Robotics: Four Deep Questions Evidence Synthesis v1
Date: 2026-06-13 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact Visibility: PUBLIC Target: site + slide Public-safety: yes. No private portfolio data. No paid-report excerpts. No trade recommendation. Source discipline: 🟢 primary / 🟡 secondary / 🟠 estimate / 🔴 guess
0. One-line answer
Hugo 指定的四个问题其实应该合成一个 public-safe field-guide section：过去三年中美机器人不是同一条曲线；中国在 manufacturing deployment scale、国内供应链渗透和 2025/2026 资本热度上更强，美国在 humanoid frontier companies、AI/model/compute stack 和高额 private mega-round 上更强。Tesla Optimus 是 capacity-intent / vertical-manufacturing curve；Figure 是 deployment-KPI + manufacturing-ramp curve；绿的谐波是 filing-backed component-volume curve。四者都值得跟踪，但公开证据仍大多停在 S3/S4，不是 S5 scaled-commercial-economics proof。🟢/🟡/🟠 as of 2026-06-13.
1. Core framing: do not compare “机器人” as one market
The right comparison has four layers:
- Installed industrial automation base: who is already deploying robots in factories.
- Humanoid / embodied-AI capital: who is funding new-generation platforms.
- Product / deployment evidence: who has moved beyond demos into pilots, production lines, and customer-site KPI.
- Supply-chain value capture: who can turn volume into revenue, margin, and bottleneck power.
A single “China vs US robotics winner” answer would be misleading. The evidence says China currently has the larger factory robot deployment flywheel, while the US has several high-signal frontier stacks around Tesla / Figure / Apptronik / Agility / NVIDIA / foundation models. 🟠 synthesis from sources below.
2. Question 1 — 中美机器人这轮发展比较：capital scale + industrial scale, last 3 years
2.1 Industrial deployment scale: China is already much larger
| Metric | China | United States / North America | Read-through |
|---|---|---|---|
| 2024 industrial robot installations | 295,000 units | US 34,200 units | China installed about 8.63x the US count. 🟢 IFR; 🟠 ratio = 295,000 / 34,200 |
| 2024 share of global industrial robot installations | 54.4% | US 6.3% | China is the central deployment market; US remains important but much smaller in factory robot installations. 🟢 IFR; 🟠 share calculations |
| 2024 global installations | 542,000 units | n/a | Global market remains large and mature enough to support supplier learning curves. 🟢 IFR |
| 2024 China operational stock | 2.027m industrial robots | US operational stock roughly 393,700 per IFR press-derived secondary references; US installation official extract gives 34,200 in 2024 | China stock crossed 2m and doubled in 3 years from >1m in 2021 to >2m in 2024. 🟢 IFR China press release |
| 2024 China domestic supplier share in China | 57%, up from 47% in 2023 | n/a | Domestic Chinese suppliers overtook foreign suppliers in China for first time; +10pp YoY. 🟢 IFR |
| 2024 North American robot orders | n/a | 31,311 robots, US$1.963bn | NA orders were roughly flat YoY: +0.5% units, +0.1% value. 🟢 A3 |
Signal: China’s advantage is not only policy narrative. It is visible in actual factory robot installations, operational stock, and domestic supplier share. 🟢
Noise: industrial robot installation scale is not the same as humanoid commercialization. Industrial arms / SCARA / cobots / AMRs do not automatically prove general-purpose humanoid economics. 🟠
2.2 Capital scale: US has mega-round frontier concentration; China has broad 2025 embodied-AI funding acceleration
| Evidence | US / Global frontier read | China read | Caveat |
|---|---|---|---|
| F-Prime State of Robotics 2025 | North America / Europe / Israel robotics + AV companies raised $70bn since 2020; 2024 investment more than doubled from 2023 and approached 2021 highs; mega-rounds >$50m drove rebound. 🟡 F-Prime / PitchBook / CapIQ methodology | Not China-covered; useful for Western frontier robotics capital, not direct China comparison. | Scope excludes China; includes robotics + AV, not pure humanoids. |
| Figure Series B | Figure raised US$675m at US$2.6bn valuation in Feb 2024, with Microsoft, OpenAI Startup Fund, NVIDIA, Bezos Expeditions and others. 🟢 Figure / PRNewswire company release | n/a | Funding does not equal deployment economics. |
| China robotics funding 2026 YTD | n/a | Crunchbase News says China-based robotics companies raised US$5.6bn across 176 deals through mid-May 2026; robotics startups accounted for US$3.3bn across 126 deals in China Q1; China’s domestic robotics market accounts for >43% of global robotics venture investment. 🟡 Crunchbase News | Current-year 2026; not full-year; Crunchbase category definitions may include broad robotics. |
| China humanoid funding 2024 | n/a | Caixin summary found in search results says China humanoid sector raised RMB 8.45bn across 71 deals in 2024; global humanoid funding reached US$1.1bn in 2024. 🟡 search-result snippet, needs full extraction before use in final public copy | Treat as secondary lead unless full article accessible. |
