How to fix your kids' education for $200m
Open source brains
"The beauty of K-12 (kindergarten through twelfth grade) is that we know what we know," McNealy says. "We know ten plus ten. We know sentence diagramming. We know modern European history. We know geography. We know all that stuff.
"Anything that is known can be reduced to an on-demand type thing."
With such a premise in mind, paying $130 for a fourth grade math text does, in fact, seem ridiculous.
So, the Curriki backers have created a system where educators can upload their own material. Perform a search for "calculus," and you get seven items back, including "Introductory Calculus 1" from the National Repository of Online Courses and a PowerPoint presentation from a chap named Mike Corsetti covering imaginary numbers.
At present, the searches prove quite thin. Much of the material, in fact, comes from the National Repository of Online Courses, making one wonder if another repository is needed.
The answer is 'absolutely,' according to McNealy, who has set ambitious goals for the Curriki project.
"Right now, we are just in collecting mode and have 3,000 plus assets," he said. "The goal over - let me be clear on the timing - the next twenty years is to get all of this information in a common user interface, a common search format, a common development architecture and a common deployment architecture, so that it all feels like it came from the same place."
Ever the Sun loyalist, McNealy describes the Curriki effort in terms familiar to server customers. He talks about an education-based services oriented architecture (SOA) and discusses a worldwide effort around the education site being governed by something similar to the Java Community Process. And the first step, as mentioned, is to build a Curriki reference architecture - similar to the RAs Sun pitches at customers buying hardware.
To pull this off, the Curriki crowd seeks close to $200m in funding.
"If I could get some rich guys to give us enough money, we could actually go pay some people to get that reference architecture going," McNealy said. "Then, it could evolve from there."
The funding aspect seems key to us. The last thing we need is a bunch of anonymous posters sending in textbooks that served as the basis for a child's education. Curriki would need the best of the best to displace incumbent textbook makers rather than the facile cruft that dominates Wikipedia.
"I don't mind paying people to get even higher quality content," McNealy said. "Some people are smart and like to donate, and some people want to see the money. I don't want to lose either set."
Oversight, of course, would help too. God forbid the playground go unattended like it does at Jimmy Wales's house.
As McNealy sees it, you could end up with professors from universities in, say, the US, China and Scotland being paid to shepherd the calculus "cell" on Curriki. They would ensure that high quality texts and course material were in place and also monitor how well material was received and how well it helped students.
It's this measurement quality that stands as key to McNealy.
"We'll build in feedback mechanisms so that every group competes against all the other groups," he said. "You might find that third grade math seems to be getting better reviews than fifth grade social studies. So, we'll find out what's wrong with fifth grade social studies."
Such a feedback loop could deliver a drastic improvements over the current textbook system where publishers often update texts with a new format simply so they can sell fresh editions of the same, old material. Move a few pictures around, bold some text and voilà, a teacher's lesson plan from 2007 fails to work for 2008 texts, since the page numbering has been skewed.