Thread by @mrvemzi

Ribbonfarm studio’s “The Dawn of Mediocre Computing” is definitely one of @vgr’s best. Can’t wait for the next. Some of my favourite bits and takeaways. A thread 🧵

studio.ribbonfarm.com/p/the-dawn-of-…

1/n
Ventakesh Rao succinctly summarizes why achieving mediocrity - achieving human parity is difficult and important by placing the problem in the context of realish domains.

2/n A realish domain is one that is sufficiently open, structured, gamified and resourced that mediocre performance against artificially constructed but illegible criteria is sufficient for real-world survival, with better-than-wilderness odds.

3/n There are gem strewn throughout the piece, like

AIs can be used to generate “deep fakes” while cryptographic techniques can be used to reliably authenticate things against such fakery.

4/n Corollary of that,

Crypto is a target-rich environment for scammers and hackers, and machine learning can be used to audit crypto code for vulnerabilities.

5/n Back to the rigor of Realish domains definition.

Sufficiently open - Boundaries exist but externalities may impact system under consideration
sufficiently gamified - some formal definition of what is success and what is excellence

6/n Realish domains exist between truly co-operative vs adversarial and deterministic vs probabilistic axes.

7/n Stylised domains, antonym to realish domains are characterised by excellent individuals or AI looking to achieve an objectively measurable Gamified metric. Stylised domains are metaphors while realish domains are inhabited by humans. Eg., Cicero

garymarcus.substack.com/p/what-does-me…

8/n Cicero is in many ways a marvel; it has achieved by far the deepest and most extensive integration of language and action in a dynamic world of any AI system built to date. It has also succeeded in carrying out complex interactions with humans of a form not previously seen.

9/n The truth is that it is still inhabited by High performance individuals and AI agents in an open probabilistic gamified environment. Mediocre computing site on the other side of the spectrum. The target is to achieve human performance parity.

10/n What does human performance parity aka mediocre mean here ? Toil reduction from idea to code

https://twitter.com/mrvemzi/status/1598679660111724545

11/n Time to generate a machine readable file from human readable text. This is not even prompt engineering. Simple k8s YAML generation or Terraform IaC code gen.

https://twitter.com/mrvemzi/status/1598680258026541056

12/n @vgr closes off after sharing the blurbs of a discussion with #chatgpt3 #ChatGPT “useful first-dump brainstorming input.” Mediocrity in this context is actually pretty damn good.

13/n @ribbonfarm

@threadreaderapp unroll