Inaugural Engineering and Humanity Lecture, Columbia University, School of Engineering and Applied Science, New York, NY, November 18 2026
Science, Engineering and Innovation: Why Are You Studying Engineering?
By Gerald Chan
It is wonderful to be back on campus as this campus means so much to me. My aunt who was the first person in my family to study in America went to Teachers College for her master’s degree. Growing up, I had always heard of Columbia as a venerated institution of higher education.
When I first came to this country, I was a teenager sent to live with relatives in Englewood Cliffs across the George Washington Bridge. I didn’t like living under the roof of a formal, older couple and plotted my escape. I found it in Columbia’s summer school. I could have gone to NYU to take a course in calculus that would have counted for credit in an engineering curriculum; I chose instead to take a course in physics here at Columbia. That choice was a landmark for me; it was the first time I got to choose what I wanted to learn for my own sake and not for the sake of fulfilling some requirement of a curriculum. It was my first taste of intellectual freedom.
I lived in John Jay. My budget was meager. I found that a few blocks north of campus, I could buy day-old bread for 35 cents a loaf and I survived the summer on bologna sandwiches. A cheeseburger from the grill in the dormitory cafeteria which cost 99 cents was my treat for the week. Every Sunday, I took the IRT and then the E train to go to the church in Queens that my aunt had attended and it was there that I met my future wife. The bliss of a carefree teenager, of life unfettered, of innocent times, and the opportunity for a supple young mind to ingest knowledge ravenously for no other purpose except the pleasure of learning — these are the memories of my summer at Columbia. They meant so much to me that twenty years later when I started in the investment business, I named the firm Morningside.
Let me move on from memory lane to the present. I have in the last two years given a series of lectures on higher education in the age of AI. Not being beholden to academia, my commentaries were meant to be provocative and challenging to stasis. Everything in the world is subject to inertia, and that is why energy is required for change. Without supplying fresh intellectual energy, I would consider myself to be of no value to you.
This being a school of engineering, and my host being the dean of engineering, I’d like to offer a few thoughts on the topic of engineering education, in particular as it relates to science on the one hand and innovation on the other. What I would like to do is to sharpen the focus of students of why they are studying engineering. My hunch from talking around is that many students are nebulous about why they chose engineering as a major. For many students who are good in math and science in high school, but find math, chemistry or physics too dry, engineering becomes the default choice. They end up in engineering by a process of elimination.
First, what is science? A field of inquiry is science not by virtue of the subject matter of the inquiry, but by virtue of the method of inquiry. Science is the application of the method of formulating falsifiable hypotheses and subjecting them to validation or invalidation by empiricism. This has come to be known as the scientific method. Hence we can have physical science, life science, social science, etc. For our discussion today, I will refer primarily to science as an inquiry into nature, hence, natural science.
Engineering, on the other hand, is the application of science to solve problems. I emphasize that engineering is based on science. Without science, engineering becomes mere craftsmanship. When engineering produces a novel solution to a problem, one that was non-existent before, we have innovation.
The goal of natural science is discovery of how nature works. One does well in science by being the first to make a discovery. Once a discovery is made, being second has no consolation prize even if it is a great service to the scientific community to provide independent confirmation of the discovery. Scientists fight to be recognized as the first to make an important discovery. A famous fight was over who first discovered HIV as the virus that causes AIDS. A bitter feud between the French scientists, Luc Montanier and Francois Barre-Sinoussi, and the American scientist Bob Gallo was elevated to a flight between the French government and the American government. The two governments eventually settled, but when the Nobel Prize for this discovery was awarded, it went only to the French scientists to the exclusion of Dr. Gallo.
Publication in a peer-reviewed journal is how scientific discoveries are recognized. Scientists therefore fight over the priority date of publication as well as the priority of authorship on a paper. Who is the first author? We now have papers with multiple co-first authors. One infamous case is that of the two Chinese physicists who won the Nobel Prize in physics in 1957, T.D. Lee here in Columbia and C.N. Yang of Princeton at the time. Because of the fight over who would be the first author, the two became estranged for the remainder of their lives. In short, scientists work for discoveries; in turn, discoveries are recognized by publications and prizes. For this whole process, a scientist’s work is judged by his peers. This is a point that I will return to later.
