Post ATyMD2PM1dQHh5DWue by ceperez@sigmoid.social
 (DIR) More posts by ceperez@sigmoid.social
 (DIR) Post #ATyBC3WfHhc3NLxwqe by ceperez@sigmoid.social
       2023-03-25T09:21:48Z
       
       0 likes, 0 repeats
       
       1/n In the understanding of ecosystems, we should always seek to identify the innovation cycle or flywheel that exists within the system. Wardley maps are an excellent way of developing a deeper understanding of this.
       
 (DIR) Post #ATyBk4JSrfBVf90gZk by ceperez@sigmoid.social
       2023-03-25T09:27:55Z
       
       0 likes, 0 repeats
       
       2/n Any complex living system has an internal innovation cycle. There are many intertwined cycles in cognition.
       
 (DIR) Post #ATyCksI19Vm2cIHVwW by ceperez@sigmoid.social
       2023-03-25T09:39:17Z
       
       0 likes, 0 repeats
       
       3/n It's a fruitful exercise to examine the latest AI developments using a Wardley Map.  Where can we find the loops?
       
 (DIR) Post #ATyDByCRHrXQlUamUS by ceperez@sigmoid.social
       2023-03-25T09:44:11Z
       
       0 likes, 0 repeats
       
       4/n If there are no loops, then you have missing information. There should always be some process that has the potential to incrementally improve.  If you can't find one, then it's not something you can incrementally build on. There is no potential to build an advantage.
       
 (DIR) Post #ATyH6mTdqM7cmEQ5K4 by ceperez@sigmoid.social
       2023-03-25T10:28:03Z
       
       0 likes, 0 repeats
       
       5/n  What I would like to see are Wardley Maps that do have loops. I want to see one for generative AI.
       
 (DIR) Post #ATyMD2PM1dQHh5DWue by ceperez@sigmoid.social
       2023-03-25T11:25:13Z
       
       0 likes, 0 repeats
       
       6/n Another question is, how can one leverage GPT-4 to help build more informative Wardley Maps?  I actually would like to talk to a expert @swardley so I can uncover the kind of question they ask. This will inform me as to how to craft a solution. Anyone interested in chatting?
       
 (DIR) Post #ATyQ5vRBwU0ZrhwNm4 by joapen@masto.ai
       2023-03-25T12:08:42Z
       
       0 likes, 0 repeats
       
       @ceperez Hello Carlos, on these maps it's difficult to find the loopsWhy? They're product of a short research study and they only represent the present. I would also say that they need more thinking to  become a better representation of the reality.IMHO, if these maps evolve through conversations and they will probably gain a level of abstraction that looks at the reality at the right level, enabling the identification of loops.Culture map is at least 6 months map, it's quite mature.
       
 (DIR) Post #ATyRkvO9861PXXd7cu by joapen@masto.ai
       2023-03-25T12:27:22Z
       
       0 likes, 0 repeats
       
       @ceperez this is a good question.As mapper, I always struggle finding data, right just data that enable me to contrast the hypothesis we are mapping.What can LLMs do for that? A lot.I did an experiment for a manufacturing research we are doing. On this case the result is poor, but I see a lot of potential on it.https://joapen.com/blog/2023/03/23/supply-chain-director/For people with availability of large data sets, they are now in wonderland.https://joapen.com/blog/2023/03/23/supply-chain-director/
       
 (DIR) Post #AU76vx154G9D9xtgJc by gfriend@mas.to
       2023-03-29T16:46:27Z
       
       0 likes, 0 repeats
       
       @ceperez @peterkaminski