What if Australians had time and personal AI opportunity by 2035?

The proposition

What if Australia made shorter paid working hours and individually controlled AI access national priorities, giving citizens and permanent residents the time and resources to learn by doing outside employment, explore their own ideas and adapt their lives through 2035? How could that opportunity be delivered so that Australia becomes more resilient to AI-related job disruption and harm to social cohesion?

This is the first programme in What If, a research family that begins from a proposition and investigates how it could be implemented by 2035. The programme context and research date were established on 9 September 2026; unpublished antecedents are excluded from the evidence for this release.

The unit of support is the individual. A person does not need an employer, degree, course, company, approved career plan or One World Community membership to receive initial access. Employment, enterprise and civic outcomes matter to the programme's evaluation; they are not tests of whether a person's lawful interests deserve funding.

1. Settled premises and open choices

The investigation accepts P1–P3 without requiring a preliminary proof that this is the only possible route to resilience. It still asks whether the specific service and working-time arrangements can function, remain affordable and avoid preventable harm. It does not promise that every person will use AI, become an entrepreneur or avoid displacement.

2. A concrete 2035 experience

The following is a design scenario, not a forecast. A 17-year-old can try models for a self-chosen project through a personal account, without needing school approval or family credit-card spending. A 58-year-old facing job disruption can explore a new role, a small service idea or an unrelated creative interest using the same entitlement. Each can decline optional coaching and keep private prompts private. Neither has to create a business to justify participation.

For employees, a shorter week means less actual work at unchanged ordinary weekly pay, with service coverage arranged by employers. Some use the released time for experimentation; others use it for care or recovery. People outside paid work can access AI immediately, so their opportunity does not depend on receiving a workplace time dividend.

An individual who wants further backing can submit an idea in an accessible format. AI prepares the assessment and uncertainty record. Named people make the final funding decision, and a failed business does not revoke the underlying personal entitlement. National infrastructure makes the service dependable across providers; infrastructure ownership is not itself the benefit offered to the person.

3. What the previous research contributes

The current published ORW collection supplies the initial evidence base and reader pattern. Its four published trilogies distinguish significant pathways, comparative claims and delivery conditions. The current index was retrieved on 9 September; the public reports and source register provide the release evidence. This is not a fresh claim-by-claim audit of all twelve reports.

Unpublished antecedents are excluded from this release. Existing public Resilient Australia and AI Help reports use different tests; neither establishes P1–P3 for this programme. What If makes its conditional reasoning explicit.

4. Keep two linked research strands

The four-day week still warrants its own trilogy: workplace feasibility, fair Australian implementation and public adoption are different questions. Individual AI access can begin before a national hours reform, and shorter hours may have value even for someone who never uses AI. These remain separately assessable strands within one What If programme.

A1 and B1 can proceed together. A2 and B2 follow the foundation review; A3 and B3 follow the relevant downstream work. This sequencing is for this programme, not a retrospective change to another series' governance.

4a. Four pathways to compare through 2035

The four-pathway comparison was incorporated on 9 September 2026. The programme uses the same definitions across A1–A3 and B1–B3. The pathways are design scenarios; they carry no assigned probabilities. WF1 is the comparator, WF2 and WF3 are distinct variations, and WF4 is the approach designed to make the proposition succeed.

WF3 is a credible option that can deliver value while the complementary working-time reform is incomplete. WF4 must explain what additional mechanisms solve that constraint. It cannot mean merely issuing a larger credit or assuming that use automatically produces better lives. Citizens/permanent residents, voluntary personal use and human final investment decisions remain the intended programme's settled principles; WF1/WF2 describe alternatives to compare, not a reversal of them.

What success in WF4 must mean

Define and assess these outcomes together, with 2035 as the horizon:

• Time actually available: reduced work is visible in actual hours and maintained ordinary weekly pay; workload and essential-service failures do not conceal the cost. Report coverage and exclusion by sector and employment arrangement.
• Usable individual opportunity: eligible people can activate and use funded AI for lawful self-chosen activity without paying first or entering a course, employer or business gateway. Record barriers, accessible support and continuity when suppliers change.
• Voluntary learning that transfers: some people use the opportunity to experiment, and demonstrate useful capability on new tasks and verification decisions beyond assisted output. Choosing leisure or declining AI remains legitimate.
• Employment resilience: examine continuity of work, adequate personal net income, options during displacement and sustainable self-employment separately from registrations, prompt counts or task speed. Show the contribution relative to the defined BAU comparator.
• Social cohesion and malicious-AI resilience: examine belonging, perceived fairness, human relationships, verification and recovery from deception, alongside overconfidence and harms. Political agreement is not the outcome measure.
• Affordable, accountable continuation: recurring funding, infrastructure, service coverage and human investment accountability can sustain the offer through 2035 and specified shocks, without a new government organisation managing the funds.

