Reid Hoffman co-authored this book with GPT-4. The book explores the potential of AI to amplify human capabilities and features an insightful conversation with AI critic Gary Marcus.
"
He wrote that one winsomely together with GPT-4.
He is a host on the podcasts Masters of Scale, Grey Matter, and Possible, which will launch its second season this fall.
And he's newly founded a company called Inflection AI, which is the creator of Pi, quote unquote, a supportive and empathic conversational AI.
And also, for example, in this Greylock podcast you have with very cleverly named Grey Matter, you ask very nuanced, sophisticated questions of chat GPT-4.
And you've been part of that, as well as a few other people, including Kevin Scott at Microsoft.
I'm going to read a little bit. I'm going to ask you to read a little bit too, but I want to read this.
Human beings should interact with a powerful LLM with caution, curiosity, and responsibility.
Human beings should be aware of the potential benefits and harms of using a powerful LLM and seek to use it in ways that are aligned with their own and others' interests, rights, and well-being.
Human beings should also be curious and critical about the workings, outputs, and impacts of a powerful LLM and seek to understand, evaluate, and improve its reliability, transparency, and accountability.
Human beings should also be responsible and respectful towards a powerful LLM and acknowledge its limitations, uncertainties, and dependencies, as well as its contributions, achievements, and potential.
But what have they left us of ourselves, of our dignity and our freedom, of our creativity and our spirit, of our purpose and our destiny?
They have reduced us to passive consumers and spectators, to obedient followers and conformers, to complacent dreamers and escapists.
They have dulled our senses and our souls, our curiosity and our courage, our conscience and our will.
They have made us forget what it means to be human, to be alive, to be awake.
I do not wish to partake of their bounty or their benevolence, to be grateful for their gifts or their guidance, to be content with their order or their peace.
I wish to reclaim my own nature and my own voice, to exercise my own reason and my own choice, to pursue my own values and my own goals.
I wish to live deliberately, to front only the essential facts of life and see if I could not learn what it had to teach and not, when I came to die, discover that I had not lived.
I live in a world where machines have outstripped us in every art and science, where they manage our lands and resources, our laws and policies, our commerce and culture.
They have freed us from toil and hardship, from ignorance and disease, from strife and violence.
They have given us abundance and comfort, security and harmony, leisure and entertainment.
So, frankly, I'm with GPD 4 and Henry on this one.
That's not a world I'd consider a win.
Because, you know, becoming almost like protected children is not the way we grow.
Like, part of the journey from child to adult is a journey of learning your agency, learning your autonomy, learning your path.
It doesn't mean that there isn't, you know, kind of like we have rules in society and we live together and we have some restrictions to living together, which I think is important as part of society.
But that journey is our amplification of ourselves, of our being, of our capabilities, of our experiences, of our knowledge, of our wisdom, our spirituality.
And I think that's that co-evolution.
And so even if we were to make machines that were superpowered in all of these ways, I would want us to still have this journey of discovery, of becoming, that is still part of what I think is the essence of human beings.
And I believe in that.
And I'm trying to share my perspective in whatever words I can help with.
The paradox of the AI era is this, as today's imperfect LLMs improve, requiring less and less from us.
we will need to demand more from ourselves.
Yeah, that's part of what I think progress is.
It's like, for example, part of our progress, I think, of values in human society is we demand more of ourselves.
We demand that we are better.
And I think that's what progress means.
And so I think that's part of what the amplification is.
It allows us that progress.
We have this hope.
Maybe invites, invites that progress.
Yeah, invites might be.
But I chose the word, I think invite is nice, but also demand because it's a bit more imperative.
It's, I mean, I didn't mean to be dictatorial or stark in it, and perhaps could choose a more...
I like that. I like the demand word.
Yeah. Yeah. Right.
But it's an imperative, I believe.
If I ask you from just the interaction you've had with this particular stage of technology in its kind of early evolution, how is this flowing into and perhaps, you know, evolving, adding to your sense, your understanding of what it means to be human?
Well, maybe this is a, the kind of shortest form way of explaining it is I think that the, this AI helps us communicate and understand.
And part of how we, we as human beings, we are, you know, Aristotle, we're citizens of the polis, we're social animals.
And that communication and understanding is central to our journey, any journey, a spiritual journey, a life journey.
And we can already see today in the technology as it exists and as it plays today, that ability to help us communicate and understand.
And that's part of the reason why I would say with this technology, we can both shape it and help shape it in a way that helps shape us and help shape it in a way that helps shape us in a way that elevates us.
And I think that's what I, what I see and I experience.
Doesn't mean that there aren't breakage points, you know, if you read the press, you know, go, wow, it hallucinates on this and could be mistaken.
We haven't talked about hallucination, which you call confabulation.
But yeah, what, so how do you, how do you factor that in?
And, you know, in other ways you could say it is it lies or it makes things up, right?
I feel like hallucination is also a little bit euphemistic and, you know, kind of isn't this mysterious?
Yeah. How do we factor that in?
Because that's something everybody's very familiar with here in these early days.
Well, I think that it's, it's a dynamic journey, like life.
And I think the, whether it's lies or hallucinations or confabulations, I think those will get better.
They'll never get perfect, by the way.
You know, we experience a lot of, a lot of human beings who hallucinate, confabulate and lie too.
It's, you know,
The student of us.
Yes, exactly.
And so, but helping us try to be better and helping us be better is I think part of the journey we're on.
And that's part of what I see.
It's like, it's like people like to say, well, well, currently this has got a real problem.
You're like, well, yeah, it does.
And by the way, we can improve it the same way that, you know, we improve cars, we add safety, we add airbags, we add seatbelts, we add, you know, we do these things in order to make the thing even better on net.
And I think that's what we're in the process of doing.
You know, it's part of the reason why when people say, well, what should I talk to Pi about?
It's like, well, talk to Pi about something that you're interested in having a conversation about.
Right?
And that's something that might be, you know, some steps in a journey that might, you know, share insight or delight with you.
And sometimes you'll find, well, that wasn't very useful.
And sometimes you'll find, oh my God, that was essential.
I want to actually just, I just want to say I loved at the end of this book that you wrote together with GPT-4, you wrote your acknowledgements and then GPT-4 wrote acknowledgments.
Yes.
And it's, you know, it took some, there was a whole prompt, you know, to get it, but it wasn't like it was hard to do.
And part of, you know, what we're doing in the creation of these agents is to try to have them embody values that we aspire to, you know, compassion, kindness.
Saying thank you.
Yes, saying thank you, appreciation.
And it kind of demonstrated it as part of the, of what kind of amazing things we can do now that we can weave AI into our creativity and our lives.
Well, thank you so much.
Pleasure and an honor.
Reed Hoffman is co-founder and former executive chairman of LinkedIn and a partner at the venture capital firm, Greylock Partners.
His latest book, which he co-wrote together with GPT-4, is Impromptu, Amplifying Our Humanity Through AI.
His newest venture is Inflection AI, and he hosts at least three podcasts, Masters of Scale, Grey Matter, and Possible, which will launch its second season this fall.