| Policy / ecosystem | USCC says humanoid development is led mainly by US and China; no general-purpose autonomous humanoid is commercially viable at scale today. 🟢 USCC | USCC says China applies subsidies, tax breaks, zones, government-backed centers, adoption incentives and public-private collaboration; MIIT aims world-class humanoid innovation ecosystem by 2025 and integration by 2027. 🟢 USCC | Policy support can inflate capacity and concept activity; needs deployment economics. |
Socratic conclusion:
- China is ahead in existing manufacturing-robot deployment and likely has a deeper component-manufacturing learning environment. 🟢/🟠
- US is ahead or at least unusually concentrated in AI/model/compute + frontier humanoid mega-rounds: Tesla, Figure, Apptronik, Agility, NVIDIA, Physical Intelligence, Skild. 🟢/🟠 from existing robotics files.
- China’s 2025/2026 embodied-AI funding surge is a real signal, but public-safe copy should avoid overstating it as “mass commercialization proven.” 🟡/🟠
What would change our mind:
- China: named paid humanoid deployments with runtime, intervention rate, ROI/payback and repeat orders across multiple factories.
- US: beyond Figure BMW / Tesla capacity evidence, more multi-customer paid deployments and cost/yield/margin disclosure.
- Capital comparison: a complete 2023/2024/2025 database by country, category, round size, and stage; current public sources use different scopes.
3. Question 2 — Tesla Optimus versions: timeline, function, structure, evidence grade
This is a public-safe version comparison using official / filing-backed evidence and existing Tesla artifacts. It is not a TSLA recommendation.
| Version / milestone | Public timing | Function / positioning | Structure / manufacturing evidence | Evidence grade | What it proves | What it does not prove |
|---|---|---|---|---|---|---|
| Tesla Bot concept | 2021 AI Day announcement | General-purpose humanoid concept tied to Tesla AI. | Concept-stage; not a working production robot. | 🟢 for announcement; 🟠 for implication | Tesla publicly entered humanoid robotics. | No product, deployment, production, economics. |
| Early Optimus prototype | 2022 AI Day | Working prototype / walking demo; early manipulation. | Prototype-stage hardware; not production-optimized. | 🟢 Tesla event/video | Moved from concept to embodied prototype. | No validated factory deployment or production economics. |
| Optimus Gen 2 | Tesla official YouTube published 2023-12-13 | Demo showed improved walking / balance / manipulation, including delicate egg handling in secondary reports. | New generation prototype; public sources emphasize improved performance, not production line. | 🟢 Tesla official video; 🟡 Ars / Electrek secondary summaries | Technical iteration speed and dexterity progress. | Does not prove deployment, uptime, cost, or revenue. |
| Production-line / Gen 3 mass-production intent | Q3 2025 / Q4 2025 SEC filings | First-generation production lines being installed; Gen 3 meant for mass production; production-primed design. | Supply-chain readiness and production-line preparation; planned start before end-2026 in Q4 2025 filing. | 🟢 Tesla SEC filings | Stronger S4 manufacturing-infrastructure signal than demo videos. | “In anticipation” and “planned” are not actual production. |
| Fremont 1M/year designed line | Q1 2026 SEC filing | First large-scale Optimus factory preparation; replace Model S/X lines in Fremont. | First-generation line designed for 1m robots/year. | 🟢 Tesla Q1 2026 Form 8-K Exhibit 99.1 | Named site + designed-capacity anchor. | Designed capacity ≠ current production rate, accepted units, or customer demand. |
| Texas 10M/year long-term line | Q1 2026 SEC filing | Long-term second-generation line preparation. | Giga Texas prepared for second-generation line designed for long-term annual production capacity of 10m robots. | 🟢 Tesla Q1 2026 Form 8-K Exhibit 99.1 | Tesla’s scale ambition is extreme and filing-backed. | Not a dated production forecast, revenue forecast, margin claim, or supplier confirmation. |
Current stage: Tesla Optimus is best labeled S4 manufacturing-infrastructure / designed-capacity evidence. It is not S5 until Tesla or customers disclose actual output, accepted units, internal task ROI, intervention rate, external paid customers, robot revenue, gross margin, and safety / reliability statistics. 🟢/🟠
Common misconception:
- “1m / 10m means current production.” Wrong. Existing Tesla filing audit notes Tesla’s own capacity-table caveat: installed capacity is not current production rate. 🟢
4. Question 3 — Figure history and product evolution
Figure is the cleanest public example of a private humanoid company moving from funding / prototype to customer-site KPI and manufacturing KPI.
