Author: Reid Hoffman

My eight years on the OpenAI board

This post is by Reid Hoffman from Reid Hoffman

AI, like most transformative technologies, grows gradually, then arrives suddenly. Headlines make AI feel abrupt and singular when it’s compared to a tidal wave, revolution or tectonic shift. In actuality, foundational work in AI has been going on since the 1950s. For more than a decade, Greylock has invested in AI startups, such as CrestaSnorkel AIAdept AI Labs, and Inflection AI, one that I personally co-founded. OpenAI started in 2015—the same year I invested and joined its board.

With all that’s been built over decades, the landscape has changed drastically over the last few years. Personally, I haven’t been this excited since I first started understanding the power of online social platforms. There’s so much afoot: The mainstream adoption of AI is happening now. Advancements like LLMs are amplifying the abilities and productivity of humans. There are future trillion-dollar companies being built and invested in right now, which will not only change markets, but launch new ones.

With all the possibilities ahead, it’s important to think beyond any one application, company, industry when it comes to AI, because I believe its impact will be on a much greater magnitude.

It’s with that in mind, that I’ve decided to step back from the OpenAI board. I made this decision after months of conversations with Sam, Greylock colleagues, and friends. There are a few reasons:

As OpenAI’s APIs become more important to the next wave of AI applications, Greylock and I will be investing in companies, like Tome and  (Read more...)

AI’s AHA Moment

This post is by Reid Hoffman from Reid Hoffman

We are all collectively experiencing an aha! moment—that flash of discovery, insight or understanding—with the recent developments in AI, such as OpenAI’s ChatGPT.

Of course, an “aha” moment is a bit longer in the tooth than AI. According to the O.E.D, “aha” as we know it goes back at least to the 1390s, when English author Geoffrey Chaucer wrote in the fable “The Nun’s Priest’s Tale”: “They crieden, out! A ha the fox! and after him thay ran”

Today, it’s AI rather than a fox that has us exclaiming, but the aha feeling is there—and maybe even more powerful. It’s not just a passing moment of discovery, but possibly a peek into something more enduring: how this technology can augment our abilities in a lasting way.

With AI, our AHA moment stands for something bigger: it’s an Augmentation of Human Ability. These AI tool advancements make us more productive, creative, and efficient. They show us more deeply and swiftly who we are and can be.

What do you think about this AHA (Augmentation of Human Ability) term as capturing what AI can do for humans and humanity? Let me know your thoughts in the comments below.

Amplified Intelligence

This post is by Reid Hoffman from Reid Hoffman

The magic on display in the growing family tree of LLMs—before and beyond OpenAI’s ChatGPT—is astonishing. In these early stages, we’ll continue to collectively find our footing with AI, but the path is there to reconfigure how we work and live.

That term itself—AI—has also been on my mind. As I’ve been so focused in recent years on AI projects, I’ve found myself questioning whether using that moniker—”Artificial Intelligence”—may actually be holding us back in some ways.

To start: has anything been celebrated—or preferred—for being artificial? Artificial food coloring. Artificial flowers. Artificial respiration. If so, it feels like a rare exception.

I believe the current language of Artificial Intelligence has somewhat misled us and, in the wider culture, painted an ominous view of an AI that develops sinister intent at odds with our natural intelligence and natural interests. 

To me, there’s a more fitting descriptor of what’s taking shape that better captures these tools’ impact on humanity: Amplified Intelligence.

Same acronym, but more accurate. Advancements like LLMs are amplifying the abilities and productivity of humans. And, yes, it will amplify our mistakes and misunderstandings as much as our achievements and insights, but that just reflects and accelerates our growth as humans.  

What do you think about Amplified Intelligence (AI) as a term? Let me know your thoughts in the comments below.

Technology Makes Us More Human

This post is by Reid Hoffman from Reid Hoffman

ChatGPT, a new AI system that sounds so human in conversations that it could host its own podcast, is a test of temperament. Reading between its instantly generated, flawlessly grammatical lines, people see wildly different visions of the future.

For some, ChatGPT promises to revolutionize the way we search for information, draft articles, write software code, and create business plans. When they use ChatGPT, they see Star Trek: a future in which opportunities for personal fulfillment are as large as the universe itself.

Others see only massive job displacement and a profound loss of agency, as we hand off creative processes that were once the domain of humans to machines. When they use ChatGPT, they see Black Mirror: a future in which technological innovation primarily exists to annoy, humiliate, terrify, and, most of all, dehumanize humanity.

I’m firmly in the Star Trek camp, because although I fully acknowledge that the tech industry is imperfect, and always in need of thoughtful, responsive leadership, I still believe that improvement through technology is how humanity most effectively makes progress. (In the interest of full disclosure, I should note that I sit on the board of the OpenAI nonprofit, and that, through my foundation, I have invested in OpenAI.)

That’s why I switched from a planned career in academia to one in Silicon Valley in the first place. In the early 1990s, I saw how software, globally distributed on the internet, was creating new opportunities to empower people at scale, and that’s (Read more...)

Talking AI with AI

This post is by Reid Hoffman from Reid Hoffman

Today we’re launching Fireside Chatbots, a special podcast miniseries running as part of Greymatter’s Intelligent Future series focused on artificial intelligence.

In this new series, I’m going to be talking about AI and the massive positive impact I believe it’s going to have on all facets of society.

But I’m not just going to be talking about AI – I’m going to be talking with AI.

In this podcast, my guest is ChatGPT, a large language model created by OpenAI, an organization whose Board of Directors I serve on.

