Education in the Internet Age
Getting an MBA to learn about doing a startup is a bit like getting a Physics PhD to learn about riding a bicycle. Startups are not spreadsheets. MBAs can be useful for scaled businesses. Good MBAs can execute and have eye for numbers. But their compensation requirements and skills are often not startup friendly.
Many people considering Business school could learn much more by taking that $100k+ and starting a company. Be a CEO instead of just reading about them. If you are admitted to a top grad school, you can likely defer admission and parlay your acceptance letter into fundraising for your company. Found a startup. If it fails, go to school.
Save the money you would have wasted on your children’s private schools and even college, and give it to them as a lump sum when they reach the age of majority. That becomes their seed funding. Far better than incinerating money on an increasingly useless diploma in a rigged higher education system.
Personally, I wish I hadn’t done my Ph.D. It wasted much of my 20s. I learned more material faster after becoming a founder. One can say the knowledge to do computational biology wasn’t widely distributed in the 2000s, and I’d agree. But it is today for the resourceful.
Degrees don’t provide proof of skill.
A college diploma is a very, very coarse illustration of somebody's skills. If somebody says they have a Harvard degree, or a Stanford degree–what does that really mean? You have some estimate for a Stanford degree in computer science. You kind of know what the curriculum is, but for this particular person, you don't know what skills they have. All the coursework they did during undergrad has been basically wasted.
Imagine if instead they had a portfolio of every problem they solved. With 25 courses, 10 weeks per course, and 10 problems per week, all 2500 problems could be organized in folders on GitHub. You could see their portfolio and drill down to see the particular skills each person has. He knows operating systems, she knows compilers, this person knows machine learning. I can see the evidence by the 15 problems they've solved in their code online.
Especially for broad disciplines like computer science, electrical engineering, and chemical engineering you can't presume that because somebody has a degree they have a particular skill set. People have strengths in different areas. Once you can see individual measurable skills, you can see just how much talent or background somebody has.
If someone is actually an expert in an area, they will have the badges to prove it. As you learn online you will earn cryptocredentials like badges in your wallet, which reflect your ability to do certain things. For example if you complete five Python problems you get a badge that says you completed five Python problems. For answering 10 questions on Russian history in Russian, you get a credential for the Russian language.
You’ll keep those badges in your crypto wallet, and use your wallet just like you log into a website today. Your wallet is your portable crypto diploma and credential. You could log into a job board and just start working because you're automatically qualified on the basis of your credentials.
The long-term plan here is for anyone who meets the requirements to login based on their task history, contribute work, and get paid for it.
Labor liquidity is far below what it should be. In theory, a next-generation version of LinkedIn should continuously be finding you opportunities, and not just full-time work–a constant stream of possible tasks, missions, and jobs. This is a hard problem because it seems to subsume everything from job boards to MechanicalTurk to consulting gigs to Uber/Lyft. But I can think of a few z-axis approaches.
This kind of thing is important because it totally destroys the concept of a college diploma.
*Note: This is a bonus section from The Anthology of Balaji, which was edited out of the final published version. Enjoy this section and join the email list for updates, new material, and upcoming products.
