The legend of the father of ChatGPT: technical genius, investment maniac, doomsday survival maniac, the next Musk
A lot has been said about ChatGPT in recent months, but this time let’s talk about its man, OpenAI CEO, Sam Altman.
About three minutes after meeting him, I remember thinking "ah, this is what Bill Gates looked like when he was 19".
In 2006, facing Altman in his early 20s, Paul Graham, co-founder of venture capital firm Y Combinator, thought to himself.
Eight years later, he himself named Altman president of Y Combinator.
Although it has only become more widely known in recent months, Altman is not an upstart in Silicon Valley. He has been in this circle for nearly 20 years.
From programmers, entrepreneurs, entrepreneurs, and angel investors, Altman may be the next star CEO in Silicon Valley after Musk and Zuckerberg.
Stanford Dropout Becomes President at 28
Altman seems to have lived the life of a technical genius.
Born to a Jewish family in 1985 to a dermatologist mother, Altman grew up in St. Louis, Missouri, got his first computer at age 8 and learned how to program and take apart Apple's Macintosh .
Similar to Jobs, Bill Gates and Zuckerberg, after two years of enrollment, Altman dropped out of computer science at Stanford University and started his entrepreneurial career with a few classmates.
In 2005, Altman's team launched Loopt, a location-based social networking application. Its valuation once reached $175 million, but it didn't spark much splash later, and was acquired in 2012 for $43.4 million.
With $5 million of his share, plus investment from billionaires, Altman raised $21 million to start a venture fund called Hydrazine Capital. Subsequently, he invested 75% of the funds in Y Combinator, a venture capital firm, and led the B-round financing of the well-known social networking site Reddit.
At this time, Altman has changed from an entrepreneur to an investor.
In 2014, the 28-year-old Altman was appointed as the president of Y Combinator by Paul Graham, the Bole mentioned at the beginning.
Altman's investment style is clear, and he is very bullish on start-up companies related to science and engineering.
Shortly after taking office, Altman wrote a blog post calling for companies in energy, biotech, artificial intelligence, robotics and more to apply for investment. He also created an angel investor program to fund at least 1,000 new businesses each year.
In 2015, Altman was named to the Forbes 30 Under 30 list of venture capitalists. Venture capitalist Marc Andreessen said of him:
At Y Combinator under Altman, ambitions rose 10x.
It was also in this year that Altman, Musk, LinkedIn co-founders and other bigwigs raised $1 billion and co-founded OpenAI. They have an almost pure mission: to develop general-purpose artificial intelligence.
Meanwhile, Altman's investments continue, and he's not afraid of big ideas, fascinated by their potential impact on the world. In 2021, Altman led a $375 million round in fusion startup Helion, his largest ever investment in a startup .
What Altman does is not a hammer in the east and a stick in the west. He has a clear plan for the future:
I think fundamentally, in today's world, the two limiting commodities that you see everywhere are intelligence and energy.
The former is trying to cooperate with artificial intelligence, which is what OpenAI is doing; the latter is betting on nuclear fusion. Altman chose the company that he thinks is the most successful, although he also knows clearly that nuclear fusion commercialization is far away.
In my experience, focus, relationships, and confidence are the way to do things in this world.
In an interview six years ago, Altman shared his experience , a mantra that has been consistent throughout his career.
Concentration and self-confidence have been reflected above. In terms of communication, at the most exaggerated time, Altman spent 6,000 minutes on the phone every month, an average of more than three hours a day. He has a stable and close circle of friends in Silicon Valley, many of whom are billionaire. He was even called "the most connected millennial in Silicon Valley" by Information.
Altman also likes to run his own blog, where he shares his thoughts intermittently, mostly around technology development, venture capital and personal career development. So, if he wants, he can publish a stack of success books.
Cautiously Optimistic Doomsday Survivalist
If Altman grew up on the trajectory of a technical genius, after rising to the top, he looks like a pragmatic technical executive.
In February of this year, Altman gave a speech on the technology podcast "Hard Fork" , bluntly saying that ChatGPT is cool, but it is a bad product, and its specific performance includes easy crashes and frequent error messages.
Fortunately, there is a lot of value in it, so people are willing to put up with the imperfections.
In fact, in December last year, Altman also reminded that ChatGPT is good enough in some aspects, but its limitations are also very strong, and it should not be relied on for any important things.
That's typical of Altman, who has always been cautiously optimistic about technology.
On the one hand, in Altman's view, the revolution caused by general artificial intelligence is unstoppable.
In March 2021, he published an article entitled "Moore's Law for Everything" , which mentioned:
To the three major technological revolutions—agriculture, industry, and computing—we add a fourth: the artificial intelligence revolution. If we manage it responsibly as a society, this revolution will generate enough wealth to give everyone what they need.
Altman, on the other hand, acknowledged the need to regulate AI.
Some time ago, after the new Bing "crazy" scared users, it was "castrated" by Microsoft. Starting from February 17, users can only ask a maximum of 50 questions per day, and can only communicate 5 times per conversation. (But on Feb. 21, the limit was eased again, increasing the number of questions per session to six, for a maximum of 60 questions per day.)