The goal of engineering, on the other hand, is not discovery, but use, or utility. The end game is not a paper, but a product or a prototype thereof, or an invention which in some instances results in patents. The hallmark of a great engineer is not what papers he had published, but what products he has designed or made. What problems have those products solved? What impact have they had on society? The adjudication here is not peer review, but market acceptance of the product. This can be illustrated by the invention of the PCR machine, or polymerase chain reaction for amplifying DNA which then enables sequencing. When Kary Mullis invented PCR, he asked a prominent scientist how many PCR machines he thought would be sold in the first year. His colleague said two, on the premise that one would be located in Harvard University and the other at UC Berkeley and between them, they would do all the PCR reactions for the world. In reality, a few hundred machines were sold in the first year after launch. What matters is not the opinion of peers, however authoritative, but market demand and acceptance.
I am painstakingly belaboring the distinction between science and engineering because I feel that engineering education has suffered from this distinction being blurred. Engineering education has veered from the roots of engineering which is to design and build things. We used to think of engineers as builders of bridges, dams, chemical plants, airplanes, cars, and all manners of machinery. The disposition of engineering is to build things. Scientists make discoveries, engineers make things. Science explains, engineering makes use. While engineering is based on science, it is a distinctly different endeavor.
The blurring between science and engineering in American universities began after WWII. In 1955, the American Association of Engineering Education issued the Grinter Report which recommended that engineering education become less apprenticeship based, less vocational, and become more math and science based and more like a professional school. The intention of the Grinter Report is reminiscent of the Flexner Report of 1910 in its attempt to base the practice of medicine on science. Sometime around 1959, Columbia changed the name of its School of Engineering to the School of Engineering and Applied Science. In those few years, many universities made the same name change, notably Yale, Princeton, George Washington, Virginia, WashU, reflecting a shifting emphasis from engineering to science.
I want to quote Herbert Simon on this point. Simon is one of my intellectual heroes. His work in computer science, considered the first true AI program, won him the Turing Prize in 1975; his work on bounded rationality, seminal to the field of behavioral economics, won him the Nobel Prize in economics in 1978. In 1969, he wrote this:
“…. it is ironic that in this century the natural sciences almost drove the sciences of the artificial from professional school curricula, a development that peaked about two or three decades after the Second World War. Engineering schools gradually became schools of physics and mathematics; medical schools became schools of biological sciences; business schools became schools of finite mathematics. The use of adjectives like “applied” concealed, but did not change, the fact. It simply meant that in the professional schools, those topics were selected from mathematics and the natural sciences for emphasis which were thought to be most nearly relevant to professional practice. It did not mean that design continued to be taught, as distinguished from analysis.”
This quote came from Simon’s book The sciences of the Artificial. In it, he highlighted the difference between science which deals with the natural and design which deals with the artificial by which he meant anything that is man-made. His concept of design may be used interchangeably with engineering which is to design and create things with the intention of solving problems. For Simon, scientists analyze, engineers create.
What Simon critiqued was Grinter gone overboard. Is an education in engineering an education in selective areas of science or is it training students for the practice of designing and creating things? Are we measuring the engineering students by their mastery of science or by their ability to create things?
When I enrolled in UCLA as an undergraduate, it was the year that UCLA’s College of Engineering was changing to the School of Engineering and Applied Science. It wasn’t just a name change, it was also a change in the curriculum. Students entering that year had the option of choosing the old curriculum of the College of Engineering or the new curriculum of the School of Engineering and Applied Science. I surveyed the two curricula and opted for the new principally because the new curriculum had eliminated the requirement for a senior capstone project. I remember feeling greatly relieved because that meant I didn’t have to make anything. The thought of having to make something scared me, so I cobbled together a hotchpotch of courses in fluid mechanics and applied math and ended up with two engineering degrees. In reality, the education I received was an education in science, not engineering. None of the courses I took placed any requirement on me to make anything.
