英文旁白、无字幕的 3–4 分钟管理层简报。3–4 minute executive brief, English narration, no captions. 在 YouTube 打开Open on YouTube ↗
起步顺序Order of work 先选一个打包方案,再建交付团队,最后做政府与媒体方案。Pick one packaged solution first, then a delivery bench, then government and media offers.
公开事实Public facts 已核对Checked
- 500 MW+Bell AI Fabric 总算力目标Target capacity for Bell AI Fabric in total加拿大最大的主权 AI 算力项目Canada's largest sovereign AI compute project
- 300 MWRegina 附近规划中的 AI 数据中心AI data centre planned near Regina已披露租户:Cerebras 与 CoreWeaveTenants named: Cerebras and CoreWeave
- 7 MW首个推理设施 Kamloops,2025 年 6 月上线First inference site, Kamloops, live June 2025与 Groq 合作Built with Groq
- 41%AI 解决方案收入同比增长Year-over-year growth in AI solutions revenue2025 年 Q3 财报披露Reported with Q3 2025 results
需要完成的三个转变Three shifts to make 判断Judgment
- 卖算力容量Selling compute capacity → 卖客户用得上的成果Selling outcomes customers can use
- 有租户的数据中心A data centre with tenants → 跑在其上的应用Applications running on top of it
- 客户自己做集成Customers integrate alone → 贝尔提供的交付团队A delivery bench Bell provides
专家组裁决Expert panel 依据公开观点的推演Inference from public views
Foteini Agrafioti
每个方案挂到收入目标;写明 AI Fabric 团队与销售的分工。Tie each solution to a revenue target; write the split between the AI Fabric team and sales.
Erik Brynjolfsson
客户价值在流程与技能;把这些打包定价,而不只卖算力。Customer value sits in process and skills; package and price those, not only compute.
Ziad Obermeyer
面向客户的方案要写明目标与代理指标,并保留清单。For customer-facing solutions, state the objective versus the proxy and keep an inventory.
Ethan Mollick
客户员工会摸到边界;用真实任务办入门工作坊。Customer staff will find the edges; offer starter workshops on real tasks.
Andrew Ng
为一类客户做一个打包方案,三到六个月内出数字。One packaged solution for one customer type, with numbers in three to six months.
Thomas Davenport
这是从卖算力转向卖成果的商业模式变化;每个方案点名负责人。This shifts the model from selling capacity to selling outcomes; name an owner per solution.
Thomas Davenport:先定商业模式。Decide the business model first.
Andrew Ng:先交付一个方案,模式随之清晰。Ship one solution first; the model will follow.
→ 现在选一个方案;模式在第一道关上再定。Pick one solution now; decide the model at the first gate.
六位专家是谁、怎样思考、有什么铁律 →Who the six experts are, how they think, their rules →
三个目标Three priorities 建议Recommendation
打包成果,而非算力Package outcomes, not capacity
在 AI Fabric 上为一类客户做一个方案。One solution for one customer type, built on AI Fabric.
建设交付团队Build a delivery bench
为中型客户提供入门项目与集成伙伴。Starter programmes and integrators for mid-size customers.
把 AI 带进 Bell MediaBring AI to Bell Media
CTV 与 Crave 上的推荐与广告定向。Recommendations and ad targeting on CTV and Crave.
场景组合Scenarios 建议Recommendation
| 场景Scenario | 做法Approach | 状态Status |
|---|---|---|
| 打包的企业方案Packaged enterprise solutions 增收Revenue | 把算力、托管与一个方案打包卖给一类客户。Compute, hosting and one solution sold together to a customer type. | 建议 · 基于已上线的 AI FabricProposed · builds on live AI Fabric |
| AI 入门项目AI starter programme 增收Revenue体验Experience | 为中型客户提供工作坊与集成支持。Workshops and integration help for mid-size customers. | 建议Proposed |
| 政府托管方案Government hosting offer 增收Revenue降风险Risk | 面向政府与受监管客户的主权托管方案。A sovereign hosting offer for government and regulated customers. | 建议 · 完整审查Proposed · full review |
| 媒体推荐与广告Media recommendations 增收Revenue体验Experience | CTV 与 Crave 的推荐与广告定向。Recommendations and ad targeting on CTV and Crave. | 建议Proposed |
12 个月路线12-month roadmap 建议Recommendation
奠基Foundation
- 选定一个方案与客户类型Pick one solution and customer
- 点名产品负责人Name the product owner
- 起草定价逻辑Pricing logic drafted
试点Pilot
- 首个客户试点First customer pilot
- 客户员工真实任务测试Customer staff test on real tasks
- 与客户共同度量价值Value measured with customer
上线Go live
- 方案卖给更多客户Solution sold to more customers
- 培训交付伙伴Delivery partners trained
- 上线即监控Monitoring from day one
扩展Extend
- 第二个方案Second solution
- 政府方案评审Government offer review
- 年度重排优先级Annual re-ranking
治理与合规Governance 建议Recommendation
- 客户数据按合同约定留存Customer data stays where the contract says
- 面向客户的方案:记录目标并建清单Customer-facing solutions: objective and inventory recorded
- 与法律顾问确认适用的数据与模型风险规则Confirm applicable data and model-risk rules, with counsel
三层指标Three-layer metrics 建议Recommendation
| 采用Adoption | 使用方案的客户数Customers active on solutions · 工作坊参与度Workshop attendance |
|---|---|
| 过程Process | 首次见效时间Time to first value · 每客户交付工时Delivery hours per customer · 方案复用率Solution reuse rate |
| 结果Outcome | AI 解决方案收入AI solutions revenue · 每兆瓦收入Revenue per megawatt |
目标值由企业自定。Targets are the company's to set.
风险Risks 判断Judgment
有算力无客户Capacity without customers
光有算力卖不动;要有方案与交付。Compute alone does not sell; solutions and delivery do.
交付瓶颈Delivery bottleneck
能为中型客户做集成的人太少。Too few people can integrate for mid-size customers.
主权承诺Sovereign promises
数据驻留承诺须在每份合同里兑现。Data-residency claims must hold in every contract.
前 90 天First 90 days 建议Recommendation
| 第 1–2 周Weeks 1–2 | 点名产品负责人;选客户类型Name the product owner; pick a customer type |
|---|---|
| 第 3–6 周Weeks 3–6 | 起草方案与定价逻辑Draft the solution and pricing logic |
| 第 7–12 周Weeks 7–12 | 招募试点客户;定价值指标Recruit a pilot customer; set value measures |
| 第 13 周Week 13 | 首次复盘First review |
我们可能错在:We could be wrong if: 客户其实只要算力本身,或合作伙伴早已占据应用层。customers only want raw compute, or partners already own the application layer.
本页是基于公开资料的分析,与所述企业没有合作关系,不代表其官方立场或内部计划;企业数字多来自其自述或媒体转述;「专家组」是依据六位专家公开观点的推演,不是他们的原话,也不代表他们或其雇主的立场。资料整理日期:2026-11-14。This page is an analysis based on public information. It has no affiliation with the company and does not represent its official position or internal plans. Company figures come mostly from its own statements or media reports. The expert panel is an inference from six experts' public views, not their words, and does not represent them or their employers. Compiled 2026-11-14.
想给你的公司也做一份?Want one for your company?
同样的方法:公开事实、专属判断、12 个月路线和前 90 天。公司大小都可以聊,中文沟通。Same method: public facts, a company-specific judgment, a 12-month roadmap and the first 90 days. Any size of business; we can talk in Chinese or English.