4.1 Timeline
| Date | Event | Quantified evidence | Signal grade | Read-through |
|---|---|---|---|---|
| 2022 founding window | Figure company release in Feb 2024 described Figure as founded 21 months earlier and based in Sunnyvale with 80 employees. | 80 employees; founded 21 months before Feb 2024 release. 🟢 | S2/S3 | Team formation speed signal. |
| 2024-02-29 Series B | Raised US$675m at US$2.6bn valuation; investors included Microsoft, OpenAI Startup Fund, NVIDIA, Bezos Expeditions, Intel Capital, ARK and others. | US$675m; US$2.6bn valuation. 🟢 | S4 capital signal | Frontier capital + AI partner concentration; not commercialization proof. |
| BMW deployment result | Figure post dated 2025-11-19 | 11-month Figure 02 deployment; 10-hour shifts Monday-Friday; 90,000+ parts loaded; 1,250+ runtime hours; 30,000+ BMW X3 vehicles contribution. 🟢 | S4 deployment KPI | Strongest public Figure evidence; still lacks customer ROI/payback and intervention rate. |
| BotQ announcement | Figure post dated 2025-03-15 | First-generation line capable of up to 12,000 humanoids/year; supply chain can scale to 100,000 robots or 3,000,000 actuators in next four years. 🟢 | S4 capacity-intent | Manufacturing ambition; capacity ≠ actual output. |
| Figure 03 production ramp | Figure post dated 2026-04-29 | 350+ Figure 03 robots delivered; 1/day to 1/hour cadence; >80% EOL first-pass yield; 99.3% battery-line first-pass yield; 9,000+ actuators. 🟢 | S4 manufacturing KPI | Actual manufacturing metrics are unusually concrete; still need sell-through/utilization/economics. |
| Catalyst commercial agreement | Figure post dated 2026-05-26 | Named distribution/logistics customer; starts at Reno DLC. 🟢 | S3/S4 | Second vertical signal; missing robot count, value, timeline, KPI. |
| Figure 03 launch | Figure post dated 2025-10-09 | 3rd-generation humanoid; 2x frame rate, 1/4 latency, 60% wider field-of-view per camera, 10Gbps mmWave data offload, UN38.3 battery certification. 🟢 | S3/S4 product signal | Product architecture optimized for Helix/home/scale; field economics still missing. |
4.2 Product evolution
| Product / generation | Main question it answers | Strongest public evidence | Remaining gap |
|---|---|---|---|
| Figure 01 / early platform | Can Figure build a humanoid platform quickly? | Funding/team/company-history evidence; prior product not the strongest current public anchor. 🟢/🟠 | Less relevant than Figure 02/03 for deployment/economics. |
| Figure 02 | Can a humanoid perform repeated industrial work in a named customer site? | BMW 11-month deployment, 1,250+ runtime hours, 90,000+ parts loaded. 🟢 | ROI/payback, intervention rate, robot count, repeat order. |
| Figure 03 | Can Figure design for scale, home use, data flywheel, and manufacturing? | Figure 03 launch; BotQ production ramp; 350+ robots delivered; 1/hour cadence. 🟢 | Economic utilization, reliability distribution, gross margin, customer deployments. |
| Helix | Can VLA / autonomy reduce task engineering and intervention? | Helix 02 official autonomy claims and product integration in Figure 03. 🟢 for company claim | Long-duration field autonomy and customer-site intervention-rate reduction. |
Figure’s key difference versus Tesla: Figure has stronger public customer-site operating KPI; Tesla has larger filing-backed designed-capacity ambition. They answer different questions and should not be compressed into a winner ranking. 🟠
5. Question 4 — 绿的谐波：能否成为 “宁德时代之于电车 / NVIDIA 之于 AI”？
5.1 Short answer
可能成为中国 humanoid / embodied-robotics 供应链里的重要 bottleneck supplier，但公开证据暂时不足以支持 “CATL-of-EV / NVIDIA-of-AI” 级别的强类比。更准确的 public-safe phrase 是：绿的谐波是目前最清晰的 filing-backed component-volume signal 之一；是否能升级为 platform-defining supplier，要看 humanoid-specific revenue, named customer, architecture durability, margin durability, and value-content expansion. 🟢/🟠
