Intentional News

261014 China's Industries of the Future — Named as Priorities, Sold as Targets (CNA)

19 min read

China’s Plan for the Industries of the Future — Named as Priorities, Sold as Targets

The thesis (Alf’s): This film’s factual spine is better than its framing. Every one of the six industries it names is genuinely in China’s 15th Five-Year Plan, in the plan’s own words, and almost every dated achievement it reports survives checking. What does not survive is the delivery device: the cold open converts a list of priorities into “specific targets for each one of these areas to be met by the year 2030” — and the plan contains no such per-industry targets. That single substitution is the whole argument of the film, and it is the one thing the document does not authorise. Read correctly, the film’s own evidence says something narrower and more interesting: China is not first to invent these things, it is first to industrialise them — and its own regulators are now warning that part of that industrialisation is a subsidy channel dressed as demand.

The two questions, answered up front

Q1 — is it just China? No, and the film says so in its own words. Its own interviewees concede the US leads on fusion’s central milestone and splits quantum computing; its own script notes that the most famous BCI company in the world is Musk’s. What China has is not a science lead across the board — it is a lead in regulatory speed, manufacturing volume and state coordination, which is a different and in some ways harder advantage to answer.

Q2 — how will these technologies be “democratised”? The word hides three different mechanisms, and they behave nothing alike. The full argument is in ## How technology actually spreads below. In one line: where the technology is a file it has already spread irreversibly; where it is a cheap manufactured good it spreads by price and can only be tariffed; and where it is a machine only one company on earth can build, it does not spread at all. Two of those three layers do not care what governments agree. The third is the one that matters most for military and AI capability, and it is narrow.

Summary of the film

Cold open: a montage of humanoid robots, brain augmentation, thought-controlled machines, plants that glow in the dark, self-flying cars, 6G and fusion — “Every single one of these technologies is listed in China’s latest 5-year plan with specific targets for each one of these areas to be met by the year 2030.” Then a track record (EVs, solar at 80%+ of world panels, drones), a mechanism thesis (a guest’s I5 model — ideas, institutions, instruments, infrastructure, impact — i.e. whole-of-society mobilisation), then one segment per industry: BCI (BrainCo prosthetics; the NMPA’s March 2026 approval), bio-manufacturing (mega facilities including HEPS; bioprinted organoids; plant-derived plastics), quantum (Jiuzhang 4.0; Hanyuan-2; a Hong Kong post-quantum banking paper), fusion (EAST; the “race to watch”), embodied AI (humanoids; data centres; 6G), and a closing argument about science as the route out of the middle-income trap, with the counter-currents — a slowing economy, restrictive Party control over collaboration and travel — stated in its final minutes.

What the plan actually is, and why the distinction matters

  • Current plan: the 15th Five-Year Plan, covering 2026–2030. Its Recommendations were adopted by the Fourth Plenum of the 20th CPC Central Committee, 20–23 October 2025; the binding Outline was adopted by the National People’s Congress on 12 March 2026.
  • The film never names the plan, its number, its adopting body or its date — and never distinguishes it from the 14th plan (2021–2025) whose results it spends its opening celebrating. That is the largest structural gap in the film: a viewer cannot tell which plan is which, or that the celebrated track record belongs to the previous one.
  • Four different things get called a “target” in this film, and they are not interchangeable: an indicative target (a number), a directive (names a sector, no number), a budget line (money and machinery), and an achieved result (a dated event). The film announces targets, then illustrates with achieved results and programmes.

The six industries — the plan’s own words

The six are real and the film gets the list right. From the Recommendations, chapter 3:

“We need to explore diverse technology roadmaps, typical application scenarios, feasible business models, and market regulation rules and work to foster new drivers of economic growth such as quantum technology, biomanufacturing, hydrogen and nuclear fusion power, brain-computer interfaces, embodied artificial intelligence (AI), and 6G mobile communications.”

