A ChatGPT Guide to the Techno-Executive — Part 1 : Are we Doomed?
Introduction & Roadmap
1st, what’s up with the Techno-Exec ? You might ask. Of course, we are not talking about “TechnoKing” as in Elon Musk the “TechnoKing”, but close !!
I do want to mention that I did not use any Generative AI for writing this blog — probably it shows. The blog might be more free flowing if I had sought the help of ChatGPT !
2nd, I use ChatGPT as a moniker to the family of Generative AI as well as the larger “increasingly Multi- or General-Purpose AI” as researchers at U C Berkeley would say. This broader perspective is important …
3rd, this is “Part 1 : Are we Doomed” of a series of blogs (definitely ≤ 7, won’t go beyond that)
In the next 4 weeks, I have teed up a set of (~7) Byte-sized, Targeted, Nimble, Visual (with appropriate pictures) and Snackable (each blog, a few short pages, goes well with a few shots of a single malt, preferably triple distilled) aiming to be “usefully wrong” — like our protagonist the ChatGPT !
We might not land in Mumbai airport, but we will be “usefully wrong” and land closer, say, land the plane in the nearby parts of the Arabian sea. You will have to breakout your (waterproof) GoPros to take pictures — but they will be interesting pictures !!
We also will have an annotated reference, in the end, for more detailed exploration into selected topics
Back to our roadmap,
- In Part 2 ([6.1.23] Done ! Here), we will introduce concepts on how to think about the ChatGPT Threat Vectors & Guardrails for LLMOps. This is an interesting and relevant topic.
- In Part 2, we will also look at the canonical use cases as well as the team, talent & organization to support ChatGPT projects. It is a little different than traditional projects !
- Two weeks from now, we will look at some relevant numbers in “Part 3 : A Few Good Numbers & Current Wisdom (of the crowds)”[Here] and
- Third week will usher “Part 4 : State of the Union - Generative AI models” [Here]
- In between I wrote an interlude blog “Part 5a. ChatGPT- The Smooth Talking Stochastic Parrot” [Here]
- I have some thoughts on Transformers that might be a good “Part 5 : An Ode to a Transformer” blog.
I had conducted a Transformers class at the Nvidia GTC 2020 (Here); it is going to a few weeks to add latest developments and summarize as Part 5
- May be a “Part 6 : Explainability of the Unexplainable”.
- As 7 is the max number for this series, I will keep Part 7 in my back pocket !!
Without further ado, let us start ….
Week 1 — Part 1 : Are we Doomed ?
The short answer — “No we are not but … “ ; the caveat has a long stick, as in “… we need to pay attention — this will not take care of itself or go away”
When folks like Yuval Harari and Jeff Hinton caution us, we should pay attention.
But where is the truth ? We turn to Prof. Domingos for insights …
I liked Ethan Mollick’s LinkedIn post - organizations need to be prepared for a broad spectrum of eventualities:
There will be threats to your business model — but there are opportunities for new products and services as well ! An analysis is a must, a plan and some initiatives are recommended.
For example. lawyers and consultants are already exploring ways of using Generative AI ! And, that didn’t end well !! A lawyer got into trouble by using ChatGPT for legal research — the ChatGPT quoted non existent cases (with bogus decisions) and the judge was not happy !
Generative AI is already impacting many industries i.e. people are afraid of the implications.
Of course, as the construction firm Impact points out, ChatGPT can’t complete buildings. But the fact that it needs to be mentioned, says a lot about the collective insanity (and the threat) !
Open AI’s CEO has interesting comments at a recent hearing. They propose licensing !
Finally, a16z has an interesting take on how to think about the effects of Generative AI on an organization’s products and services. This, I think, is a gem !
This (see below) lens is very illuminating, for example traditional search tolerates the creativity of Generative AI, but search has to be redone to benefit from the emergent attributes of Generative AI
In short, there are challenges and opportunities ahead. Unfortunately inaction is not an option …
Two Parting thoughts:
1st, if AI becomes powerful we will have opposing forces as well
2nd, we can embed the responsible principles in the AI itself. As an example, Anthropic has the Constitutional AI, that can embed principles into the pipeline, very similar to Asimov’s 3 laws of Robotics !
- In short, all is not lost !
I hope we have uncovered interesting points and that these spurred more ideas … Onward to the next week’s blog “Week 2 — Part 2 : ChatGPT Threat Vectors & Guardrails for LLMOps” introducing concepts on how to think about threats and mitigation. Slightly long blog, but worth the effort (mine and yours !)