At that time, a friend a few years my senior had gone on to study engineering in University of Michigan. He told me about a competition his school held for the students to design the simplest way to drop an egg from the rooftop of the engineering building to the ground floor without the egg breaking. The winner was a contraption like an ice cream cone. The weight of the cone would ensure that it is always below the egg as it falls. As the contraption hits the ground, the material of the cone would deform and absorb all the impact from the fall. What an elegant design! It is not surprising that to this day, the engineering school in University of Michigan is still the College of Engineering and not a school of engineering and applied science.
Let me now turn to a different approach to illustrate the distinction between science and engineering by comparing two countries, albeit of different times. The first country is Britain in the first half of twentieth century. By any account, Britain was by then a powerhouse in science, having produced great scientists like JJ Thompson, Rutherford, Bragg, Chadwick, Ramsay, Eddington, Dirac, and Dorothy Hodgkin. In 1928, penicillin was discovered in London by Alexander Fleming. He published his discovery in April 1929. That discovery lay dormant for eleven years until the Oxford group led by Howard Florey worked on characterizing and producing penicillin. They were also the first group to use penicillin in man even though they could only produce penicillin in minute quantities. In order to have enough drug in the patient for efficacy, they had to extract the drug that was excreted in the patient’s urine and recycle it.
WWII soon broke out and the race was on to make penicillin in large quantities to treat soldiers wounded in battle. With all that political exigency, none of the British pharmaceutical companies had any interest in manufacturing penicillin. In desperation, Florey sailed to America looking for help in scale-up and manufacturing. The solution came from a Department of Agriculture research lab in Peoria, Illinois and from a bulk chemicals company in Brooklyn that was primarily making citric acid as a food preservative and a flavor enhancer. That company had developed a deep tank fermentation technology that could be used for growing the mold that makes penicillin. That company was Pfizer. For the invasion on D-Day, Pfizer made 2 million doses of penicillin. In a span of a few years, the world went from recycling penicillin from the patient’s urine to making two million doses. That’s engineering! In 1945, the Nobel Prize for discovering penicillin went to three Brits, but the pharmaceutical giant that was born was in America. Britain got an A for science and an F for engineering. It was a particularly big bummers for the Brits because when they tried to produce penicillin after the War, they had to pay a royalty to the Americans.
In contrast, the second country I want to profile is China. This past summer, I attended a biotech conference in Shanghai. In 15 minute blocks from 7:45 am to 3 pm every day, one Chinese biotech company after another took to the stage to introduce their drug development programs. It was striking that so many companies were doing the same things. They were going after the same drug targets, and doing so using the same panoply of tools. Each company had a program in small molecule, bispecific antibody, protein degrader, T-cell engager, car-T, etc. There was absolutely no new science there, but there was a lot of engineering. Every company was trying to make a drug. In contrast to biotech companies in America which are formed and built on novel science or novel technology, the Chinese companies simply tweak known tools to make drugs against known targets.
In spite of the absence of new science, the Chinese biotech industry has been exceptionally productive. One example is a company that made a bispecific antibody against two old drug targets, vascular endothelial growth factor, or VEGF, and the immune checkpoint PD-1. The first anti-VEGF monoclonal antibody as a cancer drug was approved in 2004. That was 22 years ago, before you undergraduates were born. Anti-PD1 antibody as a treatment for cancer was first approved in 2014. That was 12 years ago. By making a bispecific antibody against these two old drug targets, this Chinese company showed in clinical trial a significant increase in progression-free survival for non-small cell lung cancer patients. There is absolutely no new science here, only anntibody engineering. This case is by no means an isolated one in biotech today. By taking an engineering approach to drug development, the Chinese biotech companies are making many promising new therapeutics. Last month, Roche licensed a portfolio of antibody drug conjugates from the Chinese biotech company Hansoh for a whopping $15B. It is all engineering. One third of the new licensing deals done by multi-national pharmaceutical companies worldwide are now for assets developed in China.
