After a period of relaxation and intensive writing sessions for our upcoming book on ๐๐ซ๐๐ก๐ข๐ญ๐๐๐ญ๐ฎ๐ซ๐ ๐๐ง๐ ๐๐๐ฏ๐๐ฉ๐ฌ ๐๐จ๐ซ ๐๐ ๐๐ฒ๐ฌ๐ญ๐๐ฆ๐ฌ, I’m excited to resume regular updates. A lot has unfolded, and even more is on the horizon, so stay tuned.
Last week, I had the pleasure of speaking at the Australian Software Engineering Summer School, engaging with the next generation of software engineering researchers. My talk, an extension of my ICSE ’23 keynote, touched on several new developments:
1. ๐๐ ๐๐ ๐๐๐๐๐ญ๐ฒ ๐๐ฎ๐ฆ๐ฆ๐ข๐ญ & ๐๐ฎ๐ฌ๐ญ๐ซ๐๐ฅ๐ข๐’๐ฌ ๐๐ ๐๐ข๐ซ๐๐๐ญ๐ข๐จ๐ง๐ฌ: An overview of the UK AI Safety Summit, the Australian government’s approach to responsible and safe AI, and the new AI Safety Standard.
2. ๐๐ ๐๐จ๐๐ข๐ฅ๐จ๐ญ ๐๐ฆ๐ฉ๐๐๐ญ ๐จ๐ง ๐๐จ๐๐ ๐๐ฎ๐๐ฅ๐ข๐ญ๐ฒ ๐๐ซ๐๐ง๐๐ฌ: We touched on the recent empirical studies exploring GitHub’s code quality trends. The question arises: Is AI CoPilot enhancing productivity at the cost of quality, or are we witnessing the emergence of a new way of doing coding?
3. ๐๐ก๐ ๐๐ข๐ฌ๐ ๐จ๐ “๐๐๐๐๐ซ๐”: Imagine a future where AI doesn’t generate code but becomes the ‘ware’ itself, with billions of parameters encoding more and more business logic directly. This raises a critical question: what role will software engineers play in this new era? However, I remain optimistic about the future of this field. For those interested in pioneering this space, consider submitting your paper to the inaugural AIWare conference (https://lnkd.in/gfniE7pD) by 29 March. Join us as founding members of this groundbreaking community. I am excited to be on the steering committee.
4. ๐๐๐ฌ๐ญ๐๐ฌ๐ญ ๐๐จ๐ซ๐ค ๐ข๐ง ๐๐๐ฌ๐ฉ๐จ๐ง๐ฌ๐ข๐๐ฅ๐ ๐๐: An update on our latest work in developing a reference architecture for foundation model-based agents https://lnkd.in/gADfxpac , metrics for responsible AIs , focusing on accountability https://lnkd.in/gSG5nr7S, and copyright challenges across AI life cycle https://lnkd.in/gZEEBiJp . (all accepted at the CAIN-AI Engineering Conference 2024)
5. Summer Insights: Reflections on my learnings about DevOps, MLOps, and Architectures while writing my book over the summer.
Below are a few slides from my presentation for a glimpse into these discussions.
For our new responsible AI book, you can order on
ยทย ย ย ย ย Amazon:ย https://lnkd.in/gsQz5swy
More information and resources on the book websiteย https://lnkd.in/g9BCu6nn
CSIRO’s Data61