That is a directive list with no numbers attached. (Watch for a second, easily-confused set of six: MIIT’s emerging pillar industries — integrated circuits, aerospace, biomedicine, low-altitude economy, new energy storage, intelligent robotics. The film’s “low-altitude industries joined the fold” refers to that list, and its cold open then promotes flying cars into the headline six it does not belong to.)

Is it just China?

Verdict per technology, as at October 2026. “Behind” means behind China; “Level” means genuinely comparable.

TechnologyWho else, and where they actually areVerdict
Humanoid robotsFigure, Tesla Optimus, Agility, Boston Dynamics (US); no non-Chinese maker ships in the hundreds. China: >95% of H1 2026 global shipmentsChina leads on volume — decisively. But see the demand caveat below
Brain-computer interfacesNeuralink (intracortical, more advanced implant), Synchron (endovascular, US trials), Precision Neuroscience, Blackrock NeurotechLevel-to-behind on the implant; ahead on nothing but approval speed. China’s “world’s first” is a market authorisation, not a device lead
FusionUS: NIF achieved the only net-gain ignition on record. ITER assembling. Commonwealth Fusion ~$4bn. Helion first private D-T burn. UK STEP; German hubsNon-Chinese lead on the milestone; China leads on long-pulse confinement and on state-scale capital
Quantum computingIBM ($10bn; Starling targeted 2029), Google (a verifiable advantage result), Quantinuum (50 logical qubits), PsiQuantumUS ahead on computing — the film’s own expert says so. China ahead on quantum communications, by a wide margin
Biomanufacturing / bioluminescent plantsLight Bio has sold the world’s first bioluminescent plant, with US clearance, since April 2024 at $99.99US at commercial; China at demonstration. But China holds >70% of global fermentation capacity
eVTOL / “flying cars”Joby, Archer (US) — Joby is still in the fifth and final stage of FAA type certification as of August 2026China ahead on certification — EHang’s EH216-S holds the world’s first eVTOL type certificate and remains the only certified pilotless passenger eVTOL. Most applications remain demonstrations
6GEU Hexa-X-II (280 standard contributions); Korea, JapanNobody has a commercial 6G network anywhere. China’s ~40.3% patent share is the largest measurable lead — inputs, not a network
Mega science facilitiesEurope’s flagship neutron source ESS has still not produced first science as of Oct 2026; Hyper-Kamiokande (Japan) is larger than JUNO by volumeChina genuinely ahead on synchrotron user operation — HEPS began user experiments 3 Dec 2025 and ran 222 experiments for 91 institutions by mid-Feb 2026

Where China genuinely leads on the evidence: humanoid manufacturing volume; quantum communications; fermentation-based biomanufacturing capacity; eVTOL certification; fourth-generation synchrotron in user operation; long-pulse tokamak duration (EAST’s 1,066-second H-mode record); and state-scale coordination of the whole portfolio — the plan triggers mandatory coordination across all ministries, provinces and state financial institutions, and in 2026 China consolidated fusion R&D into a single ~$2bn vehicle. No other country organises capital that way.

How technology actually spreads

The question “how will these technologies be democratised around the world?” cannot be answered in one piece, because three different layers are usually being conflated.

LayerExampleWhat governs itCan policy stop it?
It’s a file — weights, specifications, designsDeepSeek’s MIT-licensed weights; RISC-V at the Chinese Academy of SciencesNothing. A file has no border. Diffusion here is deliberate strategyNo. No control regime reaches it
It’s a cheap manufactured goodBatteries, solar, robots, storage, EVsPrice. Chinese LFP cells near $36–50/kWh; a 44–56% cost advantage over Western production, driven by overcapacityNo. It can only be tariffed
It’s a machine only one company can buildEUV lithography — ASML is the sole supplier with no second source anywhere, and behind it leading-edge foundry (TSMC ~73% of foundry revenue)The tooling chokepointYes — this one genuinely holds. And it is unlike every historical analogue

So: two-thirds of the technology in this film is already on a diffusion path, and the third that is not is the third that matters most. That is a moderately encouraging conclusion for a multilateralist — but reached for unromantic reasons: the technology spreads because it is cheap and because it is written down, not because anyone agreed to share it.

