Dustin Moskovitz, the CEO of Asana and co-founder of Facebook, has been concerned for some time about the potential hazards of out-of-control artificial intelligence. He has acquired a large stake in Asana and is now its majority owner; additionally, he is set on donating the majority of his wealth to philanthropic initiatives, with 'dangers from sophisticated AI' as one of the main focus points.
Dustin Moskovitz deviated from the typical success story of a tech founder, in which one sells off significant chunks of their company to venture investors, eventually becoming a minority shareholder before taking it public and then selling more stock. Instead, Moskovitz retained a majority stake in Asana which he took public in 2020, with his ownership standing at 36%. He proceeded to purchase 480,000 more shares of the company, pushing his total ownership up to 51%. Subsequently, Moskovitz announced his intention to buy up to 30 million more Class A shares in 2021, causing the stock to surge by nearly 19%.
Moskovitz, whose net worth exceeds $12 billion primarily due to his early Facebook stake, doesn't see his majority ownership of Asana as an act of control. Instead, it is a means of supporting his philanthropy. Open Philanthropy, founded by Moskovitz and his wife, Cari Tuna, has donated over $300 million to focus areas such as artificial intelligence, with a particular emphasis on safety and governance. Moskovitz has also contributed to OpenAI, which is now valued at $30 billion.
Though Moskovitz has the means to purchase all shares of Asana, he has agreed to keep a majority of its directors independent, per the rules of the New York Stock Exchange. He declined to comment on speculation that he is buying up shares to keep activists from coming in and forcing change, as has been done with Salesforce.
Recently, Moskovitz experienced a convergence of his two worlds when OpenAI soared from a small startup to a widely acclaimed tech company following the release of ChatGPT in November. Before that, Moskovitz had been experimenting with OpenAI's DALL-E technology which is for converting text into images. He mentioned that OpenAI CEO Sam Altman had granted him access to the startup's "labs account" already in April of 2020. Following the launch of ChatGPT, Moskovitz took the chatbot for a spin to explore solutions to California's housing shortage. Asana then joined the list of companies that utilized generative AI features that are able to yield text, visuals, and audio in response to human input. In the beginning of June, Asana declared that they had already given a few clients access to a range of AI generated features powered by OpenAI's models.
Moskovitz mentioned that "chat is just one way of deploying these technologies" and that there are more complex applications such as optimizing automation processes or aiding in making decisions. He explained that the strength of the technology lies in its ability to grasp a vast amount of information and an ambiguous goal and then produce something that matches the original purpose. Asana estimates to spend around $5 million or even more on OpenAI's technology next year. This past June, Asana's earnings call included a six-minute-long presentation that Moskovitz made on how they use AI in their system. This was more than Microsoft CEO Satya Nadella's use of the acronym in the company's earnings call in April. Microsoft is OpenAI's leading investor.
The relationship between Asana and OpenAI is much more than just business-related. In 2016, Altman had already invested in Asana and during the June earnings call, Moskovitz reminded their audience that Asana and OpenAI had "the same board member in Adam D'Angelo," a former Facebook tech chief who then built Quora. One of OpenAI's first board members was Holden Karnofsky who is the co-CEO of Open Philanthropy and later co-founded Anthropic with his wife, Daniela Amodei. Moskovitz also invested in the company in 2021, in the same year he co-invested with Altman in Helion, a nuclear fusion startup.
Moskovitz is highly optimistic about AI yet aware of the potential dangers it poses. In May, he was among the entrepreneurs who issued a warning that mitigating dangers that come with AI has to be regarded as a top priority, in the same way as nuclear war or pandemics. However, he wasn't a part of the letter that was sent by the Future of Life Institute in March which asked AI labs to take a break from training cutting-edge AI models for a minimum of six months. Elon Musk who is one of the early supporters of OpenAI and has warned about the risks of advanced AI was one of the signatories. Moskovitz declared that although he and Musk don't necessarily share the same perspective on AI censorship and 'wokeism', they both want to bring this technology to the world in a safe way.
Concurrently, Moskovitz is crafting a 12-point list of potential policy changes for the American legislators to consider. He particularly wants to make sure that all the later and advanced generations of AI technology, such as GPT-5 and GPT-6, are tested to ensure their safety prior to their launch in the world. To express his complex views, he even coined a new word: 'excito-nervous' for AI.
Comments