Investiture Address: Arati Prabhakar

Director of the Defense Advanced Research Projects Agency (DARPA), member of the National Academy of Engineering, and a fellow of the Institute of Electrical and Electronics Engineers

June 11, 2016

Thank you to Dartmouth for this very special Award — it means a great deal to me.

So! Hello graduating students — hello families — hello friends! And congratulations to all of you on this joyful day!

Graduates: I am so pleased to be among the first to welcome you as the newest members of an amazing community — a community that designs, that builds, that creates ... a community that SOLVES! I'm talking about the community of engineers!

And as one engineer to another, let me just say — you are in for an incredible ride ... an incredible journey. And let me also say ... no pressure! ... but you’re joining the club just in time!

Because there are a lot of things to fix out there. A lot of problems to solve in the practical, rational, powerful and yes, elegant, ways you've all been learning and practicing here at this exceptional engineering school at Dartmouth.

Now, we all know that engineering is going to be critical to solving our problems, to getting us ... as a nation and as a civilization ... to where we need to go. Because of course engineering has brought us the world we inhabit today ...

We create magnificent skyscrapers that simply shrug when the tectonic plates beneath them lurch...We design million-pound aircraft that leap into the sky and cross oceans….We build the infrastructures that bring us water with the twist of a faucet... Electricity with the flip of a switch...

And any episode you would like of Game of Thrones with the swipe of a finger!

And engineering has stretched well beyond utilitarian needs. It has allowed us to send probes to Mars and Pluto and beyond ... to see through clusters of stars and galaxies all the way to the edge of our universe, back to the beginning of time itself. So there is so much we've done and so much we CAN do with the power of engineering — not just to fulfill our practical and physical needs but to help us understand our universe and ourselves and to advance the cause of human progress.

But now let me take note of the other side of the equation. Because you all know, there's always the other side. And the other side of great power is great responsibility.

Everything we do as engineers, every benefit we bring, carries a potential cost. And as our power as engineers grows, these potential costs ... the potential risks ... grow as well. And these risks don't belong to somebody else. It's not like, "If we build it, they will come, and they can clean up the mess we made while we were building it!"

For example, we engineers are wrangling vast volumes of data these days — and that's why algorithms read our posts, calendars, and purchases, and figure out what you're thinking about buying next, or where you're planning to travel, or who you might want to date.

And it doesn't stop there. There are algorithms now that will measure my sentiments, who I'm likely to vote for, whether I'm inciting jihad ...

These analyses — these insights – can be incredibly valuable, be it for marketing or for security. But what does this mean for a cherished American value — the value of privacy?

If we know that our contrail of data — or our company's contrail of data — is constantly being watched, will we be willing to experiment — to be open about new ideas — to risk the embarrassment of a crazy idea? That’s something that has to happen for creativity to flourish. Will we take the risk that is part and parcel of reaching for great impact?

Some of you will tap into the richness of vast data in your work. How will YOU resolve the natural tension that exists today between big data with all the benefits that can come from it, and good old-fashioned privacy?

And classifying online data to serve up the next ad or to try to stop a terrorist is just one of the areas where engineers are doing amazing things today with artificial intelligence and machine learning. AI and machine learning are starting to be used to control systems, too — and again the possibilities are endless.

Imagine 100s of times more data through a fixed frequency band because machine learning embedded in radios can find unused portions of that band from one instant to the next. Imagine machine intelligence in networks that can automatically detect and respond to an incoming cyber attack. Imagine artificial intelligence that powers autonomous physical systems.

And there will be hard choices here too.

I work in the Defense Department where we grapple with whether and when to grant certain degrees of autonomy to weapons systems, like missile-defense systems that have to respond instantly if they are to protect the lives of hundreds of sailors on a ship. But that's not the only arena where we engineers need to think through how we grant crucial decision-making powers to machines. Think about self-driving cars, which I suspect will be cruising our streets before long. They too will sometimes have to make extremely difficult, life-or-death choices instantaneously.

How will YOU assess what machine intelligence can and can't do and when to trust it? How will YOU decide how much power to delegate to a machine, as we and our machines become increasingly intertwined?

And speaking of becoming intertwined: We engineers, in collaboration with neuroscientists, are making unprecedented strides in our ability to decode the intent and the meaning of the neural signals generated by our brains. And it's a two-way street, so we are also starting to learn how to create neurological signals to send in, to communicate with the brain.

These advances can be transformative for our wounded warriors, who will be able to operate advanced prosthetic arms and hands with just their thoughts, and for warfighters and others with traumatic brain injury who need help restoring memory function.

But beyond just restoration — neurotechnologies could help accelerate learning by inducing neuroplasticity in the brain — which would be a total game changer. Imagine sponging up a new language as easily as a 5 year old can!

Where will YOU draw the line as we move from restoration of function to augmentation?

We're okay with braces and laser eye surgery and pacemakers. Are memory-boosting brain chips different? What will YOU consider appropriate in a future of cyborg possibilities? 

These are difficult questions, and as we gain even greater power to harness matter and energy and information, more of these difficult questions will come. Which is to say, with graduation you are inheriting and committing yourself to professional responsibilities that are significant — that are profound.

OK, so why am I talking about all this heavy stuff when we’re here to celebrate your graduation and your future?

I’m thinking we’ll go home and our daughters will say: "Hey, how’d the Dartmouth talk go?" “I don’t know… They were all sitting there like this…”

The reason I’m sharing these mind-benders with you is this:

It is precisely quandaries like these that can bring you the deepest satisfaction and the greatest joy in the work ahead. Because really, would you have it any other way? Would you rather suffer the fate of small dreams?

I can tell you that my most gratifying days — my most joyful days — are the days when my colleagues and I are struggling with these kinds of questions. Because those are the days when we know we are truly at the frontier, truly in position to see the next chapter of human progress.

Nothing is more rewarding than wading into the sublimely rich intersection of technology and humanity ... shaping its contours and dimensions and bringing to that process the very best instincts we each can offer about how to make the world better and stronger ... More technically empowered and at the same time more beautiful and more simply human than ever.

So please ... revel in the mix of opportunity and responsibility you've earned. Take pleasure in the challenges ahead and infuse your work with optimism and exuberance.

I can’t wait to see the world you’re going to create! Thank you!