Back in late 2020, during the heavy downturn of all operations due to COVID-19, basically all businesses were sent to online, including education. Universities around me all became 'digital', the form of education mostly known for an inferior and less prestigious delivery of a university degree.
Till then, along with my IT start up, as a means of company promotion, I had been running a short crash course for Machine Learning / Deep Learning, but I had had growing inconfidence that students have awfullly shallow understanding of what I teach. Most of them just hoped to 'steal' a production code for data science research, but then they later complained that my code did not work for their own environment. They simply did not understand that data science is a basic form of knowledge to be modified for every different situation. I could feel that without proper fundamentals, they would not be able to differenciate code libraries for AI/Data Science and real data science w/ scientific minds.
After all, they thought data science is 'coding', not 'science'. To fix that mis-perception, I realized that I should deliver the education w/ all necessary scientific backgrounds, thus in a form of university degree. Fortunately, due to COVID-19, people became much more open-minded for online delivery of education, even for quality and prestigious degrees. If it is online, I can run the department anywhere in the world. That's how my journey for SIAI had begun.
Why University costs you $$$
It might vary depending on where you live, but generally founding a university costs a lot of money and efforts. After learning all that, I came to understand why most universities charge $$$ unless they are heavily funded by governments and/or philanthropist(s). In my neighborhood, minimum requirements for an offline university was around US$100 million, and even the online needed over US$40 million. That's just for initial financial requirement. Profitability after such mega investments is also a huge question. Most universities hunt for students thus give out easy grades and barrier-free diploma, which are known as 'diploma mills'.
I wanted to 'torture' students to be experts in data science. Founding a 'diploma mill' was the last thing that I would ever do. In fact, it's going to harm my business's reputation, thus profitability was a secondary concern. But in the process of looking for alternative, I come to understand that 'quality' over 'profitability' is the silliest strategy for a university.
The very first warning I got from a senior professor of a 3rd tier university in my neighborhood was,
Teach good, you lose students
In other words, if you put 'quality' first, you lose 'profitability'.
After months of trials and errors with my local unsupportive authority, I come to understand that it is better to tag along with another university than building my own. I was just not ready to justify that chunk of investments for 'quality' over 'profitability'.
CHF500,000, if to run a 'AI/Data Science' department?
My neighbor universities were all reluctant to put 'quality' first, so I turned my head abroad. Most countries had its own strict rule even for online education, despite COVID-19, so eventually I ended up with Swiss, the country known for asylum seekers and extreme freedom in business.
I had had talks with total 5 universities some with extensive details for contract. There were two schools eventually willing to give me a head of department for AI/Data Science, on the condition that the department is profitable enough. If not profitable, I had to pay the 'operation cost', which was CHF500,000 per year.
It is way better than a one-time investment of US$40 million w/ subsequent uncertainty in following years, but CHF500,000 per year to run my own department also sounded ridiculous. Running such a department for 'quality' should dramatically change the school's reputation, on the positive side. Back then, I was too naive to think that that they would be happy to accommodate a serious researcher as a leading member of such a popular department.
I was blatantly wrong.
If I have to pay that much, just to help them to keep profitable, then, why not build my own?
All-in for 'quality' over 'profitability' to be business-unfriendly
I began searching for alternatives. I first widened my scope to other countries, but options were not good enough. I still had to make some level of financial committment, which eventually taught me that school is just another line of a business. I had to think it from a business guy's perspective, despite the fact that I was willing to go 'all-in' for 'quality' over 'profitability'. With this kind of 'business-unfriendly' mind, there was no university in the world ready to take me in.
So, I came back to Swiss. With less strict rule for a univesity foundation than any where else in the western hemisphere (and other advanced countries where I can trust the country's institution), this is where I should build my own school, however unprofitable it will be.
Even in Swiss, the country known for open market without tough regulation, it still was not easy even for setting up a legal entity. Lawyers asked me a bunch of questions as if I plan to immigrate to Swiss, and charged me ridiculous amount of money. The Swiss local authority disallowed my company funded institution in Swiss, because they thought I would use that Swiss entity for money laundering and/or crypto currency related businesses.
Took me months to convince the authority that I was genuinely interested in offering quality education and there were not that many alternatives other than setting up a business entity in Swiss.
In the process, I was lucky enough to meet supporters across Europe, who later teamed up to build the current mother institution of SIAI, named as Global Institute of Artificial Intelligence (GIAI). Without the serial luck with people, SIAI would have never happened. I always feel indebted to my collegues.
The 'whacko' idea of scientific training in AI/Data Science not for profit but for quality certainly has really proven 'whacko'. It has not been profitable at all. People give doubts of the quality of education, because we deliver everything online, until they see our lecture notes and final exams(MBA, MSc), some selective of which are open to public, just to counter-argue with mis-understanding. Gladly, brave and adventurous students come, but they fail a lot, as the most STEM majors. Because AI/Data Science is not mere coding, but math and science.
Some jealous people go bickering. A dean from one of the Swiss universities for earlier contract told me in the process of negotiation that the school's Wikipedia had been frequently modified as if the school was not legit. He told me that I should be a part of his team to 'counter-attack'. Now, I come to see similar cases quite a lot for my own institution.
What can I do? Should I unring all the bells and close down the school? It's a temptation that I fight every day.