Reimagining Investing in Frontier Technology

Summary Report

Event Summary

Harvard Kennedy School’s Belfer Center for Science and International Affairs’  Technology and Public Purpose Project (TAPP) and Harvard Business School’s Arthur Rock Center for Entrepreneurship co-hosted Reimagining Investing in Frontier Technology on May 15, 2019. This workshop convened over 70 investors (Limited Partners and General Partners), entrepreneurs, technologists, and others investing in and building frontier technologies in areas including artificial intelligence, genome engineering, advanced computing technologies, and more. The workshop explored the challenges investors and entrepreneurs face in bringing products to market in ways that maximize their benefits to society while minimizing harms.

Opening Remarks

Tom Eisenmann, Faculty co-Director of the Arthur Rock Center for Entrepreneurship, opened the event with an interactive exercise in which participants were asked to reflect on nine facts about venture capital and frontier technologies (See Figure 1). Several aspects of venture capital investing make it challenging to incorporate considerations for ethics and public purpose. The failure rate among start-ups is very high, and fund returns are earned from a small percentage of companies. This leads to immense pressure on entrepreneurs to build, scale, and attain returns quickly. Investments are raised over several rounds with different VCs, and senior leadership teams in start-ups can turn over quickly. This turnover in capital and leadership leads to challenges in mission continuity and in accountability. At the same time, venture capital-backed startups also have opportunities to better integrate ethics and public purpose considerations. Partners sit on company boards and provide active management and advice, which is an opportunity to provide direction and support. Among frontier technology ventures, initial funding often comes from governments and technologies are often licensed via university technology transfer offices, both of which provide opportunities for additional checkpoints to consider societal impacts.

Former U.S. Secretary of Defense and Belfer Center Director Ash Carter gave an address on the challenges of the next wave of technological development. He noted that the stakes will be much higher for society as AI moves from applications in ads and enterprise software to making decisions in health care, military, policing, and justice administration. The seriousness of these implications means that technologists, entrepreneurs, and investors must be more cautious going forward in ensuring that AI systems ‘get it right’ and are explainable and accountable to those impacted. At the same time, the coming wave of biotechnological innovation will also bring new societal challenges. The biotechnology field historically has been a relatively slow-paced field, led by scientists with PhDs, high regulatory oversight, and long product development timelines. The next wave will be characterized by fast-paced investment as the field intersects more with AI and digital technology developers. Community biolabs are also opening access to a new type of biotech entrepreneur without advanced scientific training.

Investing in Frontier Tech: Opportunities and Responsibilities

The first session, led by TAPP Inaugural Fellow Susan Winterberg, focused on the roles of investors in shaping development of start-ups and on the current state of responsible investment practices. Rudina Seseri, a Founder and Partner at Glasswing Ventures, spoke about evaluating bias in machine learning algorithms and how to encourage greater diversity on boards and product development teams in portfolio companies. Peter Barrett, a Partner at Atlas Venture, spoke about investing in biotechnologies, the value of having a regulatory body like FDA to help evaluate risk and quality of new products, and the need to keep the patient at the center of decisions. David Wood, Faculty Director at Harvard Kennedy School’s Initiative for Responsible Investment, spoke about why venture capital doesn’t fit into existing paradigms of responsible investment happening in the public equities markets. Public equities markets have a strong focus on transparency and measuring materiality of ESG issues to company performance. Both public and late stage private equity investors lack the tools necessary to evaluate new technologies. One participant explained, “The [financial services] industry is still struggling to get the right data points and the right data specificity. With AI, I think we have nothing to work with.” In the venture capital space, there is a subset of ‘impact investors’ looking at specific products that could benefit groups of people or the planet, but there is a lack of attention to issues in other areas around risk, misuse, and unintended consequences. David Korngold, Director of Financial Services at Business for Social Responsibility (BSR), reflected on the experience of private equity funds in setting up ESG evaluation processes. He also noted there are many lessons that those interested in technology human rights and AI ethics can learn from other movements where businesses and investors have already mobilized, such as climate change. Key success factors include having a clear definition of what a successful outcome is, metrics, and a set of voluntary commitments that companies can sign onto.

Frontier Tech and Society: What Should Investors Consider?

In the second session, participants engaged in a case study discussion led by HBS Professor Mitch Weiss. The case featured Boston startup Voatz which had recently raised $2.2 million in seed funding to develop a mobile voting app using blockchain and facial recognition biometrics. The case featured a dilemma during a pilot with overseas absentee overseas voters in the West Virginia primary elections in 2018. A series of tweets and articles accused Voatz of Russian influence and criticisms of its data security practices calling the app ‘a horrifically bad idea’ and the ‘Theranos of voting.’¹

Weiss asked participants to take the point of view of a potential investor in Voatz and weigh the good intentions of the founders and potential benefits of the technology (increased access to voting and reduction of voter suppression) against potential risks of harms (election hacking, exposure of individual voting records, and subsequent risk of violence or loss of trust in democracy). The participants then generated a list of actions venture capital investors could take that would ‘increase the good and reduce the chance of harm’ such as developing term sheet requirements to hire a risk officer, requiring key employees to obtain security clearances, and creating governance, oversight, and accountability mechanisms on the board.

