Category: Future

Attention isn’t just personal grit; it’s a public good. Treat short‑video feeds like a safety‑critical system: labels, defaults, audits and data access across jurisdictions.

Modern economies leak attention like a faulty pipe. Average on‑screen focus now resets in ~47 seconds; daily reading‑for‑pleasure has fallen by ~40% over two decades. Meanwhile, nearly half of US teens say they are online almost constantly. None of this proves simple causation; it does establish a base rate that warrants action. In public‑health terms the markers are present: high prevalence, meaningful severity (especially for adolescents), clear externalities (lost learning, error‑prone work), modifiable risks (autoplay, infinite scroll, push alerts) and feasible interventions.

The market failure

Short‑video platforms optimise for watch‑time using designs that mimic variable‑ratio reinforcement [unpredictable rewards; the slot‑machine schedule]. Infinite scroll, autoplay and push alerts externalise interruption costs onto households, classrooms and workplaces. Crucially, when researchers removed only mobile internet for two weeks—texts and desktop intact—participants’ sustained attention and well‑being improved materially. That is a policy‑scale hint: change defaults, change outcomes.

The risk is dynamic, not static. A new real‑user study of 1,100 TikTok accounts found that even light users roughly doubled their daily watch time over six months. The curve is endogenous; usage begets usage. Regulation must therefore target system behaviour, not individual virtue.

From self‑help to systems: the attention diet as regulation

Think of an “attention diet” not as self‑denial but as market design: labels, budgets and defaults at the product layer, plus audits and data access at the system layer.

1. Labels (calories for cognition).
   • Warning labels for minors, as urged by the US Surgeon General, frame the baseline risk. But add attention‑nutrition panels—notifications/day; autoplay status; average time/session—visible in app stores and settings. Like food labels, they don’t ban; they disclose.
2. Budgets (age‑appropriate ceilings).
   • For under‑16s, autoplay off and finite scroll by default; time‑of‑day curfews that respect sleep and school hours; counters that reset only after offline intervals. Australia’s new minimum‑age law for social media provides the legal handle; make the defaults align.
3. Defaults (friction beats willpower).
   • Notification batching (e.g., three drops/day) reduces stress without FoMO spikes. Creator‑first defaults (draft before feed) and For‑Later queues convert roulette into reading lists. These are small levers with measured effects.
4. Audits and access (accountability, not vibes).
   • The EU’s DSA (EU’s online‑platforms law) already mandates systemic‑risk assessments for VLOPs (very large online platforms)—including risks to public health and minors’ well‑being—and opens Article 40 data access for vetted researchers. Export that architecture: audited risk logs, independent experiments, and public dashboards.
5. Evidence at scale (policy RCTs).
   • Run city‑ or district‑wide trials that toggle autoplay, scroll limits, or labels with pre‑registered outcomes (sleep, anxiety screens, reading minutes, exam scores). The two‑week mobile‑internet RCT shows behavioural gains are available; governments should fund the follow‑through.

A cross‑national playbook

United States. Pair warning labels and public‑health messaging with a federal baseline for age‑appropriate defaults. Re‑table KOSA (US minors online‑safety bill) with precise design mandates (autoplay off, robust age checks, independent audits) and safe‑harbour protections for privacy‑preserving research access. Use federal procurement to require attention‑friendly defaults in official apps.

European Union. Enforce DSA risk assessments and mitigations with a named “attention harm” class and require VLOPs to offer a chronological feed and to publish shrink tests (impact of disabling engagement features on minors’ outcomes). Use the new researcher‑access delegated act to open multi‑country panels.

United Kingdom. Under the Online Safety Act, Ofcom’s draft children’s codes already push algorithmic taming and age checks. Add attention metrics to compliance (notifications/day, autoplay status) and align school‑day phone rules with clear guidance on storage, exceptions and enforcement.

Australia. With an under‑16s minimum‑age regime taking effect by December 2025, set pragmatic definitions of “reasonable steps” (high‑accuracy age assurance; family‑device attestation) and require public accuracy audits of age‑checks to avoid false positives.

Schools as the high‑yield lever

School‑day smartphone prohibitions are becoming the norm (England’s guidance; France’s tightening practice). The best evidence suggests test‑score gains concentrate among lower achievers when phone bans are enforced; but bans alone don’t fix sleep or anxiety. Marry phone‑free days with print‑reading minutes and digital‑hygiene lessons.

Productivity is a public‑health externality

Interrupted workers “go faster” to catch up, raising stress and error risk. The mere presence of a smartphone can sap cognitive capacity—replication is mixed, but the prudent policy is out of sight in safety‑critical zones and batched alerts elsewhere. Governments can lead by setting attention standards for the public sector and contractors.

How to measure success

• Exposure: minutes of autoplay exposure per minor/day; notifications/day; share of time in chronological feeds.
• Outcomes: reading‑for‑pleasure minutes (time‑use diaries), sleep duration, YRBS (CDC teen health survey) indicators, and school exam variance.
• Process: share of VLOPs with published risk logs, independent audits, and researcher APIs.

Objections—and replies

• “This is parental, not public, responsibility.” Seatbelts and food labels did not replace personal responsibility; they raised the baseline. So will safer defaults.
• “It will harm innovation.” Transparency, interoperable controls and audits are pro‑innovation; they make quality legible and reduce arms‑race externalities.
• “Evidence is mixed.” True for causal links to all harms, less so for design exposures (autoplay, infinite scroll) and usage intensity. Base‑rate trends and RCT signals justify low‑regret guardrails.

The kicker

Treat attention like clean air for cognition: invisible, essential, taken for granted until it degrades. Set labels, budgets and defaults; demand audits and data. The feeds won’t police themselves.

By the numbers

• ~47 seconds: average on‑screen focus before switching. (gloriamark.com)
• −40%: decline in US reading‑for‑pleasure, 2003–2023. (Cell)
• ~50%: US teens online “almost constantly.” (Pew Research Center)
• RCT: 2 weeks without mobile internet → better attention & well‑being. (OUP Academic)
• Escalation: even light TikTok users double daily watch time in 6 months. (The Washington Post)
• Policy shift: Australia’s under‑16 minimum‑age law effective Dec 2025. (Infrastructure and Transport Dept)

Sources (selected)

1. US Surgeon General advisory on social media & youth mental health. (HHS.gov)
2. PNAS Nexus (2025) RCT: blocking mobile internet improves attention and well‑being. (OUP Academic)
3. iScience (2025): 20‑year decline in US reading for pleasure (ATUS). (Cell)
4. Pew Research (2025) Teens & social media fact sheet—“almost constantly” online. (Pew Research Center)
5. EU DSA researcher‑access (Article 40) explainer. (algorithmic-transparency.ec.europa.eu)

How low‑effort AI and high‑speed feeds could break research, politics—and our sense of what’s real

When models train on their own exhaust and voters doomscroll synthetic video, truth becomes expensive. We can still cheapen it—if platforms and policymakers act.

1) The supply shock: AI slop meets recursive training

The internet is filling with AI slop [low‑effort, low‑quality AI content]. That isn’t just an aesthetic problem. Generative models increasingly ingest the very outputs they produce. Theory and experiments show that when models feed on model‑made data, they “forget” the tails of human creativity—a dynamic researchers dub model collapse. Left unchecked, each training round narrows what systems can discover, worsening homogenisation in search, summarisation and even science.  

Add adversaries and the picture darkens. It is cheap to poison web‑scale datasets in ways curators will not spot—$60 could have tainted popular image corpora—making future models brittle or biased in targeted ways.  

Zoom out and a familiar macro risk appears: algorithmic monoculture. When many decision‑makers converge on the same systems or ranking functions, average accuracy can fall and systemic fragility rises—even before shocks arrive. In information markets, a monoculture of recommendation and re‑generation magnifies slop, then feeds on it.  

By the numbers

• 40%: global trust in news—stuck at a low plateau.  

• 33% of people use TikTok; 17% use it for news—usage growing fastest among the young.  

• ~55%: pooled human accuracy at spotting deepfakes—barely above chance.  

• −46% reposts, −44% likes after an X note is attached to a misleading post.  

• $186 per worker/month: cost of AI “workslop” in US offices.  

2) The demand shock: short video compresses attention

Truth struggles not only because supply degrades; demand is being rewired by design. Short‑video feeds reward novelty, moral emotion and speed—exactly the traits that boost diffusion of polarised content and negative headlines. That is measurable: each extra “moral‑emotional” word raises the odds a political tweet spreads, and each extra negative word in a headline lifts click‑through by about 2.3%.  

Collective attention itself is shortening across media, as more content competes for fixed cognitive bandwidth. The churn is faster; dwell time shrinks.  Meanwhile, real usage data show TikTok can train daily habits quickly; even light users escalate to well over an hour a day, pushing decision‑relevant information into ever tighter frames.  

Why this matters for research and decisions: when evidence must fight for milliseconds, methods devolve toward vibes. In such markets, slop that is emotive, familiar and frictionless beats slow, careful work.

