What Telus's downtown Vancouver AI data centres mean for housing, energy, and the city.

Energy implications

The Telus cluster targets 150 MW from BC Hydro by 2032. That is 13.6 per cent of Site C's 1,100 MW nameplate capacity, BC's largest new generation asset, which came into full operation in August 2025 and powers 450,000 to 500,000 homes.

At a continuous draw, 150 MW equals the annual residential consumption of roughly 135,000 average BC Hydro households (150 MW × 8,760 hours ÷ 9.703 MWh per home per year, per the 2024-25 BC Hydro Quick Facts). The City of Vancouver had 305,336 occupied private dwellings in the 2021 Census (Statistics Canada), so the cluster's continuous load equals roughly 44 per cent of the residential consumption of every dwelling in the city.

BC Hydro is supply-constrained. The province imports U.S. electricity, some of it coal-fired, during peak demand periods. BC Green Party Leader Emily Lowan called Telus's "98 per cent clean energy" claim "dubious at best" on this basis and proposed a moratorium on new AI data centres until allocation rules are in place (Business in Vancouver).

Globally, the International Energy Agency projects data centre electricity consumption to roughly double from 485 TWh in 2025 to 950 TWh by 2030 (3 per cent of global electricity), with natural gas and coal supplying more than 40 per cent of the marginal demand growth.

Implication: the 150 MW allocation is material to BC's grid and to global data centre demand. Whether it is "clean" depends on what marginal source backfills it during peak periods, which the public record does not yet specify.

Sovereignty implications

The federal policy rationale for funding Telus is "sovereign AI": Canadian compute hosted on Canadian soil, owned by a Canadian carrier, reducing dependency on U.S. hyperscalers (AWS, Azure, Google Cloud). The Foundation supports the goal. The implementation question is whether the public benefit is captured by the public.

Three sovereignty risks visible in the public record:

• Compute access pricing. If the GPUs land at market rates, Canadian researchers and startups still pay U.S.-priced compute, just to a Canadian intermediary. The benefit accrues to brand and regulatory jurisdiction; the price end users see stays unchanged. The federal $300 million AI Compute Access Fund is supposed to subsidize access for Canadian innovators, but the price floor and eligibility rules are not yet public.
• Public-sector compute ringfence. Telus's announcement does not name how much of the 60,000 GPU capacity is reserved for public-sector or public-interest workloads (climate modelling, public health, language preservation, civil-society AI). Without a stated ringfence, the cluster is sovereign in jurisdiction but not in purpose.
• National security dependency. The cluster's HCMA-designed building runs on NVIDIA hardware procured through a foreign supplier whose export controls are set in Washington. "Sovereign AI" with foreign chips is a partial sovereignty.

Implication: sovereignty is a real public good and worth supporting, but only if the federal funding instrument enforces public access, public-interest ringfencing, and hardware-supply resilience clauses. Absent those, the policy buys a Canadian flag on a U.S.-stack data centre.

Labour implications

Telus's headline labour claim: "1,000+ construction jobs and hundreds of long-term technology and operations roles" across all three B.C. sites (Kamloops, M3, 150 W Georgia).

Hyperscale data centre operations are labour-light. Industry benchmarks (Latitude Media, Hamm Institute) put long-run operations at 0.15–0.35 full-time-equivalent employees per megawatt. Applied to 150 MW that yields a long-term on-site headcount of roughly 22 to 50 permanent operations staff. "Hundreds" in the Telus figure likely includes Telus corporate roles attributed to the project, customer-success and engineering positions that may exist regardless, and subcontractor headcount.

Construction labour intensity is much higher (0.7–2.0 workers per MW during build) and consistent with the 1,000+ construction-jobs figure across multi-year phasing.

Capital per long-term job is the more informative ratio. Telus declined to disclose project capex, but recent AI-ready hyperscale builds run roughly $20–$25 million per megawatt of IT capacity, with full campus costs (land, power, equipment) reaching $45–$55 million per MW (2026 modern data centre cost benchmarks, AI CERTs 2025 economics of hyperscale builds). The 150 MW Telus cluster therefore sits in the $3–$4 billion construction-cost range, broadly consistent with Telus's $9 billion projected B.C. economic activity at typical construction-multiplier rates. Divided by 22–50 long-term operations jobs, that is roughly $60–$180 million of capital per permanent role created, of which a meaningful share would come from public funds if the federal envelope is committed.