5.2 Why it could work
| Bull condition | Current evidence | Grade | Why it matters |
|---|---|---|---|
| Real revenue and volume acceleration | 2025 revenue RMB 570.714m (+47.31%); net profit RMB 124.367m (+121.42%); harmonic reducer sales 425,158 units (+72.48%). | 🟢 2025 annual report | Unlike pure concept names, the operating signal exists in filings. |
| Robot-relevant revenue line | “工业及具身智能机器人零部件” revenue RMB 422.528m (+52.61%), 34.88% GM; about 74.0% of company revenue. | 🟢 filing; 🟠 ratio = 422.528 / 570.714 | Large company-level exposure to robot components, though not humanoid-only. |
| Product breadth beyond reducers | Harmonic / micro harmonic reducers, rotary actuators, frameless torque motors, PRS, linear actuators, planetary joint modules, DD motors. | 🟢 annual / sustainability report | Architecture hedge if value moves from discrete reducer to integrated actuator / linear / direct-drive modules. |
| China deployment flywheel | China installed 295,000 industrial robots in 2024, 54% of global installations; domestic suppliers reached 57% share. | 🟢 IFR | Strong domestic automation market can support supplier learning and localization. |
5.3 Why CATL/NVIDIA analogy can fail
| Failure mode | Why it matters | Current evidence / gap |
|---|---|---|
| Humanoid architecture may not standardize around harmonic reducers | CATL benefited from large, standardized battery value content; NVIDIA benefits from compute/software ecosystem. Humanoid joints may split between harmonic, planetary, cycloidal, PRS, direct drive, cable, integrated actuators, OEM self-design. | Product fit exists, but no public OEM BOM proving dominant share. 🟢/🟠 |
| OEMs may vertically integrate critical modules | Tesla / Figure evidence points to in-house design and manufacturing for core modules; this can reduce external complete-module supplier power. | Figure BotQ says much robot design from scratch; Tesla supplier map undisclosed. 🟢/🟠 |
| Component supplier value content may be lower and more competed | NVIDIA captures high-margin platform/software/network effects; reducer suppliers may face ASP pressure as robot ASPs fall. | Leaderdrive GM declined from 41.14% in 2023 to 36.91% in 2025; robot component line GM 34.88%. 🟢 |
| Customer evidence is not named | CATL/NVIDIA analog needs visible design wins and repeat purchasing by leading customers. | No reviewed public named humanoid OEM, contract value, order volume, humanoid-only revenue, or single-robot value content. 🟢/🟠 |
| Stock-price narrative can outrun operating evidence | A supplier can be real and still over-hyped. | 2026-06-04 to 2026-06-08 share price +36.00% vs SSE Composite -3.05%; company said no undisclosed major matter. 🟢 |
5.4 Better analogy framework
Instead of asking “is 绿的谐波 the CATL/NVIDIA of robotics?”, use a five-gate test:
- Design-win gate: named top humanoid OEM customer + supply scope. 🟢 required, not yet public.
- Revenue gate: humanoid-specific revenue / backlog / order volume separate from industrial robotics. 🟢 required, not yet public.
- Architecture gate: validated BOM showing reducers / actuators are durable high-value modules rather than substituted or vertically integrated. 🟢/🟡 required.
- Margin gate: robot-related GM stays ~35% or improves despite volume scale and capacity expansion. 🟢/🟠 required.
- Platform gate: product stack expands from component to integrated actuator / joint module with switching cost, qualification lock-in, and service/data/testing ecosystem. 🟠 required.
Current grade: S4 supply-chain validation. Not S5 platform-defining supplier yet.