3) Blurring reality: Sora, deepfakes and the liar’s dividend

As synthetic video matures, the boundary between reportage and make‑believe erodes. OpenAI’s Sora 2 launch is a leap in realism (now with synchronised audio) and a distribution shift: a consumer app. Sensibly, the firm ships visible watermarks and C2PA [an open provenance standard] metadata by default—important, because provenance will become the “receipt” of authenticity even when pixels look perfect.  

Alas, humans are mediocre deepfake detectors—near coin‑flip in aggregate—so realistic fabrications will often pass casual scrutiny. Worse, rising deepfake awareness turbocharges the liar’s dividend: the ability to dismiss inconvenient truths as “AI”. Together, these forces make ordinary users less able and less willing to sort real from fake.  

Platforms are edging toward better labelling. YouTube now requires disclosure of “realistic” synthetic content; TikTok has begun labelling AI media from other tools using C2PA signals. Good—but labels must travel with downloads and edits, not just live on one platform.  

4) Politics in the feed: propaganda, micro‑targeting and platform responses

State‑aligned actors already flood platforms with low‑quality influence content (China‑linked “DRAGONBRIDGE” is the canonical case), and while most such spam draws little organic engagement, the flood still clogs discovery and fuels confusion.  

Regulators are moving. The EU’s DSA [content law] is probing TikTok’s ad transparency; separately the bloc’s new TTPA [rules on transparency/targeting of political ads] has prompted Meta—and now Google—to halt political ads in the EU, rather than re‑tool for compliance. That choice reduces one channel for micro‑targeted propaganda but pushes more contestation into organic feeds, creator “newsfluencers” and encrypted groups where transparency is weaker.  

Can crowd‑fact‑checking help? Evidence is mixed but encouraging on post‑level impacts. Rigorous work finds X’s Community Notes (formerly Birdwatch) can materially reduce diffusion—roughly halving re‑shares and likes—once a note appears. But notes often arrive late; and labelling a subset of false posts risks the implied truth effect, where unlabelled falsehoods seem more credible. Design details—speed, coverage, note quality, and when labels appear—decide outcomes.  

5) When factuality becomes a luxury

If slop is cheap and attention is scarce, verification becomes a luxury good. Knowledge workers already pay a hidden tax for “workslop”—plausible‑looking but wrong memos, slides and emails that take hours to unwind. That same tax hits citizens sifting political clips and “AI‑news” in their off‑hours. Over time, the people and institutions willing (or able) to pay for provenance, corroboration and curated context will diverge from those who cannot. That split—more than any one election—would be the real democratic recession.  

6) Making truth cheap again: a policy‑and‑product playbook

A. Provenance by default. Mandate tamper‑evident credentials for realistic synthetic media across major platforms and devices; make labels travel with files and show prominently in embeds. The EU AI Act’s transparency duties and YouTube’s disclosure rules are a start; extend them to default, cross‑platform C2PA.  

B. Slow the first hop. Introduce “friction at virality”: when content spreads unusually fast (especially around politics or health), platforms should slow the boost until either (i) provenance is verified, or (ii) high‑quality notes appear. Evidence suggests crowd context can cut virality sharply if delivered before the cascade peaks.  

C. Evidence trays. For any item labelled synthetic or contested, show an expandable tray with original‑source candidates (archives, wire copies), model provenance, and fact‑checks. This offsets the implied truth effect by offering alternatives, not just warnings.  

D. Anti‑monoculture incentives. Search, social and LLM providers should publish overlap scores and diversify training and ranking inputs—akin to financial concentration limits—to avoid the slop loop. The welfare gains of diversity are real under monoculture models.  

E. Research hygiene. Journals, funders and firms should adopt “RAG‑first” practices (RAG = retrieval‑augmented generation [LLM linked to sources]) for summaries and drafts, with compulsory bibliographies and data deposits; training pipelines must include human‑origin “clean rooms” to resist recursive collapse.  

F. Ad transparency that survives platform exits. As major platforms retreat from EU political ads, regulators should require public, standardised reporting of political spend and targeting across all channels (search, influencer, CTV, creators), or risk the debate shifting to darker venues.  

7) The kicker

We do not need to choose between creativity and credibility. But we do need to price provenance and pause back into the feed. If platforms can make the wrong clip spread in seconds, they can make the right context spread just as fast.

Sources (selected)

• Shumailov et al. (2023), The Curse of Recursion—model collapse.  
• Lorenz‑Spreen et al. (2019), Accelerating dynamics of collective attention.  
• Reuters Institute (2025), Digital News Report—trust and TikTok use.  
• OpenAI (2025), Launching Sora responsibly—watermarks & C2PA.  
• Slaughter et al. (2025), PNAS—Community Notes reduce virality.  

A conservative playbook to fix the rhetoric and the rules

I. What went wrong—and what to keep

Keep the security spine. Two pillars are sound: targeted export controls on advanced compute and tools, and the new outbound‑investment screen—both aimed at genuine chokepoints. The Treasury rule took effect January 2, 2025; do not broaden it into a catch‑all capital wall. Keep it narrow, with clear licences and sunsets.  

Fix the overreach. Across‑the‑board tariff hikes (e.g., Section 301 increases on EVs to 100%) punish allies, raise input costs, and invite retaliation—without improving security. Convert broad tariffs into evidence‑based trade remedies and trusted‑lane enforcement.  

Stop equating “manufacturing” with “jobs”. Modern fabs are capital‑ and automation‑heavy. Compete for them by de‑risking siting (firm power, fast permits), not by out‑bidding allies project by project. The US already tightened NEPA timelines and page caps and overhauled grid interconnection; finish the job and make those reforms the “incentive”.  

II. Conservative principles for a durable reset

1. Security, not autarky. Protect chokepoints (advanced chips, tools, secure compute) with allied rules and licences; sunset and review regularly.  
2. Allies are force multipliers. Make “trusted trade is faster trade”: expand AEO/CTPAT mutual recognition so low‑risk cargo clears first; target enforcement at unknown shippers and de‑minimis gaming.  
3. Capacity over cheques. Out‑compete by guaranteeing megawatts and minutes (firm, clean power; predictable permits), not by writing bigger subsidies. FERC Orders 2023, 1920 and 1977 are the lever—use them.  
4. Taxes that reward investment, not lobbying. Restore neutral cost‑recovery (R&D), keep the base broad, rates competitive, and avoid bespoke carve‑outs. Current law’s Section 174 amortisation is a competitiveness drag—fix it cleanly.  
5. Law‑and‑order immigration that favors talent. Tight borders plus skills‑based inflows and a real startup parole/visa channel—simple, lawful, pro‑American dynamism.  

III. The policy deck (what to announce and do in 100 days)

A. Trade & security: enforce, don’t escalate

• 301‑to‑remedies swap. Where 301 hikes now apply, instruct USTR/ITC to prioritise countervailing/anti‑dumping cases with transparent public records; publish injury evidence and allow price undertakings. Reserve 301 for proven market‑wide coercion.  
• Trusted‑lane expansion. Direct CBP to expand AEO/CTPAT MRAs and publish an AEO “fast‑lane” dashboard (clearance times, inspection rates). Require known‑shipper IDs and full data on de‑minimis parcels from high‑risk lanes.  
• Outbound clarity. Keep the Outbound Investment rule narrow: prohibited where national‑security risk is direct (sub‑7 nm logic, AI training chips), notified elsewhere, with passive investment safe harbours. Annual report with metrics and licence timelines.  

B. Industrial policy: minutes and megawatts beat money

• Permit clock discipline. Implement CEQ NEPA Phase 2 time/page limits in agency SOPs and publish a single permit clock dashboard (median/90th‑percentile). Deem applications complete unless agencies flag deficiencies within 30 days.  
• Grid fast‑track. Use FERC Order 2023 cluster studies to guarantee energisation dates for strategic loads; roll state compacts into Order 1920 long‑term planning and deploy Order 1977 backstop where corridors stall.  
• “No MW, no money.” Condition any discretionary subsidy on a signed PPA/CfD (clean, firm power) and a dated interconnection slot. Publish the contract summary.

C. Tax & regulation: pro‑investment, rules‑based

• Fix cost recovery. Replace Section 174 amortisation with full expensing for domestic R&D, paired with anti‑abuse guardrails; keep the R&D credit; simplify accounting method changes per IRS guidance. (No bespoke carve‑outs.)  
• Sunset‑proof credits. Time‑limit any new investment credits; require outcome metrics (MW connected, units shipped, on‑time delivery) rather than headcounts.

D. Immigration: tight border, open front door for skills

• Use what exists—now. Scale the revamped H‑1B beneficiary‑centric selection (cuts gaming), accelerate STEM green‑card backlogs with visa recapture, and give teeth to International Entrepreneur Parole while Congress considers a statutory startup visa.  
• Wage‑indexed caps. Announce a proposal to tie high‑skill visa numbers to national unemployment and wage growth bands—automatic stabilisers with public triggers. (No discretion, no backlog games.)
• Enforcement first. Expand E‑Verify and surge adjudication at the border and in asylum courts; pair with lawful, skills‑oriented channels so the “front door” is faster than the fence.

