A staple of startup literature, advocating a deceivingly simple concept which is hard to get right (as is proven by the examples of startups that have failed since publication).
Re-read 2024: Even though some examples are by now pretty stale, there are still many relevant insights in there.
A brave attempt to put up a framework for assessing technological innovations, that is rich of ideas, which are in many cases [in 2023] still relevant (e.g. Cognifying in the light of GenAI), but sometimes feel out-dated (e.g. Sharing is a post-truth world).
Fun fact: Meta unleashed its Llama2 (even though there are questions on its openness) just before committing to protecting proprietary and unreleased model weights.
In any case; the USA has a totally different approach from the EU with their AI Act. These commitments provide a great opportunity to do an early check on how self-regulation in AI could shape-up.
There are three observations that stand out.
Vagueness
It has already been observed that most of said ‘commitments’ made by big tech are vague, generally non-committal, or confirmation of current practices.
Considering the success of the EU in getting big tech to change (e.g. GDPR, USB-C) I am convinced that in tech, strong legislation does not stifle creativity and innovation; but fuels it.
Data void
There are also notable omissions. The one that sticks out for me is the lack of commitment with respect to training data. And that at a moment that legal cases over data theft and copyright infringement are popping up in various places. In that context, Getty Images hopes that training on licensed content will become a thing.
Admittedly, discussions on data ownership are super interesting. But full clarity on the data going into foundational models (and the policies around it) would also sharpen the extent to which data biases may put model fairness and ethics at risk.
Content validation
By far the most interesting commitment is around identification of AI-generated content:
“The companies commit to developing robust technical mechanisms to ensure that users know when content is AI generated, such as a watermarking system“
Considering the expected amount of generated content, I expect not watermarking of AI-generated content (the vast majority of future data volumes) will be problematic.
And it also addresses the problem from the wrong side. In the future, the question will not be “What is fake?”, but rather “What is real?”
This points in the direction of watermarking of human-produced content to be the way forward. Think of an NFT for every photo you make with your smartphone of digital camera. I didn´t hear Apple, Samsung, or Nikon about this yet. But I wouldn´t be surprised if we see announcements in the near future.
Nov. 2017: Interesting exploration of the implications of AGI, faulted by the typical preference of Analytical Philosophy for construction of intricate, highly theoretical scenario’s, under-emphasizing basic challenges (in the case of AGI: lack of robustness / antifragility).
Jun. 2023: The writer has leveraged the recent rise of LLMs like ChatGPT to further fuel fear about an AGI break-out – even though other AI-related risks require more imminent attention.
It would be worthwhile research topic to map the development of ship building to the principles of disruptive innovation as laid out by Clayton Christensen.
The book’s set-up with multiple scenarios for the future works surprisingly well and is especiall concerning for European readers: Europe is almost completely irrelevant in all of Webb’s scenarios.
The authors see AI as just a new option for the division of labor which, although it can have rather dramatic consequences, does not support apocalyptic GAI fearmongering.
Great effort to democratize AI and peel off some layers of mistique that harm public debate (althought the case against technochauvinism seems at times a bit too shallow).
The history of disc drives and mechanical excavators showcases how difficult it is for incumbents to come out on top when technological innovation hits your market.