Sam Altman is out at OpenAI. Oh, wait. He’s back.
Sam has kept people busy this week from a firing last Friday, causing shockwaves in the industry and beyond, to a subsequent rehiring yesterday.
- Sam Altman was out at OpenAI: The board of OpenAI shocked the world last Friday by suddenly announcing Sam Altman’s removal as CEO and board member. Sam founded OpenAI in 2015 with 11 others, including Elon Musk. Deep into the night on Tuesday, just four days later, Sam Altman was announced to be rehired.
- Team members, investors, and partners were shocked: Altman's dismissal led to resignations from key positions like Greg Brockman as board chair, backlash from investors, a petition signed by the vast majority of OpenAI employees threatening to walk out if Altman was not rehired and the Board removed, and potential financial challenges.
- The theory of why: The leading theory of the dismissal was that co-founder Ilya Sutskever and certain board members felt Sam’s focus was on commercializing OpenAI’s products like ChatGPT at breakneck speed – in conflict with OpenAI's mission to “ensure that artificial general intelligence benefits all of humanity.”
- The AI landscape is moving faster than ever: This event occurs against rapid advancements in AI, with generative AI technologies like ChatGPT becoming mainstream and developing at the speed of light.
- Impact on us: In the future, AI will be embedded in most knowledge workers’ workstreams. Already, ChatGPT is used by over 70% of knowledge workers, more than double the popularity of Microsoft and Google’s AI tools. Which products OpenAI develops and how they bring them to market immediately impacts our work.
OpenAI, Sam Altman, and the Future of AI at Work
Late Friday, OpenAI announced it had fired Sam Altman, its founder and CEO.
This caused a flurry of upset responses from the team, investors, and partners, who see Sam as key to winning in AI.
And finally, deep into the night on Tuesday, OpenAI announced it had rehired Sam.
As newly released FlexOS research shows, OpenAI’s ChatGPT is the most well-known AI tool, with an incredible 72% of knowledge workers having used it, increasing to 85% amongst 35-44-year-olds. This kind of awareness and adoption is massive for a technology that didn’t exist one year ago.
(Update: OpenAI is getting ready to release ChatGPT 5. Check out our guide to its new features)
The power moves inside OpenAI, and the vigorous efforts of Microsoft and others to ensure its continuity, underscore the immense power the company already has in business and the economy.
So, I've dedicated the last few days to understanding this story from its earliest beginnings and how it impacts us as leaders.
But wait, what is (Generative) AI, actually?
To form a complete understanding of the situation, let’s take a short step back and look at Generative AI itself.
Artificial Intelligence, or AI, refers to the ability of machines to simulate human intelligence, including learning and problem-solving.
AI has been around since the early 1900s, but thanks to computers, it started flourishing around 1950. Then, a young British polymath, Alan Turing, explored the possibilities of “Artificial Intelligence” in a groundbreaking paper.
In this context, we’ll also often hear the terms Large Language Models (LLMs, an algorithm that can learn through human language) and Machine Learning (ML, the ability for computers to learn without explicitly being programmed.”)
However, while groundbreaking, these AI developments largely escaped most people’s attention – they were happening 'in the background.'
But in November 2022, just a year ago, this changed.
We were introduced to ChatGPT, and with it, a new category of AI: Generative AI.
Generative AI is a group of AI-powered platforms that generate new content based on user inputs, making work easier.
They use language models and machine-learning techniques to provide comprehensive answers in human language.
Unlike previous language models, ChatGPT and similar applications are based on "Transformers" (The “T” in ChatGPT.)
Transformers allow models to understand the significance of data in a sequence, making them more intelligent than previous models.
For example, where a word is placed in a sentence can change the meaning of the sentence and tell us how important that word is.
The introduction of Transformers allows the ‘machine’ to ‘understand’ and respond in human language.
Being able to interact with AI directly is why we see such rapid adoption of Generative AI tools in our workplaces.
Another reason why AI is making its way into our workplaces is that it’s improving faster than any technology before. It keeps becoming more valuable.
Just look at the example below. An image made with Generative AI just 11 months ago is laughable compared to today's output.
Imagine what this looks like in 6-12 months. Can we distinguish it from a real photo? And what implications will that have for our jobs?
This rapid progress (velocity) at which Gen AI tools develop is why experts believe that in our lifetime, we will see machines with more intelligence than all humans combined for the first time.
This level of intelligence, also known as “Artificial General Intelligence,” or AGI, is when a machine can do any cognitive task a human can do.
As I explored in “A Jobless Future? Interrogating Musk's AI Prophecy,” this worries skeptics (and even realists.)