I like to point out that the largest funder of science today is the taxpayers. This is true for virtually all countries. Taxpayers and politicians as a whole do not understand cutting edge science. Engineering is what turns the achievements of science into a payback for the taxpayers. Without such paybacks, taxpayers will see no point in continuing to support science. It is not too much to say that the future of science funding depends on what we do in engineering today.
This leads me to the topic of innovation. I said earlier that engineering is about solving problems and when the solution is novel, we have innovation. The experience of participating in some innovation work should be an integral part of an engineering education. Just doing problem sets, the staple of engineering courses, no longer suffices. Solving known problems posed by someone else does hone technical skills. Ideating what are the unmet needs in society or in technology and solving them is a higher order objective of an education in engineering.
I want to tell you the story of a company named Encora. In full disclosure, I am the chairman of the board of this company. The company came from the senior capstone project of three MIT undergraduates who were teammates in mechanical engineering course 2.009, in MIT speak. The students in that course were grouped into teams of sixteen. The first assignment of the course was for each student to come up with 100 ideas of what to work on for their project. Yes, it was 100 ideas per student. Outlandish as this may seem, this requirement in fact exercises a mental muscle that few students had ever exercised. I do the same thing with young people in my office. I tell them to take a walk to the pizza shops and come back with five new business ideas.
The process of going from 1600 ideas to one was also a valuable part of the education process. How do you make tradeoffs among competing merits and demerits of ideas? Is the basis for that decision formulaic or intuitive, or a combination thereof? How does a team of people with diverse interests and abilities come to a consensus and work together? The team had to winnow 1600 ideas down to a few finalists and vote for one project to take forward.
In this case, the vote came down to 15 to 1, with the lone dissent coming from the student who originated the idea. He did so on grounds that the project would be too difficult to succeed. The vote that carried was to make a vibrating device worn on the wrist to treat the hand tremor of Parkinson’s patients. With off-the-shelf parts, the team built such a device. Still, none of the team members had ever met a Parkinson’s patient. They finally found a patient online who was willing to be the subject to test the device. That patient was an artist who was no longer able to draw because of the tremor. Then came class demo day. There, in a packed auditorium, the patient was asked to trace a spiral with pen and paper. Without the device, what he drew were wiggles. When he put on the device, he performed a smooth tracing of the spiral. The room gasped and then broke into applause. Here is the video of that moment.
https://youtu.be/McObwG-pmus?is=yAoW3W541pfSNcDB
I am sure the students on the team got a good grade, but that is not the point. The students had tasted the power of engineering delivering real-life impact. Instead of getting a job, one student decided to keep pursuing the project after she graduated from MIT. To pay rent and put food on the table, she coached the MIT women’s volley ball team, she got jobs on Rover to walk dogs and she did whatever gig work she could find to keep going. Another student from the team had gone on to a job at Apple. He felt both bored and stunted. Remembering the excitement and the potential for impact of the device, he quit his Apple job and joined his former teammate to start a company. They entered business plan competitions to win prize money to finance their work. They raised small amounts of capital from anyone who was willing to invest.
It was in one of these pitch competitions at MIT that I met them. I became the backer of the project, but it was much more than just financial support. My colleagues who know about clinical trials and FDA worked with the founders for the next phase of the development of the device. To make a long story short, we worked together for three years and did a series of clinical trials. In March this year, the device was cleared by the FDA as a medical device for treating tremor. Today, the company is gearing up to launch the product. Thousands of patients have come to the company’s website to get on the waiting list for the device once reimbursement is worked out with Medicare. If one models the pricing of this device and the total addressable market, there are all the markings of a substantial company in the making here.