E. Allies & climate: reciprocity, not copy‑paste

• Allied recognition. Offer “Buy Allied” treat­ment for NATO/GPA partners that extend reciprocal market access—better than blanket Buy American fights.
• Carbon reciprocity. Prefer product‑standard reciprocity to blanket climate tariffs; if allies run CBAM (EU charges from 2026), negotiate credits for US producers’ lower embedded emissions rather than staging a tariff exchange.  

IV. The rhetoric reset (what to say)

• From “reshoring or bust” → “secure, not sever.” We will protect what matters—chips, tools, secure compute—while keeping trusted trade fast and cheap.
• From “jobs from cheques” → “speed is the incentive.” Firm power, fast permits, clear rules: investors come for certainty.
• From “tariffs as a creed” → “enforcement by evidence.” If there’s a subsidy or dumping, we’ll prove it and act—case by case.
• From “closed borders” → “law‑and‑order talent.” Tight border security and the world’s best front door for skills and founders.

V. Scoreboard (publish quarterly, not at re‑election time)

• Permits: median and 90th‑percentile NEPA timelines by sector (EAs ≤1 year, EISs ≤2 years; page caps 75/150–300).  
• Power: MW of firm, low‑carbon capacity contracted for strategic sites; median interconnection time under Order 2023 cluster studies.  
• Trade friction: AEO share of imports, clearance times for CTPAT/AEO vs non‑AEO cargo.  
• Security hygiene: share of export‑control/outbound actions with licence pathways and sunset reviews.  
• Tax competitiveness: average effective tax rate for new capex; share of R&D fully expensed under reformed §174.  

VI. Why this works (and avoids a subsidy war)

• It targets security rather than punishing every import, reducing collateral damage to allies and US manufacturers that need foreign inputs. (Narrow controls + outbound screening, not blanket barriers.)  
• It makes speed the incentive: NEPA Phase 2, Order 2023 interconnection, Orders 1920/1977 for transmission give investors what they actually price—time certainty.  
• It is fiscally conservative: fewer blank cheques, more capacity (power, permits) and neutral tax treatment (R&D expensing) that raise productivity without permanent outlays.  
• It rebuilds trust with partners by preferring AEO/CTPAT mutual recognition and case‑based remedies to headline tariffs—and by respecting allies’ climate instruments (CBAM) via reciprocity rather than retaliation.  

Policy Boxes

Box A — “Security, not autarky”: the five‑point carving test

A measure passes if it is (1) a chokepoint technology; (2) with documented dual‑use risk; (3) amenable to allied alignment; (4) offers a licensing path; and (5) has a sunset/review baked in. (Think BIS chip rules + allied toolmaker alignment.)  

Box B — The 100‑day wins

• Launch permit clock dashboard (CEQ/NEPA). • Publish AEO fast‑lane metrics (CBP). • Issue “No MW, No Money” subsidy guidance. • Send §174 expensing fix to Congress (one‑page bill). • Release Outbound licence guidance with safe harbours.  

Box C — Talking points for allies

We are narrowing security to the real risks, speeding up lawful trade (AEO fast lanes), and competing on capacity—not subsidies. Where you price carbon (EU CBAM), we will recognise clean US production rather than slap flags on tariffs.  

The punchline

A conservative course correction does not apologise for strength—it focuses it. Protect the crown jewels with allied rules; win investment with power and permits; discipline the tax code to reward real R&D; run a law‑and‑order, skills‑first immigration system; and measure everything. That is how America competes hard, spends smart, and keeps friends—without stumbling into an expensive, endless subsidy war.

How to defend critical technologies, keep markets open, and win the race for capacity—without bankrupting the treasury

I. The new map (what changed)

• Tariff shields spread. In 2024 Washington raised Section 301 tariffs on an array of Chinese products—EVs to 100%; several clean‑tech inputs and chips higher—while tightening de minimis loopholes used by e‑commerce platforms. Brussels, after a year‑long probe, imposed definitive countervailing duties on Chinese EVs. Expect more targeted, case‑based actions.  
• Export controls bit—and ripostes followed. The US tightened the China high‑end chip rule(Oct 2022/Oct 2023/Apr 2024) and refined licensing pathways in late‑2024; the Netherlands further restricted sales by ASML. Beijing retaliated with export controls on gallium, germanium and related materials, squeezing inputs.  
• “Outbound” screening arrived. America’s Outbound Investment Security Program took effect January 2, 2025; the EU recommended a 15‑month outbound risk review for advanced tech, a prelude to possible EU‑level rules.  
• Carbon walls go up, slowly. The EU’s CBAM runs transitional reporting (2023–25) and charges start in 2026. It will shape steel, cement and other resource‑heavy trade even if others do not copy it.  
• Inputs rose to strategy. The EU’s Critical Raw Materials Act hard‑codes 2030 targets—10% extraction, 40% processing, 25% recycling, ≤65% dependence on any one third country—and partners broadened the Minerals Security Partnership to mobilise offtakes and finance.  
• The WTO’s safety net frayed. MC13 promised dispute‑settlement reform; the Appellate Body remains paralysed. In the interim, more than 50 members use the MPIA arbitration workaround (the UK joined in 2025). Security exceptions are being invoked more often—raising the stakes for narrow, evidence‑based measures.  

II. Five principles to avoid a subsidy war

1. Security: narrow the aperture, widen the coalition. Focus controls on genuine chokepoints (advanced lithography, high‑bandwidth AI accelerators, secure compute). Default to multilateral or minilateral measures with licensing pathways; over‑broad national actions invite copycats and retaliation.  
2. Trade lanes: make “trusted” faster and cheaper. Expand AEO/CTPAT mutual recognition so low‑risk shipments benefit from fewer inspections and priority lanes, while high‑risk flows face the friction. That is a better deterrent than blanket tariffs.  
3. Investment: buy minutes and megawatts, not headlines. Offer predictable permits and firm, clean power to all comers that meet security and disclosure standards, instead of escalating chequebook bids. (See our earlier “minutes & megawatts” playbooks.) Speed beats subsidies.
4. Climate: tax carbon, not flags. Where carbon pricing exists, rely on CBAM‑style border adjustment not origin‑based tariffs. It rewards cleaner producers regardless of nationality and blunts “overcapacity” complaints.  
5. Rules: litigate again. Where politics allow, commit to MPIA appeals or other agreed arbitration, and publish the evidence base for trade‑remedy cases (injury, causation, subsidy). This is the cheapest way to keep the system open.  

III. The toolkit (what to do on Monday)

1) Build Input Security Compacts instead of copy-paste subsidies

• Join and use the Minerals Security Partnership: co‑finance mines, refining and recycling with offtake contracts; publish gap‑to‑goal tables against CRMA‑style benchmarks.  
• Stand up price‑stability CfDs for critical minerals to crowd in private capital; link eligibility to environmental & labour standards and disclosure of state support.

2) Narrow, allied export controls + licence corridors

• Keep the China chip rule focused on compute thresholds and tool lists; negotiate Validated End‑Userexpansion for secure data‑centre ecosystems so allied commercial demand is not collateral damage. Coordinate with the Netherlands/Japan so ASML/Nikon rules remain aligned.  
• Require human‑in‑the‑loop risk assessments for additions to control lists; sunset review every 18–24 months.

3) Outbound screening with safe harbours

• Copy the US model: prohibit narrow (e.g., funding PRC sub‑7nm logic, AI training chips, quantum tied to military end use); notify broader categories; exempt passive index funds and routine services. The EU’s 2025 recommendation offers a starting calendar.  

4) Trusted‑trade lanes that actually move

• Conclude AEO mutual‑recognition deals (EU, US, Japan, UK already recognise one another). Give fee rebates, inspection priority and data‑sharing dashboards to AEO shipments. Tie benefits to supply‑chain forced‑labour due diligence.  
• Fix de minimis: apply tariff‑eligibility and data requirements to small parcels subject to trade remedies or Section 301/232; mandate 10‑digit HTS and known‑shipper IDs. This dulls tariff evasion while sparing legitimate micro‑commerce.  

5) Use remedies, not broad wars

• Prefer countervailing/anti‑dumping cases with transparent evidence (as in the EU’s EV decision) over across‑the‑board tariff hikes. Publish non‑confidential case files; allow price undertakings that keep supply available.  
• Police anti‑circumvention surgically (e.g., US solar after the 2024 moratorium), with clear certification paths to compliance.  

6) Carbon clubs beat tariff volleys

• If you price carbon, link with partners and extend CBAM‑style adjustments; if you don’t, use product standards (embedded‑carbon disclosure) with mutual recognition pathways for clean producers everywhere.  

7) Standards and tech alliances

• Use the EU‑US Trade & Technology Council and IPEF Supply Chain Agreement to lock‑in testing and conformity MRAs, common data schemas, and emergency logistics protocols. These are low‑cost, high‑impact alternatives to chequebook contests.  

