Last week, Cleversafe introduced a major update to its open source software. Cleversafe is a company with a big, hairy, audacious goal: transform the infrastructure for storing data. All data. Enterprise data. Consumer data. Internet data. Transform the storage of data by radically simplifying the underlying method of storing it and meanwhile making it much more secure and reliable, not to mention cheaper.

Like many startups, Cleversafe has gone through significant transformations itself as it has figured out how to navigate toward this goal. A key decision the company made early on was to put its core software into the open source domain. But it realized earlier this year that it had to rebuild that core software to handle much greater diversity in its prospective customer set. The result: Software that allows the customer to decide for themselves how reliable and secure they want their storage versus how inexpensive. As always, technology requires you to choose between cost and benefit; in Cleversafe’s case, it’s a relative decision because even the most reliable and secure application of the technology is still considerably less expensive than existing, mainstream choices.

Cleversafe’s basic proposition is that you can obviate the need for expensive, high-performance storage by using the Internet itself instead of your own disk farms. It proposes that you carve your data into N numbers of “slices” and then store those slices in discreet locations on the Internet, instead of storing all of your data in one location and then protecting that data by backing it up and archiving it. Cleversafe’s software manages those slices of your data in a way that means you can always get all of your data back even if some percentage of the discreet locations fail. The result, mathematically speaking, means storage is more reliable and more secure by spreading it around the open Internet than it can ever be by trying to maintain close control in your own infrastructure.

That’s an amazing idea; a big, hairy idea. It’s an idea that will take customers some time to absorb and adjust to. Cleversafe is right now identifying which customers will appreciate that idea the most quickly and be willing to pay the most for it as it plans to deliver a product to customers early next year.

Cleversafe holds a dear place in our greedy, venture capital hearts because the terms we offered for investment in Cleversafe were the first terms we offered to any company. We offered those terms before we had raised our fund. We were able to do that because the investment involves old friends on both sides of the investment: Chris Gladwin, the founder and now chairman and CTO of the company, was also the founder of a company where Stewart lead an investment in 1998, MusicNow. And the investor in that case, New Enterprise Associates, is also our co-investor in Cleversafe. Both Chris and NEA agreed to support our investment in Cleversafe before we got our first commitment to fund Alsop Louie Partners. Since we did get funded, NEA and Alsop Louie Partners along with the angel investors in the original funding of the company provided a Series B financing for Cleversafe in November, 2006.

This past week, Justin.tv launched its live broadcasting system. Read the Los Angeles Times coverage of the company.

Everyone’s first question to us: How the heck are you going to make money on that investment? Good question! Answer: We don’t know, exactly. We have some suspicions.

But Justin.tv is the emblematic deal for what we believe about venture capital: The more you think you know about how you’re going to make money in an investment, the less likely you will either make money or make it the way you think you will. In other words, risk is the sine qua non of real venture capital. We prefer to bet on the people behind the venture more than our belief in the outcome.

Justin.tv is all about the live in “live broadcasting”. Everyone knows that video is now, finally, a big deal on the Internet. But the video that is a big deal is recorded video: clips on YouTube (or a bunch of other sites that aggregate or distribute video) and movies and TV shows on iTunes or Amazon.com or wherever. Justin.tv specializes in live video. What impressed us and lead us to invest in the company is that the co-founders didn’t just turn on a web site that let people broadcast live to whoever is watching; they designed a system that allowed anyone to broadcast live 24 hours a day, 7 days a week from anywhere that has a reasonable cell phone signal: the most extreme version of live broadcasting.

This accomplishment was first shown in alpha versions earlier this year by the eponymous Justin Kan attaching a web camera to his hat and committing to broadcasting his life, or “lifecasting”. Out of that experience, Justin and his co-founders, CEO Michael Seibel, CTO Emmet Shear and VP Engineering Kyle Vogt, figured out how to build an end to end system that made broadcasting a video signal with even a marginal cellular signal tolerable and, most particularly, made it possible to broadcast high-resolution video both functionally and at extremely low cost. On top of that, the team has built a system for letting viewers and broadcasters cooperate in real time to make sure that the most interesting stuff being broadcast gets promoted to the “front page.”