Several points are noteworthy in the Encora story. The device did not come from any new discovery in neuroscience. The founders were mechanical engineering students. The only input from neurology in this work came from the French neurologist Jean-Martin Charcot who wrote, “For a long time I had learned from patients with paralysis agitans [Parkinson’s disease] that they got great relief from journeys by railway or carriage.” This inspiration for Encora’s vibrational device was published in 1892.
Two clinical findings emerged from Encora’s clinical trials. First, the contralateral hand that did not have the device on the wrist also had reduced tremors. Second, after the device was taken away from the patients at the end of the three-month clinical trial, their tremor did not return for days. Clearly, there had been a reprogramming of the central nervous system by signals coming from the peripheral nervous system. This stands in contrast to textbook understanding that the peripheral nervous system is under the control of the central nervous system and not the other way around. Rather than engineering being informed by science, what we have here is new science coming out from engineering.
This point is critical because if you can intervene from the periphery to affect the CNS, many more non-invasive therapeutic options become possible. This is exactly what was done in another company I built named Cognito Therapeutics. The company just announced its Phase 3 clinical trial results which affirm that neurostimulation from the periphery can reprogram the CNS, in this case, a dramatic slowing down of cognitive decline in early Alzheimer’s patients.
This work came about by a graduate student observing that the EEG of Alzheimer’s patients lacked oscillation in the gamma frequency range. The hypothesis was to use light and sound flickering at 40 Hz to induce brain waves of that frequency. The hypothesis driven science was done at MIT, albeit, advancing only to animal models. I licensed the patents, formed the company and took this treatment into man. It took us 10 years to do the engineering and clinical trials. The data showed that this intervention can slow down progression of Alzheimer’s disease by 70%, doubling the slowing down of cognitive decline achieved by the anti-amyloid antibodies and doing so in half the time. This therapy is extremely safe, works for patients with mixed brain pathologies, and can be done by the patient at home. It does not require building out any infusion center infrastructure as is needed for the monoclonal antibody therapy. This work will change the landscape of dementia medicine and the disease’s impact on society.
There is nothing novel in looking at the brain as an electrical circuit. People have been doing EEG for many years as a method of measuring brain activity. What is novel in Cognito’s work is in how to manipulate the electrical circuit to treat a brain disease. This is a departure from just about all of life science which treats biological processes as a collection of chemical reactions and using a reductionist approach of introducing perturbations, usually in the form of an drug, and observing outcomes.
By sharing these examples of my work, I like to impress upon you that your education should result in the formation of your independent thinking, conviction, and daring to do something that no one has done before. Some call this creativity. Remember I said earlier that the work of a scientist is judged by fellow scientists in the form of peer review. The drawback of peer review is that it is inherently conservative. Consensus among peers has a tendency to regress to the mean. That is contrary to innovation. People in the mainstream anre rarely hospitable to innovators until they see that the innovation is about to become the new mainstream. Innovators have to have their own conviction even if it is against peer consensus, and have the tenacity and the risk tolerance to overcome obstacles to get to a breakthrough.
What the Encora story also shows is that the barrier to innovation is not as high as people think. When the barrier to innovation is high, ideas for technological breakthroughs tend to come from professors, post-docs or graduate students. This is still the case in much of life science. Where the barriers to innovation have come down dramatically is in the field of AI. This has led to many great ideas coming from undergraduate students resulting in unicorn or decacorn companies, the likes of Mercor, Cursor and Figma. I have a few townhouses a block from the Harvard campus that are rented to a community of undergraduate students who are all company founders. The ring leader who was always complaining about the rent being too high sent me a document last year showing that had I taken stock in their startup companies in lieu of cash rent, the take would have been valued at $3 million at that point, albeit, in unrealized gain.
I do not mean to create an impression that innovation is easy. Most startups fail, and I surely have my fill of them. That is why I have never encouraged any undergraduate to drop out of school for doing a startup. Over the years, I have struggled with whether the Thiel Fellowship is a good idea or not. I suppose there is no one-size-fits-all. May be schools can make it less binary, i.e., either stay in school or drop out to do a startup. In other words, carve out time in the curriculum for students to try doing a startup, give them the training and support to do so, and give them credit towards graduation for that work. One thing I am sure and that is the experience of doing a startup is extremely beneficial to a student’s development regardless what he ends up doing in the future.
