[email protected] House startup with Berkeley alumni will help foster innovation

coloredited_hannahcooper_ai
Hannah Cooper/Staff

Big data. Artificial intelligence. Machine learning. It’s difficult not to have heard any of these phrases, especially considering the looming presence of the technology industry here in Berkeley. However, UC Berkeley — which produces groundbreaking research in areas such as computer science and its burgeoning subfield of artificial intelligence, or AI — isn’t just known for its technical prowess. The campus is also a leading hub for entrepreneurship, with renowned initiatives, such as the Sutardja Center for Entrepreneurship & Technology, and accelerators to promote innovative startups by Cal students, faculty and alumni.

I’m incredibly fascinated by AI because of both my interest in human cognition and its potential impact in society. Like many of my peers, I’m facing the choice on whether I want to work on problems in AI that I find interesting through research or directly make an impact on society by working in industry. I’ve also had a brief taste of the pitfalls of either option. Research can often involves exorbitant amounts of effort for comparatively little rewards, and industry can become monotonous; turnover rates for jobs at companies such as Google, Facebook and Uber, where many Cal students aspire to work, are between one and two years. I’m also intrigued by how the startup world presents the chance to work on dynamic problems that can have far-reaching societal impacts.

As a result, I’m extremely pleased that UC Berkeley is leading the way when it comes to partnerships between academia and industry with the new [email protected] House initiative. The House, one of Berkeley’s many startup accelerators, is partnering with several UC Berkeley alumni and professors, who are leaders in the field of artificial intelligence, so that they can provide critical mentorship to startups leveraging this technology. [email protected] House will be a great opportunity for The House’s portfolio companies, significantly increasing their likelihood of success in a competitive environment, and will also help promote more socially responsible innovations.

Startups and uncertainty often go hand in hand. The pace of work is extremely fast, as employees work (or, at least, think about their work) around the clock in an attempt to build a product that attracts users and, in turn, investment. The technical complexity of AI projects significantly adds to this burden.

One of the biggest barriers to deploying scalable AI products is hardware. In order to perform tasks such as recognizing speech or predicting a user’s preferred purchases, a computer must learn from large amounts of existing information or training data by creating a model and then use these mathematical and statistical models to predict future trends. Such processes are referred to as machine learning and involve a substantial amount of computation. Production-level machine-learning models would overwhelm a conventional computer, so they need to run on a graphics processing unit, or GPU, which can handle much more data than the conventional central processing unit, or CPU.

Additionally, the field of AI is extremely challenging to fully understand — oftentimes, a graduate degree in a discipline such as math, statistics or computer science is needed to understand many of its computational processes, and developing the architecture to run these processes is an equally challenging endeavor.

[email protected] House will help alleviate both these issues. The House will be providing GPUs to members of its AI cohort, and founders will be mentored by experts conducting cutting-edge research in the field. To top it off, many of these researchers have entrepreneurial experience — four out of the six professors who will advise The House’s AI accelerator’s startups have founded ventures of their own, including campus electrical engineering and computer sciences professor Pieter Abbeel’s company Gradescope.

There are also key benefits for the researchers. Much of scientific research involves putting intense amounts of effort into publications, which are highly theoretical in nature and are read by a limited audience. In many cases, research is far ahead of industry; many technologies that are the bread and butter of AI in this decade were being researched in the 1990s. This partnership with startups will be critical to both implementing the results of the faculty members’ research into products that will genuinely impact the lives of their consumers and ensuring that the latest developments in AI research have a major impact on society.

Another major reason I’m optimistic about [email protected] House is that this partnership between academics and entrepreneurs has the potential to set a precedent for socially responsible innovation. In many cases, AI has been used to promote profit for companies over their users’ well-being, whether it be by curating social media feeds to only include content a user would enjoy, effectively creating an echo chamber, or for tasks as trivial as editing selfies.

However, many of the professors’ research interests lie in far-reaching areas — including robotics in medicine, self-driving cars and the intersection of deep learning and cybersecurity, just to name a few — and the electrical engineering and computer sciences department at UC Berkeley stresses the importance of serving society through technology. Given the lucrative salaries of major technology companies, many of these researchers made the decision to pursue research and teaching so that they could help inspire more students to pursue the fields they are passionate about. As a result, we can expect these professors to invest in startups that align with their priorities in research, in order to make the most of their time with The House. AI has the power to revolutionize industries from finance and business to education and health care, and with the guidance of leading researchers in the field, The House can help set a precedent of applying the latest technology to where it matters most.

 

Sathvik Nair is a second-year student at UC Berkeley studying cognitive science and computer science.

Please keep our community civil. Comments should remain on topic and be respectful.
Read our full comment policy