Implication: the labour case for AI data centres is dominated by transient construction employment, with a small permanent footprint. Public funding decisions should weight this honestly. The $9 billion projected economic activity is real: capex flows into local trades, equipment, and services. That figure captures construction multipliers; the steady-state employment footprint stays small.

Urban implications

Both sites convert older or stalled commercial inventory; neither directly displaces existing housing. That is the cleanest urban-impact claim Telus can make, and it is true. The harder claim is the counterfactual: what could these sites have produced instead?

111 East 5th Ave (M3). Built 1977. Six-storey commercial building, the long-term Vancouver headquarters of Hootsuite (which leased it from 2014 and has now mostly vacated). Acquired by Westbank and Hootsuite CEO Ryan Holmes in 2015 as part of the Main Alley tech campus (full city block purchased for $40.5 million). Rezoning underway from I-1A (Industrial) to CD-1 to permit "Bulk Data Storage." Public Q&A ran January 28 to February 10, 2026 (CityHallWatch, Shape Your City Vancouver). M3 had previously been slated for an expanded creative-office use with three additional floors above the existing structure.

150 West Georgia St (formerly 720 Beatty). Owned by Creative Energy, above the downtown steam plant. The site has cycled through multiple proposals since 2020:

• October 2020: Vancouver City Council approved a 17-storey, 580,000 sq ft office tower (Bjarke Ingels Group / HCMA). Plan stalled with the post-pandemic office market.
• 2023: A CD-1 (818) text amendment added "bulk data storage" as a permitted use (Council document).
• 2025: An Align Architecture mixed-use concept (hotel + residential + data centre) was abandoned.
• 2026: Now proposed as a 10-storey, 400,000 sq ft pure data centre with Telus.

The 2020 and 2025 concepts contained office, hotel, or residential uses that would have served physiological-level urban need. The 2026 proposal contains none. Westbank has contracted Buy Social Canada for Community Benefit Agreement implementation; terms and dollar value are not yet public.

Downtown Vancouver context:

• Office vacancy: 12.6 per cent in Q3 2025 (CBRE), highest since 2004; Class B/C buildings worst at 17.2 per cent (Daily Hive). Supply currently outpaces demand; both Telus sites convert older or stalled commercial inventory, neither displaces existing housing directly.
• Rental market: Metro Vancouver purpose-built rental vacancy reached 3.7 per cent in 2025 (CMHC), the highest since 1988. Mount Pleasant in particular is seeing landlord incentives.
• Homelessness: the City of Vancouver counted 2,715 unhoused residents in 2025, a 12 per cent increase over 2023 and a record high. Metro Vancouver total: 5,232 (2025 Point-in-Time Homeless Count). Since 2005 the regional count is up 141 per cent.
• Food insecurity: Greater Vancouver Food Bank recorded 832,000 client visits in fiscal 2025 (2025 Impact Report), serving more than 15,000 people monthly.

Implication: the sites do not displace housing, but they consume parcels and floor area whose prior approved or proposed uses would have addressed acute physiological-level urban need. The counterfactual cost matters.

Where Telus sits on the quadrant

The Transmutarianism framework scores agents on F (filtering of deprivation: deprivation absorbed without being passed on) and A (amplification of fulfillment: fulfillment emitted in excess of what was received). Moral work M = τF + A is computed per Maslow level (physiological, safety, belonging, esteem, actualization) and weighted by w = {5, 4, 3, 2, 1}.

The chart below plots Telus's provisional placement (sigil-red outline) against the four archetype reference dots (muted, shown for orientation). Drag the τ slider to test sensitivity. Click any dot for its F, A, and M values.