I can understand that students have a finite number of credit hours. Taking time to do something entrepreneurial must mean cutting back on learning something else. On this topic, I think that with AI now, people can learn anything on the fly. It is a shift from learning just in case to learning just in time. As an engineer, what you need to be able to do with any problem is to go back to first principles. From there, you will figure out what you need to learn if you don’t already know it. Professors need not worry that they have not taught the student “everything”, whatever “everything” means. They definitely need to be equipped with a basic set of skills. For the rest, if they have the right intellect, they will figure it out.
No talk on education today would be complete without touching on the subject of the role of AI in education. Once again, I will resort to story telling to make my point. I heard a talk given by a professor of aeronautical engineering earlier this year. He taught a 4-week course in which he told the students to use AI to help them design a jet engine. The students were given unlimited use of tokens. Of the ten teams of students, only one team succeeded. That team consisted of seniors majoring in aeronautical engineering. When the successful team presented their work to the corporate sponsor, the company’s engineers were stunned because the students had solved a problem in jet engine design that they had not been able to solve for years.
Of the teams that failed, teams that consisted of seniors performed better than teams of juniors who in turn did better than the sophomores. Teams with aeronautical engineering majors also did better than students from other majors. There is an important message here — even with AI, domain knowledge matters. In the words of this professor, “AI will meet you where you are at.” The more educated the user is, the more powerful is AI in his hands.
My view is that in the age of AI, the work of a professor is more important to the student’s development than ever before. They have to insure that the students master the domain knowledge in the absence of AI and at the same time, the students have to be savvy in using AI to elevate what they can do. These are two sides of the same coin. Different professors may have different views on this matter and that is perfectly fine. In fact, it is better for the students to be exposed to professors with different practices. There is no need to require all professors to adopt the same practice.
Two weeks ago, the Provost of Stanford announced a $10 million fund for academic departments to redesign their undergraduate curricula to be AI-aware. It is a recognition that how we educate undergraduates has to change in the age of AI. No one has figured this out. The Stanford fund is meant to encourage experimentation.
I also want to mention one external force that is massively changing the practice of engineering today and it comes from what is happening in Ukraine. While we hate wars, they have historically been powerful accelerators of engineering feats. WWII was definitely such a moment. Today, the warfare in Ukraine has changed the time scale of engineering work. In the drone warfare between Ukraine and Russia, one side comes up with an innovation. Within days, the other side figures it out and devises a countermeasure. The conflict in Ukraine is in essence a contest of which side can innovate faster. The clock speed for innovation in this war runs on days, not months or years. In a span of three years, Ukraine has emerged as a powerhouse in drone technology. Even the bureaucratic U.S. federal government is changing how it works with defense contractors. Companies like Anduril and many AI native startups are winning defense contracts at the expense of legacy behemoths like Lockheed or Northrop Grumman.
Because the forces that are quickening the pace of innovation are coming mainly from outside academia, it is important that engineering schools do not become ivory towers. I remain an advocate for academia and industry to be more communicative in ideas, projects and personnel. Aside from funding, industry can provide a lot of signals for how engineering education should evolve.
The undergraduate years are pivotal in the formation of a young person. These are years of maturation wherein the intellect of the student is both populated with content and is shaped to function independently. These are also the years in which the student makes a transition from being a dependent to being a contributor to society as a member of the workforce. That transition from school to work should not be a jarring one. It should feel like the natural next step in a student’s experience. This is what employers are looking for, and this is what we as educators owe the students.
In closing, I have said this before, and I will say it again. In the age of AI, it no longer suffices just to possess knowledge; what matters is what you can do with what you know. This brings us full circle back to the very kernel of what engineering is all about.