Editor’s note: Both these stories are about the same underlying failure: mistaking the transaction for the transformation, the process for the outcome, the room full of people for the actual work of thinking.

Australia produces founders who embed their identity in their companies and institutions that embed their credibility in their process. The recovery, in both cases, requires something harder: building a generative capacity that exists independently of any single vehicle.

Innovation Spotlight: The Post-Exit Void

Papers are initialled. Hands are shaken. Somewhere between the legal formalities and the congratulatory posts, a founder's working identity ends. The exit arrives not with fanfare but with a silence that nobody in the ecosystem prepared them for.

Mike Cannon-Brookes described a version of this after Atlassian's 2015 IPO. The company he and Scott Farquhar had built from a Sydney university project into a $5.8 billion public entity did not disappear on listing day. The daily urgency did. The oscillation between existential threat and breakthrough that had structured his existence for over a decade…gone. He has been candid in subsequent interviews that the transition was stranger than the building.

The pattern is consistent. Founders describe calendars that were once packed with investor meetings, product crises, and hiring decisions sitting blank. Former employees find new roles. Investors redirect attention. Media turns to the next cohort. The founder remains: wealthy, technically successful, and no longer necessary to anything.

This is not burnout. It is a structural collapse.

Founders do not simply build companies. They build operating systems for themselves: purpose loops, decision hierarchies, accountability relationships, that run entirely on the company's existence. Remove the company and the operating system has nothing to run on.

The philosophical literature has a precise term for this entanglement. Heidegger's concept of Dasein — being-in-the-world — holds that human existence is not separate from the projects and commitments that structure it. A founder's company is not merely something they do. It is the horizon against which everything else makes sense. Every sprint, every hire, every pivot does not just move the company forward, it produces the founder as a particular kind of person. The identity is not prior to the work. The work generates the identity.

Remove the work, and you do not lose a job. You lose the context in which your self could appear at all.

What makes the post-exit void particularly difficult is that startup culture has no language for it. Accelerators prepare founders for fundraising and scaling. Nobody prepares them for the day the company belongs to someone else. Conferences celebrate exits. The aftermath is silence. Founders who struggle after a successful transaction typically believe they are failing at what should be the pinnacle of their career.

The recoveries that appear to work have a common feature: identity diversification that precedes the exit. This is more radical than it sounds. Foucault described it as the care of the self: the deliberate practice of treating your own subjectivity as something to be crafted, not passively inherited from your role. Nietzsche called it self-overcoming: the continuous act of becoming something not yet defined by what you have already done.

The founder who survives exit is the one who, somewhere amid the chaos of building, began building themselves as a separate project. Not as CEO or founder, but as a person whose generative capacity exists independently of any single enterprise.

The real transition is not leaving the company. It is learning to be generative without the machine that once made you generative, with no product roadmap, no board, and no metric to confirm it is working.

Australia's startup ecosystem is producing exits at a rate it has not previously achieved. We lack the support infrastructure that matches this growth. The silence that surrounds post-exit difficulty is not incidental. It is a design failure in an ecosystem that measures success by transaction and stops asking questions the moment the deal closes.

The Policy Lab Experiment

In late 2024, three former public servants decided to start a think tank¹ . By January 2025, the Tech Policy Design Institute had launched in Canberra with bipartisan political endorsements and a mandate to "shape technology for the benefit of humanity." Whether the institutional design can deliver on that mandate is an open question.

Zoe Hawkins (formerly the Department of Prime Minister and Cabinet), Johanna Weaver (digital transformation roles across government), and Sunita Kumar (competition policy) bring credible Canberra résumés to the project. TPDi's stated approach sets it apart from the standard Australian policy playbook: rather than developing a position and lobbying for its adoption, TPDi is betting on a different bottleneck theory. The constraint in Australian tech policy, they argue, is not a shortage of good ideas. It is a shortage of spaces where different stakeholders can develop those ideas together before positions harden.

Their first substantive publication, "Tetris for Australia's Future: aligning our national AI priorities", uses the puzzle game as a governance metaphor. The choice is either a signal of confidence in communicating to non-specialist audiences, or evidence of the accessibility trap that flattens Australian policy thinking into metaphor. The answer depends on whether it reaches decision-makers or merely readers who already agree.

The Youth Ambassadors Program and the Foundations of Tech Policy Course follow the same logic. Both aim to build shared vocabulary and frameworks rather than push predetermined positions. This is the convening model of institutional influence: change the language, and the policy follows. It works when the convener is trusted by all parties and beholden to none.

TPDi's funding structure is not fully public from available materials. Reporting from InnovationAus indicates a hybrid model drawing from both government and private sources. That structure could produce genuine independence. It could also create competing obligations that are impossible to resolve without disappointing someone. The signal will come when their research produces findings that create friction with funders' preferences.

The non-partisan positioning is similarly ambitious. In an era where technology policy is increasingly splitting along partisan lines, social media regulation and AI safety being the clearest examples, claiming neutrality requires more than stating it. It requires producing analysis that is genuinely uncomfortable to at least some parties on all sides. That test has not yet arrived.

Australia has historically imported its tech policy frameworks wholesale: Silicon Valley's libertarian instinct or Brussels' regulatory approach. Neither travels well to a country of 27 million people with a concentrated media sector, a mixed economy, and a public service culture that operates differently from either model. A genuinely Australian policy institution, grounded in domestic democratic traditions and economic realities, would be valuable.

Whether TPDi becomes that institution depends less on its founding principles and more on its institutional durability under pressure. Good process is necessary but not sufficient. The question is not whether TPDi can convene stakeholders. It is whether the convening produces positions that are actually better, not merely more consensual, than what the advocacy model would have produced without it.

¹Editor’s note: this is not a set-up to a joke, but they have commissioned a working paper on why it sounds like one.