8) Procurement and subsidies: outcome‑priced, time‑limited

• Retire open‑ended “jobs” multipliers. Pay for MW connected, tonnes decarbonised, uptime delivered, or units shipped—and cap per‑project awards. The EU’s CISAF shows how to channel state aid toward clean‑industry outcomes with guardrails.  
• Ban cross‑border bidding wars with a Most‑Favoured‑Subsidy clause inside trusted‑partner pacts: if you top up for one, offer equivalent terms to others—disincentivising escalation.

9) Security exceptions: use the WTO off‑ramp sparingly

• Where you must invoke GATT XXI, publish the national‑security rationale and proportionality analysis; where possible, channel disputes into MPIA appeals to rebuild predictability.  

IV. Country dials (same logic, different settings)

United States

• Shift from tariff breadth to remedy depth. Use 301/232 sparingly; let targeted CVD/AD and de minimis reform do the work.  
• Outbound rules: maintain narrow prohibitions + licence corridors; sharpen Validated End‑User pathways to keep allied compute trade open.  
• Allied lanes: expand AEO MRAs and publish a trusted‑shipper leaderboard for priority processing.  

European Union/UK

• Stick to case‑based defence (EVs, wind) and keep CBAM timetables firm. Use CISAF to buy decarbonisation and grid capacity, not duplication of allied subsidies.  
• Security plumbing: update FDI screening (inbound) and complete the outbound risk review; join or use MPIAto keep appeals alive.  
• Wind & solar discipline: pursue FSR probes transparently; insist on price undertakings to avoid supply shocks.  

Advanced Asia (Japan, Korea, Singapore)

• Export controls aligned with US/EU while Investment Promotion (ESPA/METI) channels funds to fabs and packaging—paired with MRAs and AEO upgrades for frictionless exports.  
• Minerals strategy: leverage MSP financing windows and offtakes; publish project pipelines and ESG filters.  

V. A simple message voters and firms can both accept

To citizens: We’re protecting the crown‑jewel technologies with narrow, allied rules—and keeping everyday trade open so prices stay sane.

To firms: If you meet the security and disclosure rules, your reward is speed: faster permits, firm power, and trusted lanes. No subsidy arms race, just certainty.

Policy Boxes

Box A — “Security, not autarky”: the carving test

Use a security measure only if all five are true: (1) Chokepoint tech; (2) Dual‑use risk evidenced; (3) Allied alignment plausible; (4) Licensing path exists; (5) Sunset/review in statute. (Think BIS chip rules + ASML alignment; avoid sector‑wide bans.)  

Box B — Trusted‑trade lane (AEO/CTPAT) in one page

• Who: certified operators; What: risk‑based facilitation; How: MRAs among customs agencies; Benefits: fewer inspections, priority release, data reciprocity. (EU–US–Japan–UK MRAs already live.)  

Box C — Carbon clubs beat flags

• Rule: measure embedded emissions, not origin.

• Tool: CBAM‑style adjustment timed with ETS free‑allowance phase‑out.

• On‑ramp: MRAs for verification methods; transition aid via Social Climate Fund.  

Scoreboard (publish quarterly)

• Security narrowness: share of measures with licence pathways and sunset reviews.
• Trade friction: AEO share of import value; average clearance time vs non‑AEO.  
• Legal hygiene: share of trade‑defence cases with public non‑confidential files; use of MPIA where relevant.  
• Inputs secured: offtake coverage vs CRMA targets; MSP‑backed projects reached FID.  
• Subsidy efficiency: €/MW connected (power), €/tCO₂ abated (industry), % projects reaching milestones under CISAF.  

The punchline

Great powers will still quarrel. But most countries can compete without a subsidy war by buying capacity (power, ports, people), calibrating security (narrow, allied, sunsetted), and keeping trusted lanes wide open. That is how liberal economies stay open, resilient—and credible—when geopolitics runs hot.

Sources (selected)

• US Section 301 updates & de minimis reform: USTR notice; Reuters coverage on tightening de minimis.  
• EU EV duties (definitive): European Commission notice (Implementing Regulation 2024/2754).  
• US chip export controls (2022–2024) & VEU pathways: BIS background and Federal Register summary.  
• ASML export‑restriction impact: ASML press release, Dec 2 2024.  
• Chinese materials controls: Reuters on gallium/germanium/antimony curbs; CSIS analysis.  
• CBAM timetable: European Commission (definitive regime 2026).  
• EU Critical Raw Materials Act targets: European Commission explainer.  
• Outbound investment screening: US Treasury final rule (effective Jan 2 2025); EU recommendation & Reuters.  
• WTO dispute‑settlement & MPIA: WTO MC13 note; WTO MPIA case/news; Reuters on UK joining.  
• AEO/CTPAT MRAs & WCO SAFE: EU customs; CBP MRA; WCO SAFE framework.  
• CISAF state‑aid framework: European Commission pages and summaries.  
• US solar anti‑circumvention context: US DOC/Trade.gov brief; industry coverage.  

---

Why prepare a UBI now (but keep it dormant)?

History’s best natural experiment in “cash for all”—Alaska’s Permanent Fund dividend—suggests small, universal transfers need not kill work. Over four decades, researchers find no fall in overall employment and a modest rise in part‑time work (~1.8 ppts) when the dividend arrives, consistent with local‑demand effects offsetting any labour‑supply drag.  

Formal trials echo this. Finland’s two‑year basic‑income experiment improved well‑being and eased bureaucracy; employment effects were small/insignificant. Stockton’s guaranteed‑income pilot, by contrast, raised full‑time employment and improved financial stability—likely because volatility fell and job search improved.  

But design matters. The COVID‑era surge showed that large, deficit‑financed transfers into supply‑constrained economies can stoke inflation. Federal Reserve researchers attribute roughly ~3 percentage points of the 2021 inflation burst to unusually strong U.S. fiscal support (estimates vary). A UBI must therefore be revenue‑backed, smoothed, and conditional on macro room.  

The UBI “autopilot”: rules, not discretion

Pass the plumbing now; pay later, by rule. Enact a statute that defines the benefit formula, funding sources, and automatic triggers, but keeps the per‑capita amount small until the triggers fire.

Formula (illustrative):

Base dividend per adult each quarter = three‑year moving average of earmarked per‑capita revenues (see Pay‑fors) × safety factor, where the safety factor adjusts to keep the primary deficit within an agreed band over the cycle.

Escalator triggers (all must hold for ≥6 quarters):

1. Productivity–pay gap: real output per hour rising faster than median real earnings.
2. Jobs shortfall: prime‑age employment‑to‑population below a benchmark trend.
3. Inflation anchor: 2‑year‑ahead expectations inside the target band.

Pause rule: if market‑ or survey‑based inflation expectations breach the band, freeze the escalator. (Think of it as a fiscal version of the “Taylor rule”—transparent and reversible.) Evidence on pandemic‑era inflation underscores why this brake matters.  

“We’ll turn on more cash only when productivity is up, middle incomes aren’t keeping pace, jobs lag—and inflation expectations stay behaved.”

Keep it pro‑work: a small universal floor + an earnings bonus

A UBI that replaces earnings is a political and economic dead end. Instead:

1. A modest universal floor (predictable, monthly), high enough to insure against shocks and bargaining‑power losses as automation spreads, but not high enough to discourage work.
   • An earnings bonus (negative‑income‑tax/EITC style) that rises with wages at the bottom and phases out slowly. Decades of evidence show the EITC boosts labour‑force participation—especially for single parents—while any reductions are concentrated among some secondary earners; shallow phase‑outs keep work incentives strong.  
2. Participation “credits,” not sanctions. Add a small monthly kicker for verified hours (work, accredited study, caregiving, or community service). No punitive conditionality; just a bonus for participation.
3. Nudge saving, not a spending spike. Default 20–40% of the dividend into Skills/Retirement/Health wallets(fully opt‑out). Defaults are powerful: automatic enrolment has massively raised 401(k) saving. The same behavioural logic tempers demand surges while building household balance sheets.  

---

Pay‑fors that scale with the economy (not with headcount)

1) Carbon & climate revenues (where they exist).

• In the EU, the ETS raised €38.8bn in 2024 (cumulative ~€184bn since inception). The new ETS2 (fuels for buildings/road transport) starts in 2027 and will also feed the Social Climate Fund (SCF), budgeted at €86.7bn (2026–2032). Dedicate a fixed share of ETS/ETS2 proceeds to the dividend.  
• Note (Canada): the federal consumer fuel charge and Carbon Rebate to individuals ended on March 15, 2025; any UBI there would need alternative bases (e.g., resource royalties/LVT) or provincial schemes.  

2) Land‑value taxation / value capture.

A classic economist’s favourite: LVT is efficient (non‑distorting) and progressive on wealth. Hypothecate a slice of municipal LVC (zoning/transport uplifts) or a national LVT to the dividend.  

3) Stock buyback excise—use the increment.

The U.S. already levies 1% on net buybacks; the administration has proposed 4%. Earmark only the increment above today’s rate to the UBI fund to avoid double‑counting baselines.  