Because if a machine can do all our (knowledge) work, what is left for us to do?
The velocity of AI development and the ‘threat’ of reaching AGI too early or without the proper safeguards are central to this week's OpenAI saga.
(Note: AI is developing rapidly. Take a look at our six must-know AI trends in 2024.)
What is OpenAI
But again, let’s take a step back. Because what is OpenAI even?
OpenAI is an AI research ‘lab’ founded in December 2015 as a non-profit by 12 tech entrepreneurs, including Elon Musk, Sam Altman, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, and John Schulman.
OpenAI’s founding mission was to ensure that artificial general intelligence (AGI) benefits humanity.
This mission is vital to the past week’s conflict.
Because while OpenAI started as a non-profit, it introduced a “capped-profit” model five years ago.
As OpenAI shared at the time, they created the for-profit company to attract more funding while maintaining a broad focus on developing safe and widely beneficial AI:
We’ll need to invest billions of dollars in upcoming years into large-scale cloud compute, attracting and retaining talented people, and building AI supercomputers. We want to increase our ability to raise capital while still serving our mission, and no pre-existing legal structure we know of strikes the right balance. Our solution is to create OpenAI LP as a hybrid of a for-profit and nonprofit—which we are calling a “capped-profit” company.
Even at the time, this raised the question: if OpenAI is going for profits, how will it protect its mission to introduce AGI in a way that benefits humans?
In trying to find an answer, a unique board structure was implemented, in which the non-profit board controls the for-profit company.
This structure gave a small group of individuals, all without any equity in the commercial entity, the power to dismiss the leadership of its core operations.
Leadership, including Sam Altman. Moreover, Sam even helped put this structure into place, as he understood the power he'd have.
In an interview with Lex Fridman, the former and current CEO of OpenAI said:
"For sure [I worry that the power will corrupt me], I think you want decisions about this technology and who is running it to become increasingly democratic over time. We haven’t quite figured out how to do this.” – Sam Altman, Founder and CEO of OpenAI
Launching ChatGPT and the Generative AI Boom
With the new commercial entity in place and $1 billion in its pocket from now-partner Microsoft, OpenAI started significantly contributing to AI in the following years.
But its claim to fame for most of us, came in the form ChatGPT, almost exactly one year ago, on November 30, 2022.
ChatGPT leverages the “Transformer” architecture mentioned above, which allows it to understand and generate human-like text based on the input it receives.
The GPT-3.5 (and now GPT-4) model that powers it was trained on 300 billion words of content, enabling it to respond to various queries, simulate conversation, and assist with creative writing and information-gathering tasks.
Launching it as a chatbot raised eyebrows – why would you talk to a machine-learning model? Still, adoption was swift, hitting 1 million monthly users within two months, quicker than any technology before.
And ChatGPT has only grown since.
Fast-forward to this week and our Generative AI at Work research shows that ChatGPT is the most popular Generative AI tool:
- 72% of US-based knowledge workers have heard of ChatGPT. This increases to 84% Among Gen Z and Millennial Generative AI users and compares to only 33% having heard of Google Bard and 25% knowing about Microsoft Copilot.
- 60% of people actively using AI use ChatGPT at least once per week, compared to 27% using Bard and 18% using Copilot. (Which, in all fairness, was only released last week for a limited audience.)
- 93% of people using ChatGPT say that AI improves their productivity, 71% say AI has helped them develop creative and analytical skills, and 94% believe AI will positively impact their profession.
In short, ChatGPT has become an integral part of our work lives and will likely accelerate to do so.
What happened with Sam Altman?
With all that GPT success, you wouldn’t expect the company founder to be fired. But shockingly, that’s exactly what happened.
In the months leading up to last week's dismissal, concerns about Sam Altman's leadership style at OpenAI began to surface.
Some in the company felt he was going too fast in deploying AI products without testing them.
Last month, an MIT Director-led “Preparedness Team” was launched to study ‘catastrophic’ AI risks by asking the community to submit their most malicious ideas.
Late Friday, November 17, 2023, OpenAI's board of directors suddenly and highly unexpectedly announced the dismissal of CEO Sam Altman, citing a loss of confidence in his leadership and a lack of candor in communications.
OpenAI's CTO, Mira Murati, was announced as the interim CEO with a “search process underway to identify permanent successor.”
Ilya Sutskever, whom Sam Altman previously called “a legend in the field,” was reported to have had a significant role in removing Sam from his post. He was also the one to announce the firing to Sam on a hastily announced Google Meet.