Supporting Frontier Tech Ventures

In the final session of the day, HBS Professor Shikhar Ghosh moderated a conversation among participants with Ann DeWitt, Partner at The Engine, a venture capital fund and incubator for ‘tough tech’ start-ups, and Matt Wansley, General Counsel at NuTonomy, a self-driving car company. Wansley spoke about the role of law vs. ethics in decision-making in startups and the importance of creating a culture where every employee feels comfortable speaking up on issues of risk and safety. DeWitt spoke about the importance of mission and motivation, and ensuring that founders are working to develop the full leadership capability set of a CEO. She also stressed that founders must be dedicated to ensuring applications of their technology will be ‘net positive’ for the world.

Key Insights

The discussion surfaced several key insights.

1. The coming wave of technological innovation will look very different from those of the past thirty years.

The coming wave of frontier technologies will likely be fundamentally very different from other recent technological developments. The pace is faster of technological adoption. Automobiles took 62 years to achieve the first 50 million users. Televisions took 22 years, personal computers 14 years, the internet 7 years, Facebook 2 years. Pokemon Go took 19 days.2 This means that there could be less time to fix problems with new technologies before they have massive societal impacts.

The stakes could be higher this time. The past decade of digital technological innovation focused primarily on applications for advertisements, consumer on-demand services, and enterprise solutions. This time the innovations in artificial intelligence and sensors could impact a far broader range of applications from health care, education systems, policing, criminal justice, elections, military, and national security infrastructure. Frontier biotechnologies such as CRISPR and neurotechnologies could also open new possibilities of human enhancement, which could dramatically change the structures of societies around the world.

The players will be different this time, particularly in biotechnology, where digital technology’s culture of rapid prototyping and rapid scaling is interfacing with the slower and more fastidious culture of laboratory science and clinical care. This could lead to revolutionary breakthroughs as well as disastrous missteps.

2. Investors and entrepreneurs should deeply understand the problems they are trying to solve.

Frontier technology ventures tend to be ‘solutions-led’ rather than ‘problem-led.’ Frontier technologists often begin the process by asking ‘which application of this technology should be commercialized first?’ rather than ‘what solution would best solve this customer’s problem?’

Investors and entrepreneurs should take a step back and ask what the actual customer problem is that needs to be solved and honestly assess whether the technology is the best solution. Similarly, investors should press entrepreneurs to consider the hidden risks and unintended consequences of the technology and how they will safeguard against them. Interviewing outside stakeholders as part of the due diligence process—such as a labor union leader, cybersecurity specialist, or a person knowledgeable on the community or issue in question—is one way to uncover these.

Additionally, technological development must be inclusive of the communities whom it seeks to serve and those that will be impacted by it. This requires diversity in backgrounds and viewpoints be represented at all levels of the organization from the board, to senior management teams, to product development teams. Additionally, product development teams must deeply understand the users of the technology and the dynamics of the local communities and cultural contexts where it will be applied across the world. This can be achieved through using principles of human-centered design and including diverse populations in AI algorithm training data sets.

3. Investors have many levers to influence the shape of technological development.

Far from being passive providers of capital, venture capitalists have many levers they can use to shape the overall direction of technological development. This begins with creating fund theses that seek overall positive impact in an industry or issue area, and defining which applications of the technology the fund will not support. VCs can also investigate deeper questions of risk and unintended consequences during the due diligence process, and write specific provisions in the term sheets to ensure adequate safeguards have been put into place. VCs also have many levers as board members engaging in active management by ensuring diversity of viewpoints and experiences are represented on the board, leadership teams and product development teams. They can also support founders in articulating mission, values, and building a strong culture that fosters respect for stakeholder groups and empowers employees to raise concerns and take action on issues of safety and risk.

4. Tech entrepreneurs need additional supports to help them consider societal impacts earlier on in their product design and business model development.

Many VC firms offer services ranging from traditional mentorship and incubator services to in-house subject matter experts and talent acquisition teams. Firms have well-experienced general partners ready to assist their portfolio companies with product development and business strategy. Some, as Andreessen Horowitz does with their Short-Term Advisory Relationships and First Round Expert Network, employ dedicated teams of operational and recruiting experts who advise on issues like product, design, engineering, marketing, finance, and human resources. Typically, these programs focus on operational capacity building (i.e supporting strategy or cost centers). VC firms could offer training that provides a more in-depth understanding the first- and second -order effects of the technology’s impacts and guides entrepreneurs in factoring these considerations into business models.

Continuing the Conversation

At the conclusion of the day, participants took part in a lively discussion over a reception and dinner to discuss next steps and ideas to take this work forward. In the post-event survey, participants highlighted needs for several additional research products, tools, and trainings including:

Evidence Base

• Business Case - Data on the business impact of frontier tech ventures factoring in societal considerations
• Case Studies - Short cases that demonstrate venture capitalists and founders working together towards successful management of technological advancement and risk; cases that demonstrate ‘turnaround’ techniques of a venture that has had a crisis of public trust

Investor Tools

• Investor Needs Assessment - Survey/focus group to identify current practices, concerns, and needs of venture capitalists to assess and manage frontier tech investments
• Term Sheets - Options and draft language to include on term sheets by technology category, application, issue area and stage of investment
• Technology Evaluation Tools - Rigorous due diligence tools to evaluate and manage risks of new technologies, especially data use/privacy, cybersecurity, and AI algorithms

Trainings and Awareness

• Tech Entrepreneur Training - Provide frontier tech entrepreneurs support on evaluating and mitigating risks, and working with regulatory bodies
• GP/LP Engagement - Engage general and limited partners to broaden scope of awareness of societal impacts of technology issues and to broaden definitions of ‘impact investing’ beyond social justice and climate change