local ecosystem.

A fragmented organisation may treat these as separate observations. An AI memory can assemble them into a hypothesis and propose a focused test.

The value of memory compounds.

A pricing decision produces an outcome. That outcome becomes evidence for the next pricing decision.

A product experiment creates a reference point for future locations.

A lost renewal improves the interpretation of similar account signals.

A successful event informs future community and commercial investment.

The business becomes easier to manage because each important action leaves behind usable knowledge.

This is especially valuable in founder-led companies.

Many coworking businesses depend heavily on the judgement of a small number of people. Their experience allows them to recognise patterns that remain invisible in formal reporting. It also creates a constraint as the company expands.

A persistent memory gives the wider team access to more of that history. It can preserve why earlier decisions were made and help new leaders understand the business without relying entirely on oral handovers.

The AI carries context forward. People continue to make choices involving relationships, values, ambition and risk.

A coworking business could build this capability through three recurring rhythms.

Continuously: Remember

Capture the facts, commitments, decisions, assumptions and outcomes that may matter later.

The AI connects information to the correct member, company, location, product or management question. It preserves the source so the team can trace where each conclusion came from.

Weekly: Interpret

Review recent activity against the accumulated history of the business.

The AI identifies emerging risks, recurring errors, unusual changes, unresolved decisions and opportunities that deserve management attention.

Its output is a focused decision brief rather than a complete replay of the week.

Monthly or quarterly: Dream

Reorganise the wider memory, evaluate earlier decisions and explore possible directions for the company.

The AI tests management assumptions against outcomes, combines internal evidence with market developments, identifies gaps in the organisation's knowledge and proposes changes to products, pricing, portfolio, sales, staffing or member experience.

This rhythm allows the memory to become more accurate and more useful over time.

An AI recommendation should carry its evidence with it.

Management needs to see which records were used, which conclusions are supported by facts, which are inferences, and where the AI has limited confidence.

Sensitive information requires clear access rules. Recommendations involving contracts, pricing, payments, member relationships or employee performance require human review.

The AI should also expose gaps in the business memory.

When a recommendation depends on incomplete records or conflicting information, it should say so. A credible memory does not create certainty where the organisation lacks evidence.

This transparency makes the AI easier to challenge and improves the quality of the underlying information. It prevents the dream from becoming a black box that produces attractive proposals without accountability.

The immediate AI use cases in coworking are useful and visible. They handle messages, summaries, administrative work and routine coordination.

The larger opportunity sits in the quality of management.

A coworking business with persistent AI memory can approach each decision with a clearer view of what happened before, why it happened, what the company learned and how the market has changed.

Its dream cycle can maintain that memory, question outdated assumptions and explore credible versions of what the business could become.

The result is a company that retains more of its own intelligence.

It remembers what the team believed. It connects decisions with outcomes. It recognises patterns across locations and years. It brings that accumulated understanding into the next pricing discussion, renewal strategy, product decision or investment choice.

Which decisions in your business are still being made without access to the history that shaped them?