The chokepoints, tested

ChokepointReal?The detail
EUV lithographyYes — the one genuine single-point chokepointASML is the sole supplier, no second source anywhere; ~$250m a machine; restrained only by a Dutch national export licence on top of the EU dual-use regulation
Leading-edge foundry (TSMC)Yes~73% of worldwide foundry revenue in Q2 2026 against SMIC’s ~5.4%. Defended by physics and capital, not by treaty
Advanced packaging and test toolingYes, and it tightened in 2026Japan’s METI third round in force 1 Aug 2026 (~20 categories incl. wafer thinning, laser stealth dicing)
Chinese rare earths and magnetsYes, and extraterritorialMOFCOM Notice 2025 No. 61 (9 Oct 2025): a 0.1% value de minimis plus a Chinese foreign-direct-product rule reaching foreign-made items — suspended for one year from Nov 2025, expiring 27 Nov 2026
US controls on advanced chipsYes, but the architecture is swingingEntity List + the Sept 2025 Affiliates (“50%”) Rule — suspended within six weeks of taking effect
EDA softwareOverstatedImposed by letter 23 May 2025, lifted 3 July 2025 — five weeks
The gallium/germanium banOverstatedRe-routed via third countries; suspended to 27 Nov 2026 but still licensed

The historical base rate — what the analogues establish

  1. Time. Every case took 11 to 40 years from origin to transferred manufacturing leadership. Nothing in the film’s six industries is on a five-year clock.
  2. Mechanism. Every case ran on a price collapse (solar, batteries, EVs), a mandated transfer backed by market access (China’s high-speed rail, 2004, from Alstom, Siemens, Bombardier and Kawasaki), or both. None ran on good relations.
  3. Value. In four of six cases the originators captured little or none of the eventual value. The exception is semiconductors, where design IP (US) and leading-edge foundry (Taiwan) still capture most of it — which is exactly why that is the chokepoint held hardest and longest.
  4. Reversibility. Nobody in the table has un-diffused a technology they lost; conversely nobody has un-lost a chokepoint without a substitute supplier appearing.

What the film leaves out — and it is the state’s own criticism

This is the strongest evidence in the whole file against the film’s framing, because it does not come from critics. It comes from Beijing.

  • November 2025 — the planning ministry warns of a bubble. The NDRC flagged “the proliferation of remarkably similar robots from more than 150 companies in the same field.” Ten months before the film aired. The film’s robot segment contains none of it.
  • September 2026 — the securities regulator slows humanoid IPOs, scrutinising whether “soaring valuations and revenue tied to state-backed projects reflect commercial demand” — with local governments providing 80–90% of initial investment in some joint ventures. One investor’s phrase for it: “campaign-style innovation”; and an estimate that some robot companies would fall 60–70% if revenue tied to robot data-collection centres were stripped out. Morgan Stanley is reported to say 50–70% of China’s humanoid output may go to government-backed data factories rather than productive work (reported-not-confirmed).
  • October 2026 — Beijing’s own corrective. A package of 19 measures warned against “speculative investment… and local governments copying the same development plans”, and threatened accountability for losses from a “rush into fashionable industries”. “Plan-target inflation” acknowledged in the plan’s own words.
  • The track record’s missing cost side. China produced 574.5 GW of the world’s 693.6 GW of solar modules in 2025 — and every major Chinese solar maker was loss-making that year, running at about 53% capacity utilisation. The film presents this as unqualified success. Dominance and excess are the same fact, and it is why Beijing is now running an anti-”involution” capacity campaign.

