I grew up inside machines. I studied AI when it still lived in a classroom. I have three children now, and I want this to belong to people like us.
Matt Davis. Bundaberg. As a kid I took computers apart because I could not leave them alone. At the University of the Sunshine Coast I did ICT — AI and language work included — years before a chat box was something you paid a rent on. I have been in IT ever since. The week is school, invoices, and a PC that already sits in the house. For a long time the serious tools lived somewhere else: a lab, a capital-city office, a US cloud bill. They do not have to any more. That is why I built this. I would like my kids to grow up in a country that actually uses the next layer, not one that watches it from the cheap seats.
Most of us opened a chat once and went back to the real week.
I know that week. Three children do not leave a research department in the evening. Neither does a workshop, a clinic, or anyone doing the BAS after pickup. A local model is not meant to be another thing you stay up for. It is meant to sit on the hardware you already paid for and take a job off the desk so there is a little more of you left. The figures below are firms and governments, not a survey of every Australian with a phone. They still feel familiar: we are behind countries we usually compare ourselves to, and most of the use we do have is still shallow. I would rather we caught up in our own houses than rented the future from somewhere else.
Share of businesses currently using some AI
Same questions, four countries, November 2025–January 2026. Almost 6,000 senior executives. Australia is last of the four.
- United States 78%
- United Kingdom 71%
- Germany 65%
- Australia 59%
Four-country average 69%. About 75% of those firms expect to be using AI within three years. The most common job is text generation with a large language model (41% of applications). Source: Yotzov et al., Atlanta Fed Working Paper 2026-3 (also NBER w34836).
Treasury’s own advice to the Treasurer goes further on depth: around two-thirds of Australian businesses report some adoption, and less than 10% report it as significant. Widespread, shallow. Source: ABC, 31 August 2026, reporting Treasury advice to Jim Chalmers.
Where central governments actually put AI
OECD Digital Government Index, 2025 window. This is whether a country’s central government reports using AI in that job — not how well, and not every council. The United States and Germany did not file 2025 answers in this table.
Across the OECD: 31 of 36 countries (86%) use AI for internal processes; 27 of 36 (75%) for public services; only 13 of 36 (36%) for policymaking; 12 of 36 (33%) for oversight.
| Country | Internal work | Public services | Policymaking | Oversight |
|---|---|---|---|---|
| Australia | Yes | Yes | No | No |
| United Kingdom | Yes | Yes | Yes | No |
| New Zealand | Yes | Yes | Yes | No |
| Canada | Yes | Yes | No | No |
| Estonia | Yes | Yes | Yes | Yes |
| Korea | Yes | Yes | Yes | Yes |
Source: OECD Digital Government Outlook 2026, Annex Table 4.A.1 (Survey on Digital Government 3.0). Australia’s public-service AI plan also says gen-AI uptake across agencies is inconsistent — strategy on paper, uneven on the desk. APS AI Plan 2025.
What working with me looks like
- Straight answers, including “you don’t need this”
- A fixed quote in plain English
- Written down so you are not locked in
- The same stack whether it is a house or a company with sites — I will not publish the names of the machines in mine