Struggle to keep AI research as non- profit

Struggle to keep AI research as non- profit

Updated: Mar 22, 2019



The San Francisco based highly funded research organization OpenAI has a mission to ensure that artificial general intelligence (AGI) benefits all of humanity, primarily by attempting to build safe AGI and share the benefits with the world. The mission and the talent pool has attracted great fans and prominent backers such as Elon Musk, Peter Thiel, Reid Hoffman who have pledged over US$1 billion to support OpenAI’s ambitious plans.


But, as many other such moonshot initiatives have come to learn, OpenAI too has had to come to terms with the financial realities of running an organization. This is heightened by the amount of cash burn at OpenAI whose recent project included OpenAI Five for bots trained to compete against humans in Dota 2 required renting multiple high-performance CPUs and GPUs for training. Also, with the great hype around AI, the cost of good AI engineers and scientists has shot up multiple times. The talent crunch is such that PhDs from reputed institutes can draw salaries exceeding a million dollars annually. As OpenAI employs some of the tops AI talents, the annual employee cost itself would run in 10s of millions of dollars as it employs over 100 pricey talents. Considering its ambitious plans of research, even the billion dollar cash chest seems to be running out for OpenAI and getting new funding has become essential.


So, it was not much of a surprise that OpenAI came up with OpenAI LP, a new ‘capped profit’ company which is a hybrid of non-profit and for-profit where the profits for investors would be capped to 100x. The remaining profit if successful would go the OpenAI non-profit organization. OpenAI’s blog on 11th March 2019, goes in detail about this new company while elaborates heavily on how the mission of OpenAI is still the paramount consideration for decision making. The blog claims that the overall mission of OpenAI LP would be ensuring the creation and adoption of safe and beneficial AI and this would be kept ahead of generating returns for the investors. To achieve this, OpenAI LP would be controlled by OpenAI Non-profit’s board and all investors and employees have to sign agreements in this regard that the obligation of OpenAI LP to the Charter would always come first. The blog also states that “Only a minority of board members are allowed to hold financial stakes in the partnership at one time. Furthermore, only board members without such stakes can vote on decisions where the interests of limited partners and OpenAI Nonprofit’s mission may conflict—including any decisions about making payouts to investors and employees.”


OpenAI is just the new prominent moonshot research organization to face this. Google’s Deepmind also had raised alarm bells amongst the finance people at Google’s parent company Alphabet. Deepmind’s pre-tax loss in 2017 was US$ 368 million, a significant jump from a loss of US$ 164 million in 2016. Deepmind was made into a separate entity in 2015 in the restructuring exercise at Alphabet, a move targeted to calm the Wall Street. While with the Alphabet’s cash flow, Deepmind is able to survive, it would be interesting to see how long will its investors be patient about this? A similar moonshot technology project in advanced robotics, Boston Dynamics was earlier sold to Softbank by Alphabet to cut down the losses on its P&L. Some other Google X projects were shut down as well. Even Google’s much talked about self-driving car projects recently announced partnerships to commercialize the services for grocery delivery or in some cases ride-hailing applications. It also announced that it would even sell its Lidars for non-competing applications. The pressure to monetize is clearly visible in all these cases.


It is important to note that in spite of idealistic vision about the future of technology, financial realities have to be faced sooner or later by most projects. It would be interesting to see the future of OpenAI as one reason it attracted the top talent in AI was its non-profit, open source nature. The long blog post by OpenAI explaining its primary focus to stick to its mission along with forming OpenAI LP is indicative of this struggle to manage open source research while balancing the financial realities.

To deep dive and stay continuously updated about the most recent global innovations in Artificial Intelligence and learn more about applications in your industry, test drive WhatNext now!


Image Courtesy : www.datarobot.com

+44 20 3478 7711

5 Chancery Lane, London, EC4A 1BL, United Kingdom

Priva

©2020 by WhatNext Global plc.