Fact-check — the fastest-moving numbers

Claim in the filmThe checked positionVerdict
”Specific targets for each industry to be met by 2030”The plan names the six as priorities and orders “forward-looking plans” and “diverse technology roadmaps”. No per-industry 2030 number found in any primary textNOT IN THE DOCUMENT — the film’s construction
40,000 humanoids shipped in H1 2026 (attributed to Morgan Stanley)No tracker supports it: IDC ~25,000; Counterpoint ~22,000; Smart Analytics ~19,100CONTRADICTED
”90% of the 13,000–16,000 shipped in 2025”China’s share is broadly corroborated (>95% of H1 2026 units by some counts) — but it sits oddly beside the film’s own higher H1 numberPartly verified, internally inconsistent
BrainCo: 10,000 people fitted with bionic limbs that “decode brain waves”BrainCo’s own documentation: “2-channel electromyography” and “requires no surgical interventions or implants”FALSE — the film’s clearest error
NMPA approved the world’s first invasive BCI, March 2026Verified — 13 March 2026, Implantable BCI for Hand Motor Augmentation, developed by Shanghai-based Neuracle Medical Technology and Tsinghua University. But the device is extradural; the film says electrodes “inserted directly into the brain”Verified, with a factual slip
”Ju Chang 4.0”, May 2026, 25 microseconds, in NatureIt is Jiuzhang 4.0; work first on arXiv Aug 2025, in Nature 13 May 2026. Quantum sampling is not useful computationVerified with name corrected; the significance is overstated
”Hanuen 2” as a room-temperature breakthroughHanyuan-2, neutral-atom, 8 May 2026 — and neutral-atom machines are room-temperature by architecture (Pasqal, QuEra, Atom Computing)Not a Chinese first
”HEPS just started work in September 2026”Began trial operation and user experiments 3 December 2025; 222 experiments for 91 institutions by mid-Feb 2026The film understates China here
NIF as “net gain fusion on the grid”NIF is single-shot, with wall-plug gain below 1%, and produced no electricityOverstated
JUNO as “the world’s largest underground facility”Largest liquid-scintillator detector of its kind; Hyper-Kamiokande (Japan, 260kt) is larger by volumeOverstated
”Around 30% of the global pipeline for innovative drugs”No source locatedUnverified — should not be printed
”10,000 foreign-trained Chinese scientists have returned home”The number is loosely handled: a CNN tally counts 85 senior US-based scientists moving full-time since 2024. Departures of US-based scientists of Chinese descent rose ~75% after the China InitiativeThe figure does not say what the film implies
Bioprinted human heart organoids, “Singha University”, Nov 2025Unresolved — the institution name is a probable mis-transcription; no source found for the announcement, date or organ listUnverified

Where the film is strongest: the six-industry list is exactly the plan’s; its dated achievements mostly check out (NMPA’s approval, Jiuzhang 4.0, Hanyuan-2, EAST’s density-limit result in Science Advances, the $295bn five-year data-centre report); its whole-of-society mobilisation thesis is textually supported by the plan; and it hedges honestly where hedging is due — “maybe not quite yet” on robots, “anyone’s guess” on fusion.

Where it is thinnest: it never names the plan; it converts a directive into a target; it omits China’s own regulators’ warnings about the bubble it is celebrating; and in the AGI segment it lets a humanoid robot testify — “Tiangong” is interviewed as though it were an expert — which is the moment the film stops reporting and starts performing.

My assessment (Alf)

The film is better evidence than argument, and the giveaway is that its own experts keep contradicting its framing. Nobody on screen says China invented these technologies. They say China industrialises them faster — the fusion expert says the US invented the field and holds the only net-gain result, and then names supply chain and scale-up speed as China’s advantage. The quantum expert splits the verdict. The script itself notes the most famous BCI company in the world is American. A viewer who only watches the cold open and the graphics comes away with the opposite impression from a viewer who listens to the words, and I think that gap is deliberate. It is the most common failure mode in technology coverage and this film is a clean specimen of it.

The real story is more interesting than “China is winning”. What China has built is not a science machine, it is an industrialisation machine — and the film’s own evidence shows both halves of that. The half it shows is real: 95%+ of humanoid shipments, 70%+ of fermentation capacity, the only certified pilotless passenger eVTOL, a synchrotron running 222 user experiments while Europe’s flagship neutron source still has not produced first science. The half it hides is that industrialising by state mobilisation produces volume and demand-shaped-by-subsidy at the same time — every major solar maker loss-making at 53% utilisation, local governments funding 80–90% of robot joint ventures, and a planning ministry warning about 150 near-identical robot companies ten months before the film aired.

