Pure and Applied

Responsible AI Reader #2: who’s in charge of the AI, productivity, budget-busting

Adrian Wiggins
The HTI Corporate Leaders Survey on who’s responsible for an organisation’s AI.

Two great guides have come up in the past two weeks – AICD’s A Director’s Guide to AI Governance, co-created with UTS Human Technology Institute, and the Gradient Institute’s for everyone else, The Everyday Guide to Using AI Safely.

A Director’s Guide to AI Governance – AICD and HTI, Mon 29 Jun. “AI is reshaping how all Australian organisations operate, innovate and create value. The AICD, in partnership with the Human Technology Institute (HTI) at the University of Technology Sydney, has updated its key artificial intelligence (AI) governance resources to account for rapid AI technological change and maturing governance practices.”

The Everyday Guide to Using AI Safely – Gradient Institute, 3 Jul 2026.  “Are you unsure how to use chatbots and other tools powered by AI? This is the guide for you. Knowing how to use Artificial Intelligence (AI) tools in your everyday life is a worthwhile skill, whether you choose to use them or not. Understanding how they work helps you navigate a world where they are increasingly present.”

Australian productivity

In her talk at the 2026 Australia Oration Productivity Commissioner Danielle Wood argued there’s upside for Australians doing well with AI, but like CEDA laments how far down the leaderboard of countries Australia is currently placed. Meanwhile, there’s four research studies for the argument that if AI replaces jobs, it won’t always do good job.

Article content
The steam engine created nearly a third of the UK’s productivity growth in the 40 years to 1910.
  • Foundations for a flourishing Australia – Danielle Wood, Productivity Commission Chair Danielle Wood delivered the 2026 Australia oration in Sydney, 25 Jun 2026.  “AI is a general-purpose technology – something that can be put to a wide range of uses, from medical research to graphic design – with the potential to transform our economy. Previous general-purpose technologies – from the steam engine, to electricity, to personal computers and the internet – reconfigured economies and supercharged growth. The steam engine alone was responsible for about a third of the UK’s productivity growth in the 40 years to 1910. The ICT revolution in America contributed about half of its productivity growth in the 30 years to 2005. … AI has the potential to be just as transformative. We’ve estimated that it could add about 0.4 per cent a year to labour productivity over the next decade – which would be a substantial step up for a country averaging 0.3% over the past one. “
Article content
IMD’s technology assessment table for Australia.
  • IMD World Competitiveness Yearbook 2026 – CEDA (via BCSDA), Jun 2026 “With regard to new technologies, Australia ranked a very average 32nd for access to AI and 28th for AI skills, and only slightly better for AI investment (21st) and cybersecurity (22nd). Australia’s employers must continue to enable adoption of AI tools where appropriate, including by providing workforce training, to ensure we are not left behind. “

AI Productivity: it works best for the people losing their jobs – AI Weekly,  June 29th 2026. “Four of the most-cited workplace studies of the AI era, read side by side, tell one story: AI productivity gains are real but radically uneven, and they bend toward the inexperienced, the well-scoped, and the verifiable — not toward experts doing hard, familiar work, and not toward enterprises buying pilots.

Moore’s Law, again

Perhaps Moore’s Law will fix things…

Article content
In 2022 AI could do 30 second coding tasks. Today it’s fourteen hours.

Rutger Bregman: An Inconvenient Truth About AI, 7 Jun 2026. “Sometimes I want to shake people: have you actually used it?! Many skeptics seem to have opened ChatGPT in 2023, asked it to write a limerick, watched it fumble, and closed the tab. Well, that was three years ago. Three years in AI is a geological era. Judging today’s models by GPT-3.5 is like judging smartphones by a 2007 BlackBerry.”

A new Moore’s Law for AI agents – AI Digest, Jun 2026. “When ChatGPT came out in 2022, it could do 30 second coding tasks. Today, AI agents can autonomously do coding tasks that take humans over fourteen hours. The length of coding tasks frontier systems can complete is growing exponentially – doubling every 7 months.” Via Rutger Bregman.

From The Most Interesting Thing in AI: How AI Agents Are Changing Business – PwC’s Dan Priest with Nicholas Thompson, 17 Jun 2026. “Agents still have a ways to go in terms of consistently performing, especially long task cycles. .. [If] you gave a task that required eight hours of concentration, that agent is showing up on average across models with about two and a half hours of concentration. That means their success rate is dipping below 50 percent.” Also YouTube + Apple

Listening

Discover more from Pure and Applied

Subscribe now to keep reading and get access to the full archive.

Continue reading