The research must set defensible coverage, cost and outcome thresholds in A2/B2/B3 before any demonstration protocol is accepted; this foundation does not invent numerical pass marks. Access delivered without credible evidence of the intended employment and cohesion outcomes is partial delivery, not full WF4 success. Conversely, do not presume BAU produces no improvements: compare observed or explicitly modelled alternatives fairly. Accepting P1–P3 as design premises does not count as measuring these outcomes.

What each report must contribute

A1 compares the time arrangements across WF1–WF4; A2 prices sector coverage and distribution for each; A3 compares the public proposal and adoption route without inventing voter thresholds. B1 compares the individual's access and learning experience; B2 costs the complete personal-funding and infrastructure offer in each pathway; B3 integrates the success criteria, outcome comparisons and One World Community action proposal. Maintain common external assumptions and apply the same shocks to all four. A2/B2 must establish a dated inventory of existing measures for WF1 before calculating incremental costs; WF1 is not a zero-cost baseline.

5. Starting implementation recommendations

Research A1 supports advancing sector-specific reduced-hours design, with unchanged ordinary weekly pay and independently visible hours, workload and service results. It does not justify a universal claim of cost-free conversion. Work already performed at night or on the nominal day off must count as work.

Research B1 develops a personal AI allowance that pays participating suppliers directly on the individual's instruction. It uses overseas examples to reduce setup and payment friction while discarding business-registration and compulsory-course conditions. China contributes both March OpenClaw policy drafts and a newer July operating-platform announcement; US child and education accounts contribute account mechanics, not proof of AI-learning effectiveness. Small direct project grants outside educational hierarchies contribute a route from personal experimentation to optional backing.

Proposed budget comparisons use A$600, A$1,200 and A$2,400 annual service ceilings in constant 2026 dollars. These are scenarios to price, not procurement quotes, approved entitlements or national cost estimates. B1 recommends taking A$1,200 as a central costing case while preserving the alternatives. Support and accessibility costs must be separately visible so they do not silently consume the opportunity of people who need more help.

6. Backcast from 2035

The following schedule is now explicitly the WF4 implementation backcast. WF1 supplies the continuing-policy comparison, WF2 shows incremental progress and WF3 shows what individual access can achieve before time reform is complete. They are alternatives, not mandatory stages. The A1/B1 pathway sections identify the additional barriers and decisions needed to reach WF4; no path receives easier external conditions solely to make its conclusion favourable.

The dates are a proposed dependency schedule. The Parliamentary Library timetable gives an ordinary simultaneous House/half-Senate window of 7 August 2027–20 May 2028; a House-only election can be earlier. There is no fixed next-election date. Prepare the proposal early enough to survive an earlier election rather than assuming a full term. A3 must distinguish representative informed support, policy salience and institutional commitment. A majority in one poll, petition signatures or reach are not a demonstrated adoption threshold.

The proposed action is an open national public campaign for a genuinely shorter four-day working week at unchanged ordinary weekly pay, seeking sufficiently broad voter support before the next Australian election to make political adoption plausible. A3 will research that public-interest proposal, its costs and trade-offs, and what evidence could support a defensible adoption threshold. No numerical “critical mass” is assumed. The proposed campaign must explain who gains time, how essential services are covered and what remains uncertain; political agreement is never a condition of personal AI access. This desk-research task does not launch the campaign.

7. Current status and unresolved choices

The programme supplies the reusable method and six substantive reports. The four-pathway amendment is incorporated. A2/B2 carry the stated age, service and project-backing recommendations as explicit working cases into the comparison. Concrete costed eligibility, funding and delivery choices remain unresolved; citizens/permanent residents, individual beneficiaries, personal learning premises and human final investment decisions remain the stated design boundary.

No administrator, appropriation, recruitment, award or campaign launch is presumed. CC BY 4.0 is the candidate reuse choice for human review, not evidence of implementation or funding. The One World Community's proposed role is to convene and scrutinise the action proposal, offer optional peer support if later resourced, and make delivery understandable. Its legal form and capacity are not assumed.

The six reports and public source register identify what this programme establishes and what remains unresolved. The What If collection explains the method and keeps the evidence, assumptions and implementation judgements separate.

Suggested citation

How this page was derived: the full substantive source body was rendered after the declared public-context transformations, removal of private delivery/workflow material and omission of the private change log.