by Dennis Crouch
Berkely Center for Law & Technology is hosting a great half-day virtual-conference this week: “AI as an Inventing Tool – it’s Implications for Patent Law” organized by Prof. Rob Merges, Dr. Yuan Hao (PhD), and Prof. Colleen Chien.
- FREE, but please register: https://www.law.berkeley.edu/research/bclt/bcltevents/2023-asiaip-aitool/
- 1:00 – 4:30 PACIFIC TIME.
I’ll be there participating along with a great set of academics, government officials, law practice leaders, and tech developers. Hope to see you online!
In a recent talk about the Future of IP, I noted that I have never been comfortable with the Winslow Tableau. The basic idea from this old CCPA case is that we assume that the PHOSITA is sitting in his shop with all the relevant prior art posted to the wall.
We think the proper way to apply the 103 obviousness test to a case like this is to first picture the inventor as working in his shop with the prior art references — which he is presumed to know — hanging on the walls around him.
Application of Winslow, 365 F.2d 1017 (C.C.P.A. 1966). The Winslow Tableau presents an unrealistic conception of the PHOSITA possessing encyclopedic knowledge of all prior art. As Judge Rich described, the PHOSITA is envisioned sitting in a shop with all relevant references figuratively hanging on the walls around him, available for mental combination. While a useful thought experiment, this assumption of comprehensive awareness of the state of the art has always been out of touch with how ordinary creation truly occurs. However, recent advances in AI provide a more natural model for the hypothetical PHOSITA. Importantly, LLMs know the references and have their key points accessible for dynamic consideration and analysis.
And, although AI may be an automated system, its responses are closer to the “common sense” reasoning suggested by Chief Justice Roberts in KSR rather than the discarded “automaton” approach.
With these new models, Winslow is suddenly less ridiculous and instead much more prescient. It is then interesting to think about whether the level of skill in the art is changed by the addition of AI. One way to answer this is ‘no’ – AI does not change the level of skill in the art in any legal sense. Rather, its addition changes the potential creation methodology to no match the level of skill already required in cases like Winslow. The legal fiction is becoming reality.