4) Sovereign‑asset dividends (resource royalties, spectrum, landing rights).

Where resource rents or licence auctions exist, follow Alaska’s playbook: seed a Future Dividend Fund whose investment returns co‑finance the UBI. Alaska’s labour‑market record—no job loss, slight part‑time increase—suggests such funding can be macro‑benign.  

5) CBAM/ETS linkages (EU/UK).

As the carbon border levy (CBAM) phases in from 2026, reserve a modest, rules‑based share to the dividend to keep the politics of climate policy durable.  

Delivery plumbing: identity, rails, and fraud

• Payment rails: use real‑time systems (e.g., the U.S. FedNow, launched 2023) to pay monthly and reduce leakage.  
• Identity & compliance: leverage existing tax/benefit IDs; apply KYC/AML proportionality (low risk for domestic, small, recurring transfers).
• Transparency: publish quarterly dashboards: revenue per capita, the moving average, the safety‑factor setting, and the state of triggers and pause rules.

Inflation safety: pair cash with supply unlocks

Cash without capacity can be inflationary. The autopilot therefore requires supply levers to move in step with any benefit increase: housing permits, grid/energy build‑out, and childcare places. Automatic stabilisers literature supports rule‑based fiscal tools; align UBI escalators with pre‑committed supply timetables to prevent bottlenecks from turning into price spikes.  

Country dials (same logic, different sources)

United States

• Funding mix: buyback‑tax increment + LVT/value capture + spectrum and other licence revenues; state‑level carbon where applicable.  
• Integration: fuse the earnings bonus with the EITC (monthly advance option), maintaining shallow phase‑outs that preserve labour‑supply gains documented in the literature.  

European Union/UK

• Funding mix: dedicated shares of ETS/ETS2 and CBAM receipts (complementing the SCF), plus municipal LVC. Do not crowd out the SCF’s poverty‑mitigation mandate; add a dividend slice on top as revenues scale.  
• Governance: a euro‑area template for autopilot triggers avoids a subsidy race; member states adapt payout size and earnings‑bonus parameters to local labour markets.

Resource‑rich & emerging economies

• Funding mix: sovereign‑asset dividends (royalties, spectrum) seeded into an endowment; gradual LVT adoption; carbon‑market proceeds where present. Alaska’s experience shows how to build legitimacy early.  

---

Policy Boxes

Box A — UBI Autopilot (one‑page explainer)

• Base: quarterly per‑capita payout = 3‑year average of earmarked revenues × safety factor.

• Escalator: turns on only if (productivity > median‑pay) and (prime‑age EPOP below trend) and (inflation expectations anchored) for ≥6 quarters.

• Pause: automatic if expectations breach band.

• Nudges: default 20–40% to Skills/Retirement/Health wallets (opt‑out).

• Earnings bonus: a monthly EITC‑style top‑up with shallow phase‑outs.

Box B — Funding menu (country mix differs)

• EU/UK: ETS/ETS2 shares + CBAM, with SCF unaffected; LVT/LVC.  

• US: Buyback‑tax increment + LVT/LVC + spectrum; carbon where present.  

• Resource‑rich: royalties/licences into a Future Dividend Fund (Alaska‑style).  

Box C — Why this stays pro‑work

• Small universal floor + an earnings bonus that raises returns to work at the bottom.

• Evidence: EITC expansions increase participation; careful phase‑outs avoid second‑earner penalties.  

Implementation playbook (12–24 months)

Phase 0 (0–6 months): Law & rails

• Pass the Autopilot Act: define revenue streams, formula, triggers, pause rule, and dashboards.
• Map revenue baselines: ETS/CBAM shares (EU/UK), buyback‑increment/LVT/spectrum (US), royalties (resource‑rich).
• Stand up real‑time payment integration (e.g., FedNow in the U.S.).  

Phase 1 (6–12 months): UBI‑Lite

• Start a small dividend, strictly from earmarked revenues; launch the monthly earnings bonus (advanceable EITC in the U.S.).
• Default 20–40% into Skills/Retirement/Health wallets (opt‑out).  

Phase 2 (12–24 months): Trigger readiness

• Publish the trigger dashboard quarterly (productivity vs median pay; prime‑age EPOP; inflation expectations).
• Adopt paired supply milestones (housing permits, grid capacity, childcare seats) that must be met before any escalator step takes effect.

KPIs

• Median real earnings vs productivity (gap closing).
• Prime‑age EPOP trend.
• Inflation expectations within band.
• Poverty among working‑age adults and child poverty (monthly). (The 2021 expansion of the U.S. Child Tax Credit offers a benchmark: it lifted ~2.9m children from poverty and pushed the SPM child‑poverty rate to 5.2%, illustrating the power of monthly payments.)  

The politics

Tell voters: “As machines get better, your life gets safer.”

Tell investors: “We won’t yank the wheel—benefits scale by rule, not whim.”

Tell central bankers: “This is pay‑as‑you‑go and pause‑able; it won’t fight your anchor.”

Bottom line: Build the UBI rails now. Fund them from rents, pollution and land, not from payrolls. Keep a small universal floor, add a work‑boosting earnings bonus, and only scale when the economy can absorb it. That is how a liberal, future‑proof settlement shares the gains of abundant machines—without stoking inflation or dulling the urge to work.

Sources (selected)

• Alaska dividend — employment effects: Jones & Marinescu, AEJ: Economic Policy (2022); NBER WP (2018).  
• Finland basic‑income results: Kela/European Commission summaries (no significant employment effect; better well‑being).  
• Stockton SEED outcomes: University of Pennsylvania / SEED briefing (full‑time employment up; financial stability improved).  
• Pandemic‑era inflation & fiscal contribution: SF Fed Economic Letter (≈3 ppts); recent Fed staff survey of estimates.  
• EU ETS/ETS2 revenues & SCF: ICAP EU‑ETS profile (2024 revenue €38.8bn, cumulative €184bn); EC pages on ETS2 and SCF.  
• CBAM timeline: European Commission/Parliament (transition 2023–25; full from 2026).  
• Canada Carbon Rebate change (2025): CRA (federal fuel charge and rebate to individuals stopped March 15, 2025).  
• Buyback excise (US): IRS final rules (1%); proposals to raise to 4% (FY2025 Budget).  
• Land‑value tax efficiency: IMF working paper (Schwerhoff, Edenhofer, Fleurbaey, 2022).  
• EITC & labour supply: Eissa–Liebman (1996 QJE) and subsequent reviews.  
• Defaults & saving: Madrian–Shea (2001 QJE) on automatic enrolment.  

The market failure no one campaigns on

Left to themselves, companies train less than society needs. When skills are general—valuable to many employers—rivals can poach newly trained staff. That “poaching externality” means the trainer bears the cost while others free‑ride; rational firms under‑invest. Modern labour‑economics has formalised this problem for decades. 

The result shows up in the data: across the OECD, adult learning is unequal—high earners participate far more than low earners—precisely the wrong way round if automation is to lift all boats. 

Fixing the externality takes a simple instrument: a hypothecated levy on payroll that every medium‑large employer pays, and a rebate (with a small kicker) that firms earn only when training delivers verified results. It is not a new idea. Singapore has long run a 0.25% Skills Development Levy (SDL); France makes firms contribute 0.55% (≤10 employees) or 1% (≥11) to continuing training; Korea uses its Employment Insurance system to reimburse training costs, with extra generosity for SMEs. The lesson is not that levies exist—but which designs work. 

What we have learned from other people’s mistakes

Britain’s Apprenticeship Levy offers cautionary tales. Starts fell sharply after 2017; audit bodies and think‑tanks flagged “rebadging” of existing management training to draw down funds; billions in employer payments went unused or expired; and the system skewed toward older, higher‑level courses. The cure is flexibility + outcomes + ring‑fences for young starters, not scrapping levies. 

The model to copy: an Employer Levy & Rebate (ELR)

Thesis: don’t tax robots—tax not training. Make training the automatic companion to automation.

A. The levy

1. Base: 0.5–1.0% of payroll on employers above a size threshold (e.g., >50 workers), collected with existing payroll remittances.
2. Pooling: SMEs pool via sector councils so they can host apprentices without running full academies.
3. Hypothecation: proceeds go to a ring‑fenced Skills Fund audited annually.

B. The rebate (paid only for outcomes)

• Milestones: completion → recognised credential → job/role placement → 6–12‑month earnings gains.
• Rates: higher rebate for young entrants, SMEs, and scarce occupations (mechatronics, maintenance, controls, industrial IT).
• What counts: short, modular courses and apprenticeships; on‑the‑clock training during automation roll‑outs (see short‑time‑work below).
• Dashboards: publish provider‑ and employer‑level completion, placement, earnings metrics.

Why outcomes? Cross‑country evidence shows firm returns to apprenticeships vary widely—positive in Switzerland on average; costlier in Germany unless tasks are well designed. Paying for results aligns everyone’s incentives. 