ChatGPT is not human and not conscious. It does not reason like we reason, but there are interesting parallels between how it works and we work, since it is trained on a huge amount of human data from the internet.

And, ChatGPT does have an uncanny ability to speak coherently and informatively on a huge range of subjects, because it’s been trained on a massive corpus of text from the internet, then fine-tuned on examples of human dialogue.

Over the last decade or so, we’ve seen an amazing resurgence in Artificial Intelligence. A lot of it started with DeepMind and the neural networks it trained to play classic arcade games like Pong and Breakout, and then its development of even more powerful algorithms like AlphaZero, which it created to master far more complex games like Go and Chess.

As DeepMind has continued to contribute many additional stunning results, including AlphaFold for protein folding, other organizations started making incredible progress too. (Read more...)

You’re more ready than you think

This post is by Reid Hoffman from Reid Hoffman

“I don’t feel ready.”

 That’s what I hear again and again from people starting—or restarting—their careers.

 But I have good news: you’re more ready than you think. 

 I’m not saying this just to make you feel better. It’s true.

You’re more adaptive than you think. If you don’t have enough relevant work experience, you may think that you don’t have much to offer. But you’ll be surprised how valuable your existing skills and connections are to those who don’t have them.

You can take more risks than you think. Rather than fleeing risk, you should learn to take intelligent risks. An intelligent risk is an opportunity with enough upside to justify the possible downside. You’ll find an opportunity that others miss, if others perceive something as riskier than it is.

Your network is bigger than you think. You may think your network is small or that you don’t know the right people. But if you’re connected to a couple hundred people on LinkedIn, you’re actually at the center of an extended network more than 2 million people strong. 

You’re more ready than you think not because you’ve learned everything you need to know, but because you’re likely underestimating what you already have. 

You’re more ready than you think

Watch & listen:

Watch the narrated story in full screen ↗

How to Authentically Manage Your Personal Brand

This post is by Reid Hoffman from Reid Hoffman

Your personal brand at work is not just what you say about yourself. It’s also what your network says about you.

While this isn’t something you can fully control, there are ways to better understand and shape it. 

On the seventh episode of the Startup of You Podcast, Ben Casnocha and I discuss how to more effectively build and leverage your personal brand at work and across your career. 

Tune in to hear about the three components of your personal brand in professional contexts, as well as specific tactics that you can use to manage it:

How to Take Smart Risks When Making Career Moves

This post is by Reid Hoffman from Reid Hoffman

One of the riskiest things you can do in your career is to try to eliminate risk entirely. 

In fact, it’s impossible—and unadvisable—to avoid risk. For example, if you’re not aware of the risk involved in a potential career move, it’s probably not a breakout opportunity.   

The world changes—as do our jobs, companies, and industries. Risk is omnipresent. An essential skill is to know how to navigate it. 

On the sixth episode of the Startup of You Podcast, Ben Casnocha and I discuss how to conceptualize, navigate, and endure risk.

Listen here:

Don’t Make Decisions Alone

This post is by Reid Hoffman from Reid Hoffman

After a wave of early customer feedback, LinkedIn shifted its primary product development focus from individual to corporate subscriptions. 

One problem: I had never built a corporate product. 

I didn’t know who could best advise us when I called David Sze at Greylock. He introduced us to Workday CEO Aneel Bhusri, who connected us to Mike Stankey on his team. Mike had the exact skills and knowledge to help us. 

My network helped me pinpoint what I needed, and guided me to an expert better than I could have managed alone. 

There will be many times when you won’t have the answer or the expertise, but your network might. Network intelligence is not only the collective know-how of your network, but also how you access and integrate its wisdom into your decision-making process.  

On the fifth episode of the Startup of You Podcast, Ben Casnocha and I cover:   

  • Why you should see—and map—your network as a sensor network
  • What network literacy is, why it’s important, and how to develop it
  • How to conduct better reference checks—and with whom

Listen here: 

If you like what you hear, please subscribe and let us know your thoughts and questions in the comments below. 

Uniquity: DALL-E, NFTs, and the emergence of limited-edition abundance

This post is by Reid Hoffman from Reid Hoffman

When you use DALL-E, a new AI system that generates images from natural language descriptions, words become your pencils, paints, brushes, cameras, lenses, chisels, and more.

In the DALL-E universe, though, there’s never just one way to translate written language into graphic language. Inverting the old proverb on its head, a single word – or more accurately, a single phrase or premise – can be the starting point for a thousand pictures. 

Each time you enter a text prompt, DALL-E generates four different images in a matter of seconds. Here, for example, is the initial set of images that DALL-E generated when I entered “A subway train on Mars”:

No alt text provided for this image

If you don’t see anything you like, you can just hit the GENERATE button again, and DALL-E will produce four new images:

No alt text provided for this image

You can also modify DALL-E’s output by modifying your text prompt. Here’s what DALL-E produced when I entered “A 3D render of a subway train on Mars”:

No alt text provided for this image

Here’s the results for “A panoramic 3D render of a bullet train speeding through the Mars desert as spaceships circle overhead”:

No alt text provided for this image

With DALL-E, you can literally produce hundreds of unique images within minutes. As I experienced this amazing generativity first-hand, I quickly began to think in terms of multiples and scale. What kinds of books or other projects, I wondered, where you might need dozens, or hundreds, or thousands of images, could you use DALL-E for?

One of the first ideas my team and I hit upon was the idea of an (Read more...)