6. Signal vs noise across the four questions
Signal
- China 295,000 industrial robot installations in 2024, 54% of global demand, 2.027m operational stock, domestic supplier share 57%. 🟢 IFR.
- US / Western frontier capital rebound in 2024 and Figure’s US$675m Series B at US$2.6bn valuation. 🟡/🟢.
- Tesla filing-backed production-line / 1m Fremont / 10m Texas designed-capacity language. 🟢 SEC filings via existing Tesla audit.
- Figure BMW 1,250+ runtime hours / 90,000+ parts / 30,000+ vehicles and BotQ 350+ Figure 03 / 1-hour cadence metrics. 🟢 Figure official.
- 绿的谐波 revenue, profit, harmonic reducer unit sales, and robot-relevant revenue line. 🟢 filings.
Noise unless upgraded
- “China leads everything” or “US leads everything” without layer distinction. 🟠
- Tesla 1m/10m capacity as current production. 🔴
- Figure commercial agreement as scaled economics without robot count / contract value / ROI. 🟠
- 绿的谐波 as confirmed Tesla/Figure/Unitree supplier without named evidence. 🔴
- CATL/NVIDIA analogy without value-content, margin, design-win and platform-lock-in evidence. 🟠
7. Public site draft section
中美机器人：不是一场简单 horse race
过去三年，机器人产业出现了两个不同的加速器。中国的加速器是制造业部署和供应链本土化：2024 年中国安装 29.5 万台工业机器人，占全球 54%，工厂中的工业机器人存量超过 202.7 万台，国内机器人厂商在中国市场份额从 2023 年的 47% 提升到 2024 年的 57%。这说明中国拥有更大的自动化部署场景和供应链学习曲线。
美国的加速器更集中在 frontier humanoid、AI/model/compute stack 和 mega-round 资本：Tesla 把 Optimus 写进 SEC filings 并给出 Fremont 100 万/年和 Texas 1000 万/年 designed-capacity anchor；Figure 在 2024 年以 26 亿美元估值融资 6.75 亿美元，并在 BMW 和 BotQ 上给出少见的客户现场和制造 KPI。
但这两条曲线都还没有完成 S5。中国的工业机器人规模不等于 humanoid customer economics；美国的 Optimus capacity 和 Figure manufacturing KPI 也不等于行业已进入回报周期。下一步真正改变判断的证据，是 paid deployment, repeat orders, runtime/intervention rate, customer payback, production yield, robot revenue and gross margin.
绿的谐波：强供应链信号，不是已验证 CATL/NVIDIA
绿的谐波是一个好案例，因为它的证据比概念更强：2025 年收入 5.707 亿元，同比增长 47.31%；归母净利润 1.244 亿元，同比增长 121.42%；谐波减速器销量 42.5 万台，同比增长 72.48%；“工业及具身智能机器人零部件”收入 4.225 亿元，毛利率 34.88%。这是真实的 filing-backed supply-chain signal。
但它还不是 CATL 之于电车、NVIDIA 之于 AI 那种级别的已验证平台。缺口也很清楚：没有具名 humanoid OEM、没有 humanoid-only revenue、没有单机价值量、没有订单 volume、没有 humanoid-specific margin，也没有证明谐波减速器会成为所有主流 humanoid 架构中的不可替代瓶颈。更好的判断方式不是贴类比标签，而是看五道门：design win, revenue split, architecture durability, margin durability, platform expansion.
8. Think Deeper questions
- 如果中国的优势是部署规模，美国的优势是 frontier AI/compute，humanoid value 最终会在 deployment scale 侧还是 model/data 侧沉淀？
- Tesla 如果首先内部部署 Optimus，最重要的 KPI 是 robot revenue，还是 Tesla factory productivity / payback？
- Figure 的 1/hour manufacturing cadence 如果领先于 customer onboarding，真正瓶颈会从生产转向 deployment engineering 吗？
- 绿的谐波如果从 harmonic reducer 扩到 integrated actuator，价值量是上升，还是被 OEM 自研和低价竞争压缩？
- CATL/NVIDIA 类比里最容易误导的一点是什么：市场规模、单机价值量、毛利率、客户集中度，还是技术路线标准化？
9. Packaging request for Codex
Recommended slide / site package:
- Site section:
/robotics/us-china-capital-and-company-evidenceor/robotics/four-deep-questions. - Slide title: “中美机器人：两条证据曲线，不是一场简单 horse race”.