Guardrails that keep the system honest

1. Anti‑rebadging rule. No using levy cash to relabel generic management CPD as “apprenticeships”. (Britain learned this the hard way.) Impose caps on generic management programmes and require new‑to‑role hires or clear upskilling evidence. 
2. Quality gates for providers. Admission and continued access depend on measured outcomes; serial under‑performance triggers rapid defunding.
3. Transferability across supply chains. Allow prime contractors to transfer levy credits to their suppliers to develop regional ecosystems—without recreating the UK’s rigidities.
4. Young‑people ring‑fence. Reserve a floor share of rebates for under‑25 apprenticeships at Levels 2–3 (or national equivalents) to avoid hollowing out entry routes. Evidence from the UK shows the entry‑level collapse when this is ignored. 
5. Open credentials, portable benefits. Credentials must be stackable and recognised across employers; benefits should follow the worker.
6. Fraud policing. Independent audits; randomised checks; heavy penalties for “ghost trainees”.

Make automation a training window, not a redundancy machine

Short‑time‑work + training should sit inside the ELR. When factories install new lines, hours can fall temporarily; the state tops up wages while workers retrain on the clock. OECD evaluations show these schemes preserved jobs in shocks when tightly targeted and time‑limited. Germany kept millions attached to firms during downturns; Spain’s ERTE reforms locked in similar mechanics. Use that same logic for automation roll‑outs. 

Fiscal arithmetic (the “adult” bit)

• Gross in, net out. The levy collects every month; rebates are earned only on outcomes, so unearned balances finance public seats at community colleges/polytechnics.
• No unfunded promises. Tie any investment tax credit for automation to neutrality—i.e., the credit vests fully only if value‑added per worker rises and payroll hours are maintained or higher over a rolling window—with the ELR as the enforcement arm.
• Complementary revenues: where appropriate, dedicate slices of carbon/ETS/CBAM receipts and land‑value capture around upgraded industrial corridors to the Skills Fund (steady, non‑distorting bases).

Country dials (same logic, different settings)

United States.

• Introduce a federal ELR (exempt <50 workers; phase‑in rates).
• Let firms draw rebates through community colleges, union JATCs and accredited in‑house academies.
• Piggyback on state UI systems to deliver short‑time‑work + training options.
• Make CHIPS/IRA/IIJA awards universally conditional on apprenticeship ratios and local training endowments (already true in the stronger awards). 

European Union/UK.

• Modernise existing levies: allow modular courses and supply‑chain transfers; publish outcome dashboards; ring‑fence youth places.
• Align with the Net‑Zero Industry Act permit clocks so training seats are ready before plants open.
• Use CRMA targets to guide sector priorities (mechatronics/maintenance for battery, solar, wind, refining). 

Advanced Asia.

• Singapore/Korea already have the bones: keep 0.25%+ rates, scale outcome‑based payments, and expand SME pooling.
• Embed on‑the‑clock training into short‑time‑work rules during upgrades. 

Implementation in 12 months (with KPIs)

Month 0–3: Law & plumbing

• Pass the ELR Act (levy bands; rebate milestones; dashboards; audits).
• Set priority occupation lists; publish approved provider criteria.
• Stand up a Major Projects Unit so training timelines match permit clocks.

Month 4–8: Pilots that matter

• Launch two regional pilots (heavy industry + electronics).
• Fund regional training labs with OEMs; build simulator/digital‑twin capacity.
• Begin short‑time‑work + training trials tied to real automation projects.

Month 9–12: Scale & publish

• Expand to national coverage; enforce young‑people ring‑fence.
• Publish the first set of provider league tables (completion, placement, 12‑month earnings).
• Tie public capex grants to verified training seats and open‑jobs‑first rules.

KPIs (public, quarterly):

• Training uptake: +15–20 pp among non‑degree workers.
• Outcomes: ≥80% placement in field; +8–12% median earnings for completers at 12 months.
• Fairness: share of rebates going to under‑25 and SMEs.
• Resilience: layoff spikes −25–35% at plants using short‑time‑work + training. (Benchmarks from OECD/IMF studies on job‑retention schemes.) 

Policy Boxes

Box 1 — ELR in one page (printable)

• Levy: 0.5–1.0% of payroll on mid/large employers; SMEs pool via sector councils.

• Rebate: paid only on completion → credential → placement → earnings milestones.

• Uplifts: youth, SMEs, hard‑to‑fill roles.

• Transparency: open dashboards; randomised audits; fraud penalties.

• Compatibility: short courses and apprenticeships; on‑the‑clock training during upgrades.

Box 2 — What not to do (lessons from the UK)

• Don’t lock funds into long programmes only. Allow modular training.

• Don’t reward attendance—pay for outcomes.

• Don’t tolerate rebadging. Cap generic management programmes; demand new‑to‑role evidence.

• Do ring‑fence youth places and allow supply‑chain transfers. 

Box 3 — Why levies beat tax cuts (the economics)

• Levies fix poaching externalities; general tax cuts don’t.

• Evidence: 75 countries operate levy‑financed training funds; well‑run schemes (Singapore, France, Malaysia) show administrative feasibility. 

Footnotes & sources (selected)

• Poaching externalities & under‑investment: Acemoglu & Pischke, classic papers on general training; related empirical work. 
• Inequality in adult learning: OECD Trends in Adult Learning (participation gaps by income). 
• Levy exemplars: Singapore SDL 0.25% (official guidance); France CFP 0.55%/1% (URSSAF/OPCO); Korea Employment Insurance rebate‑based training (OECD profile). 
• UK pitfalls: NAO on post‑2017 starts; House of Commons/think‑tanks/CIPD on rebadging and unspent funds. 
• Cost‑benefit of apprenticeships: OECD syntheses; Swiss vs German firm returns. 
• Short‑time‑work effectiveness: OECD/IMF assessments (Germany, Spain’s ERTE). 

A practical guide to winning the automated‑industry race

Why these three levers decide who wins

Megawatts. AI and automation are energy‑hungry. The International Energy Agency reckons global data‑centre electricity use will double to about 945 TWh by 2030, nudging 3% of world demand; AI‑optimised facilities drive most of the rise. Jurisdictions able to deliver firm, low‑carbon power at predictable prices are pulling ahead in landing both automated production and compute.  

Minutes. Speed beats chequebooks. Europe’s Net‑Zero Industry Act (NZIA) sets hard permit clocks—9 months for smaller strategic manufacturing projects, 12 months for large ones and 18 months for CO₂ storage. America’s revamped NEPA rules lock in time limits—one year for Environmental Assessments and two for Environmental Impact Statements—plus page limits of 75 and 150/300 pages respectively. If you cannot match these clocks, you will lose projects even with richer subsidies.  

Minds. Robots lift productivity but change the mix of tasks. The winning model is not empty job talk but binding training finance: a tiny levy‑and‑rebate on payroll that firms earn back by delivering outcome‑verified apprenticeships and short, stackable credentials. Existing systems in Singapore (0.25% levy feeding the Skills Development Fund) and France (mandatory contributions to continuing training) show the plumbing; the lesson is to pay on completion, placement and earnings, not attendance.  

A fourth lever matters too: input security. Europe’s Critical Raw Materials Act sets 2030 targets (at least 10% of strategic‑material needs mined in the EU, 40% processed, 25% recycled; ≤65% from any single foreign supplier), signalling that offtakes and recycling are as strategic as factories.  

Finally, relative energy costs still shape siting. Analysis by Bruegel finds 2024 EU gas prices averaged nearly five times US levels and industrial power about 2½× America’s—though the gap varies by country. Policy must cut volatility and secure long‑term, low‑carbon supply.  

I. Megawatts — turn power into your strongest incentive

Offer an Industrial Power Guarantee. For strategic loads (fabs, robotics, automation campuses, data centres), provide 10–15‑year PPAs or Contracts‑for‑Difference for clean, firm power—renewables plus storage, uprated nuclear/SMRs, and long‑duration storage—indexed to inflation but insulated from gas spikes. Publish capacity mapsshowing where headroom exists and when it will be built.

Fix interconnection. In the United States, FERC Order 2023 moved to cluster‑based queues and firmer milestones to slash delays; Order 2023‑A tightened the rules. Emulate the spirit: transparent application windows, cure deadlines, and enforceable service‑level agreements between grid operators and strategic users.  

Make load flexible and useful. For AI campuses and automated plants, negotiate flexible‑load contracts (load shedding and shifting for a discount) and require heat‑reuse into district networks where feasible. Tie siting to water‑recycling plans.

Finance it without gimmicks. Channel ETS/CBAM or carbon revenues (where they exist), and land‑value capture from upgraded corridors, into a “Power & Interconnection Fund” that co‑finances storage, feeders and substations serving certified industrial parks.

Rule of thumb: No MW, no money. Every subsidy euro or dollar should be contingent on an executed power contract and a dated interconnection slot.