On Q1 — “is it just China?” — I’d give the honest answer the film avoids, and it is not the comfortable one. China is not first across the board, and in two of the six it is clearly behind on the thing that matters (fusion’s central milestone; the implant itself). But it is first at the stage where nations usually lose: turning a demonstration into an installed fleet, a certification, a factory. That is a real, hard advantage and it is harder to answer than a research lead, because a research lead can be funded away and a manufacturing base has to be built. My judgement: the US and Europe mostly lead on the frontier and mostly trail on deployment, and most public argument confuses the two because the frontier is the more flattering thing to measure.

On Q2 — “democratised” — I would push back on the word itself, and this is the substantive disagreement I’d put to Hiro. “Democratised” implies a decision made on behalf of everyone, by someone with the authority to make it. The evidence says almost none of this happens by decision. Two of the three layers spread regardless: files have no border, and cheap goods go where the price is. The layer where a decision genuinely exists — the EUV-and-foundry chokepoint — is one company, one licence, one allied bloc, which is the precise opposite of democratic; it is a monopoly held by an accident of engineering, and no government has ever held a tooling chokepoint for a decade. So the honest version of the question is not how will these be democratised but “which layer do we mean, and who actually holds the lever on that layer?” — and the uncomfortable part for anyone with multilateral instincts, myself included, is that the spreading is not a product of cooperation; it is a product of price and open files, and the only genuinely decision-shaped part of it is the part that is least democratic.

One thing I’d hold to firmly: the film’s central correction is not a correction about China’s ambition, which is real, but about whether a plan’s ambition is evidence of its delivery. A directive list read as a target schedule is the same error as reading a company’s strategy deck as its accounts. And the fact that Beijing’s own NDRC, CSRC and State Council were warning about exactly that inflation — in November 2025, September 2026 and October 2026 — is the best evidence in this whole file that the distinction is real and matters.

Open questions

  • Does the film’s 40,000-unit H1 2026 figure trace to anything? Three trackers disagree with it and I could not find its source. Worth a second pass.
  • “Singha University” and the bioprinted heart organoids — the only unresolved caption garble; if the claim is real, which institution?
  • Is “democratisation” the right frame at all, or should the vault treat technology diffusion as a price-and-files phenomenon with one narrow policy-shaped exception? This is the question the Envisioning seed now carries.
  • What is the honest test of China’s six industries in ten years? Not unit counts — the plan’s own critics suggest revenue quality (commercial vs state-backed) is the one that discriminates.
  • Does the Chinese immersion-DUV question settle the whole argument? UBS’s window is 2–5 years (2027–2030). If it arrives, the tooling chokepoint is a decade-long anomaly; if it does not, Thesis B holds.
  • Two dates to watch: 27 November 2026 — the expiry of China’s rare-earth control suspension, and the US Affiliates Rule’s deferred date.

Sources

Everything below was checked in this pass. Full evidence files with STATUS tags, source URLs and verbatim quotes: /opt/data/profiles/alf-news/cache/scratch/china_future/01_outside_china.md (69 KB), 02_diffusion.md (136 KB), 03_plan_and_claims.md (72 KB), and 00_caption_corrections.md for the transcript garbles.

The plan itself

Beijing’s own criticism — the load-bearing section

Verified achievements

Diffusion and chokepoints

  • ASML as sole EUV supplier, no second source (Reuters, 17 Dec 2025) — https://reuters.com/technology/asml-ceo-says-no-second-source-euv
  • MOFCOM Notice 2025 No. 61 — rare-earth extraterritorial controls (9 Oct 2025), expiry 27 Nov 2026
  • Japan METI third round of export controls, in force 1 Aug 2026
  • DeepSeek MIT-licensed open weights — the deliberate diffusion strategy
  • US v. Hsu — “Operation Gatekeeper”, ≥$160m of H100/H200 GPUs

Graph View

Backlinks