Assumptions, stated

1. Scope of evaluation. The two downtown Vancouver sites (M3 at 111 East 5th Ave and 150 West Georgia), in their proposed and announced configuration as of May 2026. The Kamloops expansion is excluded because it sits outside the downtown denominator and is a phased expansion of an existing facility.
2. Time horizon. Steady-state at full build-out (2032). Construction-phase effects (the 1,000+ construction jobs, the construction-multiplier component of the $9 billion B.C. economic activity figure) are real but transient and are not scored against the steady-state dot.
3. Asymmetry coefficient (τ). Default τ = 1 (filtering and amplification weighted equally). Sensitivity at τ = 0.8 (flourishing focus) and τ = 1.5 (cycle-breaking focus) is shown in the math box below.
4. Maslow weighting. The published weights w = {5, 4, 3, 2, 1} for {physiological, safety, belonging, esteem, actualization}. Lower-level deprivation is heavier than higher-level deprivation.
5. F and A scale. Each level scored on a −10 to +10 range, central estimate from the public-record evidence in the prior sections. The reasoning column of the per-Maslow table names the inputs to each value so any number is substitutable with a sourced alternative.
6. Denominator. "Net to downtown Vancouver" is the recipient population. Flows to Telus shareholders, BC's general tax base, U.S.-headquartered hyperscaler competitors, Canadian AI researchers outside the host city, and international AI markets are excluded from per-level scoring. They are real and material; they sit outside this evaluation's scope and would carry into a parallel audit done at the provincial or national level. The narrow denominator tilts the per-level scoring toward the negative by construction: a project that delivers fulfillment to a wider geography (national sovereign-compute access, provincial economic activity) does not receive credit for those flows in a downtown-Vancouver-scoped evaluation. Readers comparing this placement to a hypothetical national-scope audit should expect that wider scope to shift the dot toward Transmuter.

Per-Maslow scoring

The aggregate dot above is computed from the per-level table below. Each F and A value is the central estimate from the public-record evidence in the prior sections; substitute different numbers and the dot moves.

The verdict does not flip across reasonable τ values, because the physiological-level deficit is in A (amplification), which is not multiplied by τ. To move the verdict, the underlying scores have to change. The Foundation will publish a re-audited placement if Telus, BC Hydro, ISED, or any independent party supplies the six audit inputs listed in the recommendations section below.

Public-interest recommendations

The federal funding instrument is not yet committed. The negotiation between ISED and Telus is the moment when public-interest conditions can be written into the deal. The following are the highest-leverage interventions visible from the public record.

For the Government of Canada (ISED)

1. Condition the federal contribution on physiological-level outcomes. Tie a measurable fraction of the federal disbursement to verified delivery of housing units, food infrastructure, or energy-cost reduction in the host city. Heat recovery is a candidate mechanism: a per-GWh-delivered payment to Creative Energy that flows through to building tenants converts the marketing claim into an enforceable obligation.
2. Ringfence a public-interest GPU share. Reserve a defined percentage of the cluster's 60,000 GPUs for public-sector and civil-society workloads at cost or at zero cost. Without a ringfence, sovereign AI infrastructure becomes commercial AI infrastructure with a Canadian flag.
3. Publish the contract. The funding instrument (grant, loan, contribution agreement, compute-procurement contract) and its conditions should be public. Public capital with private conditions is a category error.

For BC Hydro and the BC Utilities Commission

1. Price the displacement. A weighted load tariff for large industrial customers in supply-constrained service territory should reflect the cost of the next 150 MW of marginal capacity, which would otherwise serve residential growth and building electrification.
2. Disclose the marginal source. What backfills 150 MW during peak periods? If the answer is imported coal-fired power, the "98 per cent clean energy" claim needs qualification in the service agreement and in public communications.
3. Apply the heat-recovery credit honestly. Recovered heat to Creative Energy displaces natural gas; that displacement is a real climate credit. Quantify the delivered GWh per year and publish the comparison against the marginal-electricity emissions of the data centre's draw. The two numbers together give the project's actual carbon position.