II. Minutes — a permit clock investors can set their watches by

Legislate the clock. Mirror or beat the NZIA and NEPA standards: ≤12 months for strategic factories (≤9 months for smaller ones); ≤1 year for EAs and ≤2 for EISs; page caps at 75/150 (300 for complexity). Build in “deemed complete” rules and clock‑stop disclosure so agencies cannot bury files. Publish dashboards of median and 90th‑percentile times by sector.  

One counterparty, not a paper chase. Create a statutory Major Projects Unit empowered to corral agencies, utilities and localities and to issue consolidated decisions. Allow parallel reviews (environment, grid, water) with a single record.

Enforce with teeth. If the state misses its own clock, projects receive automatic fee refunds or schedule‑damage payments; for applicants who delay, the clock pauses transparently.

Rule of thumb: Fast, fair, predictable beats bigger cheques nine times out of ten.

III. Minds — train or pay: a levy‑and‑rebate that makes skills automatic

Adopt an Employer Training Levy & Rebate (ELR). Levy 0.5–1.0% of payroll on medium/large employers; rebate(with a small kicker) only when firms deliver verified outcomes: completion → industry credential → job/role placement → 6–12‑month earnings. SMEs pool via sector councils. Singapore’s SDL and France’s formation contributions are the working templates; your added value is hard outcomes and open dashboards.  

Guarantee apprenticeships in priority occupations. Under‑25s should be able to secure an apprenticeship within 90 days in mechatronics, maintenance, controls, robotics, and industrial IT—with on‑the‑clock training funded through ELR.

Couple automation with work‑sharing and training. When plants upgrade, use short‑time‑work schemes so hours fall temporarily, wages are topped up by the state, and staff train on the clock. Evidence from the OECD and IMF shows such schemes preserve jobs and speed recoveries when well‑targeted and temporary.  

IV. Markets — secure inputs without a subsidy war

Use offtakes and price floors, not blanket bans. For lithium, nickel, copper and rare earths, deploy sovereign off takes and CfDs to crowd in refining and recycling. Align targets with Europe’s CRMA benchmarks (10/40/25 and ≤65% single‑country dependence). Publish an annual gap‑to‑goal.  

Keep lanes open. Pair critical‑mineral alliances with trusted‑trader lanes and mutual recognition of conformity assessments, so strategic inputs flow even in rough geopolitical weather.

V. Local statecraft — investor concierge and power‑ready sites

Build an Investor Concierge with real powers. Governors and mayors should stand up a single team with authority over permits, utilities, land assembly, housing and training seats. KPI: time‑to‑yes.

Certify power‑ready sites. Copy the American “certified site/megasite” playbook—pre‑zoned, geotech done, wetlands cleared, utilities sized, and interconnection studies complete. North Carolina’s programme offers a representative checklist; TVA‑style megasites show how pre‑work de‑risks large projects.  

Do aftercare, not just attraction. Make property‑tax abatements contingent on expansion milestones, apprenticeship ratios, local supplier development and heat‑reuse from data centres into nearby districts.

VI. Money and fairness — incentives that buy real outcomes

Condition subsidies on productivity and payroll. Replace job‑count box‑ticking with a Productivity Complementarity Credit: the investment credit vests only if value‑added per worker rises and aggregate payroll hours are maintained or higher over a rolling window. Publish the metrics.

Profit‑sharing and broad‑based ownership on public‑money projects. Evidence from NBER’s “shared capitalism” work and other studies links profit‑sharing/ownership with higher productivity, lower turnover and better worker wealth (effects vary but are positive on average). Make it tax‑neutral to choose ownership.  

Earnings boosts with shallow phase‑outs. Expand EITC‑style top‑ups so automation does not hollow out lower‑wage households; avoid cliffs. In cyclical shocks, scale short‑time‑work to keep workers attached while they train.  

Don’t forget the rulebook. For AI‑heavy projects, regulatory certainty matters. The EU AI Act entered into force in 2024 with phased obligations—notably for general‑purpose AI models from August 2025 and most high‑risk rules by 2026–27. Align sandboxes and compliance roadmaps to those dates to de‑risk investments.  

VII. The scoreboard — how you’ll know you’re beating America

Publish a simple dashboard, updated quarterly:

• Power: MW of firm, low‑carbon capacity contracted; median interconnection time to energisation; average industrial PPA price.
• Permits: median and 90th‑percentile time‑to‑decision for strategic projects; share hitting statutory clocks.  
• People: ELR enrolments and completion/placement/earnings rates; apprenticeship seats per 1,000 workers.  
• Pipelines: critical‑mineral offtake coverage vs. CRMA targets; recycling capacity additions.  
• Fairness: share of subsidised megaprojects with profit‑sharing/ownership; growth in median pay relative to value‑added.  

Policy boxes

Box A — The Three Guarantees (for your investor term sheet)

Power: A firm, clean price‑and‑access contract (PPA/CfD) plus an interconnection SLA tied to grid investment.  

Permit: One counterparty and a 12‑month decision clock (9 for smaller projects). 1‑year EA / 2‑year EIS with page caps.  

People: An apprenticeship guarantee and pre‑approved training rebates via the ELR.  

Box B — Training Levy‑and‑Rebate (ELR) in one page

• Levy: 0.5–1.0% of payroll on medium/large employers; SMEs pool via sector councils.

• Rebate: Paid only on completion → credential → placement → earnings milestones.

• Guardrails: Fraud audits; cap in‑house training without third‑party assessment; publish provider scorecards.

• Comparators: Singapore (0.25% levy to Skills Development Fund); France (mandatory training contributions).  

Box C — Permit Clock (what to copy & what to avoid)

• Copy: NZIA’s 9/12/18‑month deadlines; NEPA’s 1‑/2‑year time limits and 75/150(300) page caps; clock‑stop transparency.  

• Avoid: Silent resets, serial reviews, and unpriced utility delays—tie grid works to the decision clock.

The politics that make it stick

Promise investors certainty (power and permits) and citizens fairness (training, earnings boosts, profit‑sharing). Tie every public euro or dollar to verifiable outcomes—more value‑added and stronger payrolls—and publish the evidence. Add a UBI autopilot in statute as long‑run insurance if displacement outpaces job growth; keep it pay‑as‑you‑go and work‑friendly.

North Star: Build fast, share fair.

Translation: Megawatts, minutes, minds—with markets secured and money that buys real things.

Sources (selected)

• Data‑centre demand: IEA, Energy and AI – Energy demand from AI (projection to 945 TWh by 2030).  
  • Permit clocks: EU NZIA briefing (9/12/18 months); US NEPA Phase 2 (time and page limits).  
• Interconnection reform: US FERC Order 2023/2023‑A explainers (queue reform, certainty).  
• Energy price gap: Bruegel policy brief (EU gas ≈5× US; industrial electricity ≈2.5× US in 2024).  
• Critical Raw Materials targets: European Commission (10/40/25 and ≤65% single‑country dependence).  
• Training comparators: Singapore Skills Development Levy; France formation professionnelle contribution.  
• Shared capitalism evidence: NBER/IZA syntheses on profit‑sharing and employee ownership (productivity, turnover).  
• Short‑time‑work effectiveness: OECD and IMF analyses.  

Industrial policy for when robots are cheap and megawatts are dear

The new robot normal

The era of abundant machines has arrived. The global average robot density reached 162 per 10,000 manufacturing workers in 2023—more than double the level seven years earlier. The European Union averages 219, North America 197, and East Asia now leads outright. South Korea and Singapore sit at the top; China has vaulted into the global top tier after a blistering adoption spree.  

What turbocharges this trend is the parallel boom in data centres. The IEA projects that data‑centre electricity demand will more than double by 2030 to ~945 TWh (just under 3% of global consumption), with AI‑optimised facilities driving the surge. This is not a sideshow: where electricity is firm, clean and affordable, automated factories and AI campuses follow.  

Power is the new bottleneck

For much of the past decade, Europe’s factories faced a structural energy‑price gap with the United States. In 2024, EU wholesale gas prices averaged nearly five times US levels; industrial electricity averaged roughly two‑and‑a‑half times higher in the EU. Volatility is itself a tax on investment. Long‑term contracts and storage help, but the siting calculus starts with megawatts.  

Grid interconnection matters just as much as price. The United States’ FERC Order 2023 overhauled interconnection queues to cut delays and increase certainty—an under‑appreciated competitive edge for large loads like fabs and data centres. Jurisdictions that make grid hookups predictable and fast will win.  

Speed beats chequebooks

It is fashionable to offer giant cheques. Better to offer speed. Europe’s Net‑Zero Industry Act (NZIA) now imposes hard permit clocks: 9 months for smaller strategic manufacturing projects and 12 months for large ones (with 18 months for CO₂ storage). America’s revamped NEPA Phase‑2 rules implement page limits (EIS ≤150 pages, EA ≤75) and time limits (EIS ≤2 years, EA ≤1 year). If you cannot match these timelines, you will lose projects—even if you offer richer subsidies.  