Call us crazy but we think that a business that can reliably and systematically show people the most interesting things going on in real life in real time anywhere in the world might well be a really popular destination on the Internet. I like to think of it like this: If you can get a million people out of the six billion in the world to create a live broadcast at any one time, there’s bound to be something interesting enough to get a significant percentage of the people sitting in front of their computers to check in and see what’s happening.

Indeed, we believe that Justin.tv could well represent the first instance of a new form of media. With each new turn in technology comes a new form of media. Usually the first thing people do with a new technology is to adapt the old thing. When movies where invented, everyone stuck a camera on a tripod and filmed plays. The first radio broadcasts were simply the audio for staged plays being sent over the airwaves. We think YouTube represents early broadband video, essentially redistributing home movies and other recorded videos. Justin.tv is the first instance of a new media.

As always, there is plenty of competition for Justin.tv including Ustream, Stickam and others. In our case, we have bet on an extremely talented team of youngsters who have shown clear vision and motivation to build a business. Alsop Louie Partners lead the Series A financing of Justin.tv, which closed in August, after the company was seed-funded by Y Combinator and other angel investors and bootstrapped by the founders in an apartment in San Francisco. Check out the team on their own office camera.

“If you show me yours, I’ll show you mine.” That’s how I describe Cake Financial to people who haven’t heard of it before. The essential idea of Cake is to use modern web technology and standard data formats to share investment history with like-minded people. More simply, Cake is a way to publish your stock portfolio.

Why in heck would you want to tell everybody how you invest your money?!?! Because it will help you be better at investing and make more money. If you are an investor in public companies, you have a choice: Make your own decisions or trust someone else to do it for you. If you make your own decisions, it requires a lot of work to be any better organized than throwing darts at the stock tables. If you trust someone else, you rapidly find out that “experts” often do no better than monkeys (who are throwing darts at the stock tables). So it’s a conundrum, one that actually leads most people not to invest in stocks at all. They might buy into mutual funds, which are large accumulations of stocks (or other investments) where you trust the fund manager to make your decisions for you or just give up and leave the money in a money market fund or CD or whatever represents as little risk as possible, so you can live the rest of your life without worrying about losing your money.

Cake is a system for identifying who actually does know what they are doing – using actual transaction data – and then using their expertise to make your own decisions. For instance, if you find that Joe Blow is really good at creating a risky, high return portfolio of stocks. You could just decided to make the same decisions Joe makes. (Eventually Cake will automate that, so you can let Joe make your decisions for you, which gets really interesting.) And if John Smith is really good at getting steady, long term results without taking chances, you could have half your account traded like Joe Blow and half traded like John Smith. The possibilities become endless; indeed, the brilliant insight of the company is that, if you can get people to share their trading history, you can accumulate an accurate history of how to actually make and lose money, rather than guessing and forecasting.

The team behind Cake believes that this company has the potential to define a new approach to investing in public stocks. We don’t think that we can easily get people who have been investing in stocks for a long time to break old habits, so we’re targetting investors who are younger and newer to the game and who don’t believe that the existing brokerage firms or the advisors they hire are trustworthy. There’s 40 million people in Generations X and Y with about $5 trillion in investable assets; a big percentage of these investors like to do their own research and make their own investment decisions. We plan to become their best investing tool.

There is already competition; a company called Covestor, based in London, announced a very similar service before Cake launched. But they have to use an intermediary, Yodlee, to link to brokerage accounts; that means they can’t get the entire account history, which means they have limited transaction data from which to calculate performance. Cake has its own technology for linking to accounts and getting the source data; we think (obviously) that makes Cake a better service for our customers…

Alsop Louie Partners lead the Series A financing of the company in January and helped the company define the product, which was introduced earlier this month at TechCrunch40. If you’re interested in learning more about the company, registration is now open to the public for what it is calling a public alpha test: There are two steps, registering for the site (which lets you see a little bit of data about what people are doing), and link one or more brokerage accounts to that Cake can get and normalize your trading history (which lets you look at individual’s portfolios and compare performance).