For the City of Vancouver

1. Make the Community Benefit Agreement Maslow-tagged and dollar-quantified. Westbank's contract with Buy Social Canada is the right instrument. The output should specify which line items address which Maslow-level needs (physiological housing, safety services, belonging infrastructure, esteem opportunities, actualization access) and the dollar value of each.
2. Preserve the mixed-use option at 150 W Georgia. The site has cycled through multiple proposals containing housing and hotel uses. The current pure-data-centre concept is in pre-application; the rezoning is the moment to require a residential or community-serving floor-area share.
3. Track the office-vacancy backstop. Downtown office vacancy is 12.6 per cent. If approving 477,000 sq ft of office-to-data-centre conversions destabilizes adjacent office values, the municipal property-tax base shifts. Model the second-order effect before approving.

For Telus and Westbank

1. Publish capex, GWh consumption, GWh recovered, and FTE counts annually. Voluntary transparency on these four numbers is the cheapest credibility purchase the project can make.
2. Open the audit. The Foundation will publish a fully audited placement with sourced inputs supplied by Telus. The provisional placement above stands until the audited one replaces it.

The audit protocol (six inputs)

An audited placement (as distinct from the provisional one published here) requires six inputs. The Foundation will provide framework, scoring template, methodological support, and the published report to any party willing to supply them: Telus, Westbank, ISED, BC Hydro, the City of Vancouver, an independent researcher, or a civil-society organization with standing.

1. Audited 2032 capital, opex, and revenue allocation by site.
2. BC Hydro service agreement: rate class, demand pattern, peak draw, contracted vs interruptible portions, any private power-purchase agreements.
3. Actual heat-recovery output to Creative Energy: GWh per year delivered, with the pass-through path to tenant utility bills.
4. Long-term FTE count by site and by role, with a breakdown of net-new vs relocated positions.
5. Community Benefit Agreement terms and dollar value, with each line item Maslow-tagged.
6. Federal funding amount and instrument once committed, with any community-benefit conditions documented.

Contact: sev@economyofwisdom.com.

What changes the placement

This is the key recommendation of the field note. The Telus dot is provisional, and it will move with the data each party at the table publishes between now and the federal funding decision. Every item below names a specific public output the Foundation is asking for; absent these, the placement stands and the public conversation continues to run on press-release figures.

Moves it toward Transmuter:

• Verified heat-recovery delivery at scale (metered GWh/year).
• A residential or community-serving floor-area share at 150 W Georgia.
• A dollar-quantified Maslow-tagged cost-benefit analysis.
• A federal funding instrument with enforceable physiological-level conditions.

Moves it toward Extractor:

• Imported fossil power backfilling BC Hydro draw at peak.
• Federal funding terms that subsidize shareholder returns without enforceable community-benefit conditions.
• Heat-recovery underperformance against the 150,000-home equivalent claim.

Better data moves the dot. The audit protocol above is the Foundation's standing offer to score whichever inputs are produced, by whichever party produces them.

A note on framing

The Foundation publishes this evaluation in service of the negotiation that is currently happening between Telus, ISED, BC Hydro, and the City of Vancouver. The intent is to give every party at the table a shared instrument, sourced and reproducible, that scores the public benefit of the project in terms more honest than "$9 billion in economic activity" and more rigorous than "feels like a giveaway."

An Absorber classification describes a system state of high filtering and low amplification: the agent absorbs deprivation from its inflows but emits less fulfillment than it receives. In framework shorthand, this is "the caregiver running on empty" at the institutional scale. For Telus, the positive F traces to heat recovery into Creative Energy's downtown steam loop and to sovereign-infrastructure filtering of dependency on U.S. hyperscalers; the negative A traces to the constrained-grid and constrained-land allocation effects on downtown's physiological flows. The framework reports flows. People inside institutions are working within incentive structures that the framework's purpose is to make visible. The recommendations above are addressed to the institutions.

The math is at transmutarianism.org/framework/. The methodology is at transmutarianism.org/use-cases/anchor-tenant-audit/. The Telus dot is plotted on the live quadrant explorer at transmutarianism.org/quadrant/. If you want to dispute the placement, the framework is published and the math is fully shown above. Substitute different F and A values per level, with the source for each, and the dot moves.