The board did not share a clear explanation of the dismissal, stating only:
Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI.
Some theorize that the OpenAI reached AGI, and Sam wanted to push ahead while Ilya wanted to slow down.
Either way, the decision was met with mixed reactions in the AI community, with some supporting the board's decision and others defending Altman as a visionary leader.
Then, a series of sudden and intense events occur rapidly, causing confusion and chaos:
Friday, November 17:
- Three key researchers resign upon hearing the news about Sam Altman. Saliently, this includes Aleksander Madry, the above-mentioned MIT director responsible for the Preparedness team.
Saturday, November 18
- While OpenAI said Greg Brockman, OpenAI's president and co-founder, would stay with the company, Greg resigns.
- Significant backlash occurs from the OpenAI team, its investors, and partners. This includes Microsoft CEO Satya Nadella, who invested $10 billion into the company and was reportedly “fuming” after finding out about the firing only one minute before the public announcement.
- Then, reports circulate of board discussions about the possibility of Sam and Greg returning to OpenAI amid pressures from investors and partners.
- At the same time, other reports, including by the New York Times, speak about Sam reaching out to partners and investors to set up a new AI company, which he had been working on before his firing.
Sunday, November 19
- While OpenAI team members have heard about Sam potentially returning, they are disappointed when, late Sunday, Ilya introduces OpenAI’s new interim CEO: ex-Twitch CEO Emmett Shear.
- Emmett, critical of overly fast AI development, has ties to the effective altruism movement, which expresses reservations about advanced AI. This aligns well with the OpenAI board's desire to be cautious rather than commercially-minded.
- Hours later, it’s clear that Sam doesn’t have to register for unemployment: Microsoft, in a potential stroke of genius, announces that Sam will join their company to lead an advanced research lab.
Monday, November 20
- Ilya, the board member who fired Sam, publicly shared that he ‘deeply regretted’ his decision and is on a list of OpenAI team members threatening the board to resign.
- He later cosigns a letter with over 90% of OpenAI employees to resign unless the board does.
- Satya Nadella goes on CNBC and Bloomberg, saying that Sam Altman may return to OpenAI after all. (He also notes that he still doesn’t know why Sam was fired.)
Tuesday, November 21
- Around 4 AM, technically on Wednesday, the news is that Sam will be returning as CEO to OpenAI alongside a revised board. Sam tweets that “everything I’ve done over the past few days has been in service of keeping this team and its mission together.”
Quite a ride!
As Dan Toomey sums it up nicely in Good Work:
“Over the past few days, we've seen the people in charge of perhaps the most important technology in the world descend into a level of chaos and pettiness that would make the cast of Jersey Shore sweat.”
The Bottom Line
The recent leadership changes at OpenAI, a pioneering artificial intelligence organization, are not just industry news.
They signal potential shifts in the AI tools and technologies your company may (soon) rely on.
As our research shows, 57% of knowledge workers and 71% of those under 45 years, frequently use Generative AI tools.
As such, these developments will directly impact the tools you use to do your work. Whatever happens in AI, impacts you and your team.
Understanding how to use AI effectively in our work processes and what is happening to their development is crucial for staying competitive.
As Q Hamirani, Paper’s Chief People Officer, told me for this newsletter:
“In an era where technology and employee experience are increasingly intertwined, my focus on artificial intelligence is driven not just by its technological marvels but by its profound implications on people and leadership. As a people leader, I closely follow AI advancements to understand better how they can enhance human potential, foster innovation, and redefine how we work and lead in the digital age.”
I wholeheartedly agree and will keep you posted on the most pressing developments in (Open)AI. If you’re not yet, subscribe to Future Work here.
For more on (getting started with) AI, check out my other articles:
- Overview: AI in the Workplace
- AI in Management: How AI will change the way we manage (+37 tools to try)
- 10 AI Marketing Tools to Streamline Your Marketing Workflow
- The 10 best AI Productivity Tools to work smarter, not harder (some of my favorites)
- Best AI Accounting Tools for Improved Accuracy and Efficiency in 2024
- 6 AI Recruiting Tools for Smart and Effective Hiring
- Our top 10 Free AI Websites to Work Smarter and Faster in 2024 (for common tasks)
- Barriers to AI adoption: Why are certain companies so slow to adopt AI?
- Women and AI: Why the data shows women are at risk of falling behind.
- From Coders to Writers: An honest look at jobs AI will replace
- 18 Ways On How To Use AI at Work in 2024 (+39 Tools Reviewed)
- And another 7 Free AI Tools to Make Your Life 10x Easier