Around the same time, Altman tweeted:
Regulation will be critical and it will take time to figure out what to do. While the current generation of AI tools isn't terribly scary, I think we may not be that far away from potentially scary tools.
The risks are indeed foreseeable. Some people have used ChatGPT to forge news reports, forged AI-generated pictures into news photos, and even criminals used magic to defeat magic and wear extra prosthetic fingers to make surveillance videos look like they were synthesized by AI. It can no longer be used as "evidence in court".
Such things are hard to guard against, after all, human nature is unfathomable. A company like OpenAI alone cannot completely cure it. So Altman believes that this is a "shared responsibility."
People open source the model anyway, and it's generally good, but there are some bad things that happen. Companies that build on the model, that are directly connected to end users, are going to have to take some responsibility.
Altman's optimism and caution extend from work to life.
He loves racing cars and renting planes to fly all over California, and has a family of five cars, including two McLarens and an early Tesla.
But he is also a "doomsday preparer" who is prepared for danger in times of peace.
This group is not uncommon. In European and American countries during the Cold War, some people began to plan for survival because they were worried about the threat of nuclear war.
Today's Altman is also worried about nuclear weapons, synthetic viruses, and out-of-control artificial intelligence, so he always prepares guns, gold, potassium iodide, antibiotics, batteries, water, and military gas masks.
He also has a large piece of land in Big Sur, California, which he can fly to for refuge if necessary.
The new era of AI has begun, but he wants to see the future more
You may be wondering how Altman himself uses ChatGPT.
He often uses ChatGPT to summarize long articles and emails, and also likes to discuss programming issues with it and use it to debug code, which makes him feel "talking to a super smart programmer."
OpenAI also has another well-known generative AI tool-DALL-E 2 for text-to-image conversion. An interesting detail mentioned in the Forbes report is that Altman refused to be photographed during an interview, but drew himself with DALL-E.
But judging from many reports, Altman does not seem to care about today's AI, but about what will happen next.
For example, on the question of whether ChatGPT will replace Google , Altman seems to have said everything, but said nothing.
He believes that the current ChatGPT will not replace search engines, but one day in the future, there may be an AI system that can do it.
I think whenever someone talks about a technology being the end of other big companies, it's usually wrong. I think people forget that they can fight back here, and they're very smart, very capable. I do think search is likely to change at some point — but not as dramatically as people think in the short term.
What makes Altman even more excited is the possibility of large language models, which go far beyond the scope of search engines.
What I'm excited about with these models is that it's not "how do you replace the experience of going online and typing a search query", it's "what we do is completely different and cooler".
At the Greylock Summit last September , Altman described a future world in which AI is everywhere.
He believes that AI has become the next real technological basic platform, and it is another brand new basic platform after the mobile Internet.
In the future, a large number of AI startups will not need to start from scratch, but fine-tune the basic model trained on a large amount of data to create their own model for each vertical field. Such AI companies that provide differentiated services will "explode".
Looking back at the present, OpenAI still hasn’t made any money today, and the average single conversation between users and ChatGPT burns a few cents.
OpenAI was founded as a non-profit organization and transformed into a limited profit company in 2019. When Altman was asked how to make a profit, his answer was "There is currently no plan to make money, and we don't know how we will generate income in the future."
This is still a problem today. Recently, OpenAI launched an advanced version of ChatGPT for $20 per month, plans to make money through the developer platform , and has a deeper partnership with Microsoft-Microsoft has added ChatGPT to Bing and plans to invest billions more.
Musk, one of the initiators of OpenAI, is very dissatisfied with this partnership.
Musk pointed out that OpenAI was created as an open source, non-profit company, so the name contains "Open", the goal is to check and balance Google, but OpenAI is now a closed-source, profit-maximizing company controlled by Microsoft. This runs counter to the original intention of the establishment.
But Altman believes that, by far, Microsoft is the technology company that best aligns with their values.
Amidst the controversy, OpenAI is still moving forward.
In addition to DALL·E 2 and ChatGPT, OpenAI has other AI projects, such as OpenAI Codex, which translates natural language into code, and Whisper, which is an automatic speech recognition system.
It is also believed that the GPT-3.5 natural language model used by ChatGPT is just an appetizer, and GPT-4 is the main event.
In 2015, Altman had a public interview with Musk. He talked about the experience that Musk took him to visit SpaceX many years ago, “You know every detail of rockets and engineering technology. I don’t think many people can do this ".
At that time, OpenAI was only established for 6 months, and this interview was later described as "Kobe interviewed Jordan before becoming famous" . The details related to Musk were also written in Altman's own blog :
What I remember most is the certainty on his face when he talked about launching a large rocket to Mars.
When technology companies lay off tens of thousands of employees, giants have not launched breakthrough products for many years, CEOs and programmers have started to use AI to work , artificial intelligence has become another hot spot after Web3, and OpenAI, which opened a new era of AI, seems to be standing in the pyramid tip.
Standing at the forefront of Altman, his personal style is not as clear as Musk, but he has a world view similar to his predecessors, and he is about to become the next big man in Silicon Valley, giving a glimpse of the possible future of AI.
#Welcome to pay attention to Aifaner's official WeChat public account: Aifaner (WeChat ID: ifanr), more exciting content will be presented to you as soon as possible.