Netizen

Thursday, September 11, 2014

Neuromorphic Chips

Dr. Isaac Asimov, head-and-shoulders portrait, facing slightly right, 1965 (Photo credit: Wikipedia)

Is this what you see on the way to Singularity?

Neuromorphic Chips

Traditional chips are reaching fundamental performance limits. ..... The robot is performing tasks that have typically needed powerful, specially programmed computers that use far more electricity. Powered by only a smartphone chip with specialized software, Pioneer can recognize objects it hasn’t seen before, sort them by their similarity to related objects, and navigate the room to deliver them to the right location—not because of laborious programming but merely by being shown once where they should go. The robot can do all that because it is simulating, albeit in a very limited fashion, the way a brain works. ..... They promise to accelerate decades of fitful progress in artificial intelligence and lead to machines that are able to understand and interact with the world in humanlike ways. Medical sensors and devices could track individuals’ vital signs and response to treatments over time, learning to adjust dosages or even catch problems early. Your smartphone could learn to anticipate what you want next, such as background on someone you’re about to meet or an alert that it’s time to leave for your next meeting. Those self-driving cars Google is experimenting with might not need your help at all, and more adept Roombas wouldn’t get stuck under your couch. “We’re blurring the boundary between silicon and biological systems” ...... Today’s computers all use the so-called von Neumann architecture, which shuttles data back and forth between a central processor and memory chips in linear sequences of calculations. That method is great for crunching numbers and executing precisely written programs, but not for processing images or sound and making sense of it all. It’s telling that in 2012, when Google demonstrated artificial-intelligence software that learned to recognize cats in videos without being told what a cat was, it needed 16,000 processors to pull it off. ..... “There’s no way you can build it [only] in software,” he says of effective AI. “You have to build this in silicon.” ...... Isaac Asimov’s “Zeroth Law” of robotics: “A robot may not harm humanity, or, by inaction, allow humanity to come to harm.” ..... glasses for the blind that use visual and auditory sensors to recognize objects and provide audio cues; health-care systems that monitor vital signs, provide early warnings of potential problems, and suggest ways to individualize treatments; and computers that draw on wind patterns, tides, and other indicators to predict tsunamis more accurately.

Wednesday, September 10, 2014

Tech Startup Equity Distribution

English: Defunct Belgian shares in a Russian Tram company for the city of Saratov. With the Russian revolution these shares became worthless. Nederlands: Oude Belgische aandelen in een Russische trammaatschappij in Saratov. Na de Russische revolutie zijn deze waardeloos geworden. (Photo credit: Wikipedia)

100% is like the speed of light. The total has to add up to 100%.

33-33-34

33 to the Cofounders, 33 to the investors in all rounds all the way to IPO, 34 to the team.

When you start a company, you probably have two Cofounders, or three. One is often the Senior Cofounder, the possible Founder CEO. It could be 10-10-13, with 13 going to the Founder CEO. Or it could be 5-10-18. Or they could agree to cap themselves at 30% and give another 3% to the team. And so, 5-7-18.

But the investors have not come yet. The team has not been built yet. The Founder, or the Cofounders could keep the rest of the equity in a caretaker capacity. So, 5-7-18 could become (5+15)-(7+25)-(18+40), with the caveat, should a Cofounder leave the company, the equity held in caretaker capacity reverts back to the company. Even the 5-7-18 should have a vesting period of, say, five years. So if the Cofounder at 5+15 leaves after one year. The 15% goes back to the company. And he/she walks away only with 1%. Because the other 4 has not vested yet.

Investors

I am for an anti-dilution clause. Say you raise 50K in your first round to give away 5% of your company at a million dollar valuation. The anti-dilution clause makes sure that 5% stays 5% in all future rounds. Otherwise, without the anti-dilution clause, a round 1 investor could put in money, feel like he/she owns 25% of the company, but by the time there is an exit, that person ends up seeing that the 50K they invested is still 50K when they walk away from a successful startup's exit. I think that is a shame. And it happens all too often.

Say, it is 30-35-35. Then 5% is gone. And 30% to go.

Round 2 you raise 500K at a 10 million dollar valuation. That is 10% gone with 25% to go. Round 3 you raise 2.5 million at a 50 million dollar valuation. That is 15% gone. There has been an anti-dilution clause at each stage. That makes each round of fundraising independent of each other.

You go IPO at a 100 million valuation. You raise another $5 million at that valuation. That is 20% gone. Hopefully your company keeps doing well, and you raise another 50 million dollars at a billion dollar valuation.

The Team

This is of utmost importance. You have 35% to give away. And you have to assume some of your best people might come years later. The formula is, the higher up you are, the earlier you come, the longer you stay, the more you get. And here you want to move to number of shares as soon as possible. Because 0.1% feels like little, but 100,000 shares feels like a lot.

Your team could be five deep, or 10 deep. The Board is layer 1. The CEO is layer 2. The executive committee (COO, CTO, CFO) is layer 3. All three are not equal. The CTO might get more. The CFO might get less than the COO. Depends. Assume there are five other layers all the way to the receptionist.

I think tech startups should pay below market rate salaries as a rule. You only want people who really see the value of your equity. And the equity you give out should vest over five years.

The Board we took care of. The Founder CEO we took care of. Assume the executive team is not any of the Cofounders. Say the CTO gets 2.5%, the COO gets 2%, the CFO gets 1.5%. To vest over five years. That is 6% gone. But a COO coming in in year 3 would get less, something like 1.5%, or even 1%. Also to vest over five years. Unless you get a star CFO in year 2 who is good enough to take the company IPO, then you can be a little more generous. How about 1.5% to that CFO in year 2?

A superstar CTO might be worth 3%. I don't know.

The thing about the lower layers is there are more people in those layers. That further reduces the size of their cake.

Layer 4 might get 3% total. Or you might say, 3% for each layer with a cap of a certain percentage points for each layer.

Layer 4 gets 3% total, but an individual may not get more than 0.5%, something like that. But if there are 10 people in layer 4, they get 0.3% each on average. Some might get 0.2%.

Layer 10 people might get 0.05%.

Number Of Shares

At a 100 million dollar valuation, the company has 100 million shares to give at a dollar each in value. So at a two million dollar valuation, the company shares are worth two cents each.

So 0.05% in equity is actually 50,000 shares, which is a lot. The junior most members of the team will probably be looking at something like that, again, to vest over five years. You don't have to buy them. You just have to stick around, work hard, and your shares keep vesting.

Tuesday, September 09, 2014

Hello iPhone 6

Via Vic Gundotra

Brain Mapping

Medial view of a halved human brain, labeled in latin. (Photo credit: Wikipedia)

The brain is the true Final Frontier, the most unknown part of the universe. But this still ain't it.

Brain Mapping

At the start of the 20th century, a German neuroanatomist named Korbinian Brodmann parceled the human cortex into nearly 50 different areas by looking at the structure and organization of sections of brain under a microscope. “That has been pretty much the reference framework that we’ve used for 100 years,” Evans says. Now he and his coworkers are redoing Brodmann’s work as they map the borders between brain regions. The result may show something more like 100 to 200 distinct areas, providing scientists with a far more accurate road map for studying the brain’s different functions.