The politics of fabs—and the risk of hollow promises

Washington’s CHIPS awards—to Intel (up to $8.5bn proposed), TSMC ($6.6bn final terms) and Micron ($6.1bnfinalised)—herald a wave of capital‑intensive and automation‑heavy plants. They are vital for strategic supply chains, but they do not recreate legions of low‑skill jobs. The jobs multipliers skew to suppliers, construction and local services; inside the fence, robots rule. That is why the better awards now tie subsidies to workforce and community benefits—a cue worth copying.  

Supply chains are now strategy

Control over inputs is the quiet front in this race. The EU’s Critical Raw Materials Act sets 2030 benchmarks of 10% extraction, 40% processing, 25% recycling, and ≤65% dependence on any one country—targets that are already steering capital toward refining and recycling. Similar alliances and offtake contracts will define where advanced factories land.  

How to win without wasting money

Compete on power. Offer 10–15‑year industrial PPAs/CfDs for clean, firm power (renewables + storage, nuclear uprates/SMRs or long‑duration storage). Pair every megawatt of new demand with megawatts of new build and interconnection SLAs. Energy + speed is the new incentive.  

Compete on minutes. Adopt statutory permit clocks that match NZIA and NEPA. Publish clock dashboards, limit page counts, and use “deemed complete” rules to stop procedural churn. Investors price uncertainty far higher than tax rates.  

Compete on people. Replace slogan‑heavy “jobs” talk with binding training finance. The model that works is a training levy‑and‑rebate: firms pay a tiny payroll levy (0.5–1.0%); they get it back—with a kicker—only when they deliver outcome‑verified apprenticeships or stackable credentials. Singapore (0.25% levy), France (0.55–1.0%) and Korea (skills via Employment Insurance) show the plumbing; Britain’s experience shows the pitfalls to avoid (rigidity, “re‑badged” courses).

“Build fast, share fair. In an age of abundant machines, the scarce inputs are megawatts, minutes and minds.”

Money without gimmick

Subsidies with teeth. Tie any public capex grant (chips, batteries, robotics) to apprenticeship ratios, local training endowments and open‑jobs‑first posting rules—just as the stronger CHIPS awards already do.  

Tax neutrality, not a “robot tax”. Today’s codes often favour capital over labour (full expensing vs payroll taxes). Instead of a blunt robot tax, equalise the wedge and condition investment credits on value‑added per worker rising and payroll hours being maintained or increased. (The IMF and others have mapped sensible AI‑era tax options; the principle is neutrality.)  

Pay‑fors that scale. Use a mix of:

• the levy itself (self‑recycling via rebates);

• a modest buyback‑tax increment to fund skills and broad‑based ownership;

• land‑value capture around upgraded grids and industrial parks; and, where it exists, slices of carbon/ETS/CBAM revenues (note Canada’s removal of the federal consumer carbon charge in 2025—industrial pricing remains).  

Spreading the gains (so the politics work)

Profit‑sharing and ownership. Big recipients of public money should adopt broad‑based profit‑sharing or employee‑ownership plans. The evidence across “shared capitalism” studies points to higher productivity and lower turnover—and a healthier distribution of gains. (Make it tax‑neutral to choose ownership.)  

Earnings boosts and work‑sharing. Expand earnings top‑ups (EITC‑style) with gradual phase‑outs; during automation roll‑outs, use short‑time work so people train on the clock instead of churning through layoffs. (Germany and others show it works.)  

A UBI autopilot in reserve. If automation pushes output faster than it pushes median earnings or employment, switch on a small, universal dividend financed from earmarked, smoothed revenues and paired with an earnings bonus. The Alaska dividend’s long‑run evidence shows no drop in employment and a modest shift toward part‑time work; city‑level pilots find improved job‑search stability. Guardrails—pay‑as‑you‑go, inflation “pause” rules, and partial defaulting into skills/retirement wallets—keep it fiscally and macro‑prudently sound.  

The “Three Guarantees” (investor term‑sheet)

Power: a firm, clean price‑and‑access contract (PPA/CfD) + interconnection SLA.

Permit: a single counterparty and a 12‑month decision clock.

People: an apprenticeship guarantee and pre‑approved training rebates.  

The Training Levy‑and‑Rebate (ELR)

• Levy: 0.5–1.0% of payroll (SMEs can pool via sector councils).

• Rebate: paid only on completion → placement → earnings milestones; transparent dashboards; tough anti‑fraud.

• Flexibility: funds usable for short, modular courses; allow supply‑chain transfers; bar “re‑badging” HR seminars.

UBI Autopilot (work‑friendly)

• Small universal floor + earnings bonus.

• Triggers: productivity‑pay gaps persist; prime‑age employment lags; inflation expectations on target.

• Pause: automatic if expectations drift.

• Partial illiquidity by default (skills/retirement wallets) to curb demand spikes.  

The punchline

This is the century of abundant machines. The countries that thrive will build fast (power + permits), share fair(people + pay), and budget honestly (self‑financing skills, targeted subsidies, and a UBI safety‑valve that doesn’t blow the boiler). The rest will throw money at ribbon‑cuttings, then wonder why the middle class keeps shrinking.

Sources (selected)

• Robot density & rankings: International Federation of Robotics, World Robotics 2024 (global average 162; EU 219; North America 197; China moves into top tier).  
• Data‑centre demand: IEA, Energy & AI: Energy demand from AI (to ~945 TWh by 2030).  
• Permit clocks: EU NZIA overview and EPRS briefing (9–12 months; 18 for CO₂ storage); US NEPA Phase‑2(page/time limits).  
• CHIPS awards: US Commerce/NIST and company releases (Intel up to $8.5bn proposed; TSMC $6.6bn; Micron $6.1bn).  
• Energy‑price gap: Bruegel policy brief (EU industrial electricity ≈2.5× US; 2024 gas ~5×).  
• Grid interconnection reform: FERC Order 2023 explainer.  
• Critical minerals: European Commission, CRMA benchmarks.  
• UBI evidence: Jones & Marinescu (Alaska AEJ 2022); Stockton SEED.  

and the striking similarity that resembles Dot-com Bubble

As the market continues to trend downward and internet conglomerate becomes stale in innovation, our economic growth is going to experience a precipitous downward trend. In addition to the natural market cycle that predicts a contraction in or around 2020-2021, trade tension and contractionary economic and financial policy might accelerate the progress, making deep depression inevitable.

The manufacturing industry is going to be the first old guard of the economy to fell victim to the tariffs, social insurance policy and contractionary monetary policy, resulting in mass unemployment and social unrest. The threat of economic stability is then going to trickle out towards other traditional industries, and negatively affect equity market’s price evaluation. At this period, the entertainment and fast food industry are going to observe extraordinary growth, as we currently witnessed in Chinese market.

After manufacturing industry collapse, retail service industry, a complementary industry, is predicted to be next to experience pain from economic slowdown. However, since the retail industry is already starting to slowly transform into e-commerce, the economic effect will be less observable. The newly established e-commerce law is also going to negatively affect the cash flow of e-commerce portal like Alibaba, JD and WeChat, and their revenue from transaction fees and advertisements.

The aforementioned problems are also going to raise the price of consumer goods. With low confidence and high purchasing price, rapid inflation is next to ensue. Currently, inflationary pressure comes from three sources: tariff related inflation, stricter taxation related inflation, and monetary policy related inflation. If the central bank decides to push expansionary monetary policy further, then the rapid inflation is going to be unstoppable.

Then we discuss whether housing market is going to absorb expansionary monetary policy like 2008 and 2013. Although the purchasing leans heavily towards property market, the restriction placed on buying secondary houses and the mortgage policy is going to stop most of the newly minted money from flowing into the market. Further, since the restriction placed on property market serves to maintain social stability, it is unlikely to be lifted.

Next, since Chinese wage growth is stagnant, the entire country is slowly slumping into middle income trap. In order to escape the middle income trap, the only two potential solutions are to innovate revolutionary technologies that breaks productivity barriers, and/or transform the economy to service-based economy. Since the second solution is long and painful, the only way out is through rapid technological innovation that increases the productivity exponentially.

There are a couple candidates for innovation breakthrough. Robotic automation technology that promises less reliance on human workers, automated vehicle that reduces wasted productivity time. Artificial intelligence that learns repetition in work. These revolutionary technologies depend on breakthrough in telecommunication technology (5G) for its 20 Gbps theoretical speed and its Massive MIMO that can support tens of thousands of device simultaneously.

5G technology is predicted to be delivered in 2020, with infrastructure deployment planned in years following the commercialization of technology. Although there isn’t a specific plan on the date of implementation, we can map this plan on 4G LTE deployment and predict that it would have wide coverage starting 2022-2023.

If the bubble burst before the 5G deployment, then the new wave of company is going to establish around 2024-2027, which mirrors the Made in China 2025 plan, and coincidentally, the advent of online economy begins at around 2004-2007, after 2001’s Dot-com Bubble burst. Most famous among them: Facebook, Google, Netflix and Amazon.

Escaping the middle-income trap is extraordinarily hard but breaking it would promote the country into the ranks of developed countries. Only with precision of execution and the open economy can this country become a revolutionary force to be reckon with.