- Four blocks:
- China scale: 295k installations / 54% global / 57% domestic supplier share.
- US frontier: Tesla capacity + Figure deployment/manufacturing KPI + AI/model capital.
- Tesla/Figure: capacity curve vs deployment KPI curve.
- Leaderdrive: real S4 supplier signal, not CATL/NVIDIA proof yet.
- Footer: “Evidence map only. No winner ranking. No trade recommendation. Industrial robot scale ≠ humanoid economics; designed capacity ≠ production; supplier product fit ≠ named customer proof.”
10. Source list
Primary / official sources:
- IFR, “China Tops World Record of 2 Million Factory Robots”, 2025-09-25: https://ifr.org/downloads/press_docs/2025-09-25-IFR_press_release_China_in_English.pdf 🟢
- IFR, “World Robotics 2025 report – INDUSTRIAL ROBOTS”, 2025-09-25: https://ifr.org/ifr-press-releases/news/global-robot-demand-in-factories-doubles-over-10-years 🟢
- A3, “North American Robotics Market Holds Steady in 2024 Amid Sectoral Variability”, 2025-02-17: https://www.automate.org/market-intelligence/insights/a3-reports-north-american-robotics-market-holds-steady-in-2024-amid-sectoral-variability 🟢
- U.S.-China Economic and Security Review Commission, “Humanoid Robots”, 2024-10-10: https://www.uscc.gov/sites/default/files/2024-10/Humanoid_Robots.pdf 🟢
- Tesla Q1 2025 to Q1 2026 SEC Form 8-K Exhibit 99.1 sequence, cross-referenced in
robotics-tesla-optimus-filing-evidence-audit-v1.md🟢 - Tesla official YouTube, “Optimus - Gen 2”, published 2023-12-13: https://www.youtube.com/watch?v=cpraXaw7dyc 🟢
- Figure / PRNewswire, “Figure Raises $675M at $2.6B Valuation and Signs Collaboration Agreement with OpenAI”, 2024-02-29: https://www.prnewswire.com/news-releases/figure-raises-675m-at-2-6b-valuation-and-signs-collaboration-agreement-with-openai-302074897.html 🟢
- Figure, “BotQ: A High-Volume Manufacturing Facility for Humanoid Robots”, 2025-03-15: https://www.figure.ai/news/botq 🟢
- Figure, “F.02 Contributed to the Production of 30,000 Cars at BMW”, 2025-11-19: https://www.figure.ai/news/production-at-bmw 🟢
- Figure, “Ramping Figure 03 Production”, 2026-04-29: https://www.figure.ai/news/ramping-figure-03-production 🟢
- Figure, “Introducing Figure 03”, 2025-10-09: https://www.figure.ai/news/introducing-figure-03 🟢
- 绿的谐波 2025 annual report, 2026 Q1 report, 2025 sustainability report and 2026-06-09 abnormal-trading announcement, cross-referenced in
robotics-leaderdrive-case-study-slide-evidence-v3.md🟢
Secondary / estimate sources:
- F-Prime Capital, “State of Robotics 2025”, PitchBook / CapIQ / F-Prime analysis of North America, Europe and Israel robotics / AV companies: https://fprimecapital.com/wp-content/uploads/2025/08/F-Prime-State-of-Robotics-2025-1.pdf 🟡
- Crunchbase News, “Embodied AI Fuels Record Robotics Funding In China As IPO Momentum Builds”: https://news.crunchbase.com/robotics/embodied-ai-fuels-record-funding-china-ipo-momentum-builds/ 🟡
- Calculations in this file: China/US installation ratio 8.63x, China global share 54.4%, US global share 6.3%, Leaderdrive robot-relevant revenue share 74.0%. 🟠
11. Public-safety flag
Public-safe: yes.
Exclude:
- Hugo private portfolio weights, purchase prices, trade rationale, or watchlist priority.
- Any buy / sell / hold language for TSLA, 688017.SH, private humanoid companies, or suppliers.
- Paid-report excerpts.
- Unverified supplier-customer rumors.
- Claims that Tesla 1m/10m designed capacity is current output.
- Claims that Figure BMW / BotQ metrics prove full economics.
- Claims that 绿的谐波 is confirmed supplier to Tesla, Figure, Unitree, AgiBot, UBTECH, or any named OEM unless primary evidence appears.