I agree. I think the distance in capabilities between great mathematicians can be so much more vast than other fields as well. Some of them need every step to be derived, while others can skip ten in their head.
If he wants to succeed, he will need to reconsistute the board. That's a tough one since they appointed him, but otherwise it won't work. The type of transformation Intel needs to go through won't withstand a myopic, short term oriented bureaucracy.
Have you measured the difference in speed in moving data to the GPU? For enterprise AI workflows this is a bottleneck to utilization, so improved speed can help reduce compute costs.
This is a great book to learn about the Silk Road in particular Central Asia and the role Russia and UK played in its transformation. It reads like a Game of Thrones novel.
Interesting note is Russia colonized Central Asia with the end goal of invading India.
The payments could increase in the future as this approach becomes more successful and the ROI is better understood. Before this most companies would give into the trolls. Now there may be an increasing incentive to fight back. Few companies have the time, capital and governance freedom to do so.
Something like this is going to be needed to evaluate models effectively. Evaluation should be integrated into automated pipelines/workflows that can scale across models and datasets.
I tend to agree. Cory Doctorow has a good essay on how the deck is stacked against people who don't own homes. This doesnt include some of the health effects of loneliness and other issues when people can't live close to their friends/loved ones.
"The Big Sort" by Bill Bishop is a well researched book that explains the political fracturing of America.
The topic is explains is expressed visually in these two maps, which shows county election results from 1976 to 2004.
http://www.thebigsort.com/maps.php
One of the point is that due to economy, mobility and choice, Americans decide to settle where people are like them the most, and that this results in deeper political boundaries, especially in the context of our electoral system.
One takeaway is that individually we are not all different, and a single Democrat and Republican in a room would likely come up with compromises on differing issues. However when on Democrat/Republican is in a group of many like themselves, they tend to take the most extreme view, on average.
The other is that there used to be more friendly relationships between senior senators on both sides of the aisle. They would gather regularly to have a few cocktails even if they have strongly differing views. These bonds made it possible for them to work as a bi-partisan team to get legislation passed. These relationships no longer exist the way they used to, and many politicians are much more transient in the time they spend in DC.
Yes that is true. You're exposing an assumption of mine which is that many of these companies will not necessarily IPO.
However even when they do, they will be a lot more operating history and financial information than what companies typically had in 2000. This is why you're seeing some recent tech IPOs that went public at a valuation lower than their last private round. An example of this is Hortonworks.
Yes. Not always easy to answer. Some factor of revenue growth, margins, how much capital the company has already raised, understanding the burn rate, how much has been spent vs dry powder, liquidation preferences assumption. Probably a lot I'm missing. There should be an easier way for a prospect to simulate a cap table with a simple set of assumptions.
This type of approach works well in image processing, where you can use C++ for the repeatable, hierarchical patterns and assembly for optimizing some key loops.
One of the reasons that you hear CEO's want a $B valuation is that it becomes "easier to attract top talent". I'm not sure that is valid. You want talent that believes in your mission. Not all $B valued companies have the same risk/reward scenario. It'd be helpful to have some metric of the capital efficiency a company has to get to the $B mark.
The one good side effect is that private investors are taking all of the risk. If there is an adjustment in valuations across the industry, it should not effect the public markets the way it did in 2000. It could however impact the LP market and the number/size of VC funds could go through another cycle.
Society is at a point in time where many forget the reasoning for principles of the judicial system. The presumption of innocence in particular is critically important and written into most modern democracies constitution.
The alternative of being presumed guilty can lead to a lot more corruption and injustice.
Ben Horowitz's actual topic is the challenge of incentivizing management.
His subject is covered very well by one of the best to write about this subject, Charlie Munger (Warren Buffets right hand man and Director of Berkshire Hathaway).
Charlie Munger discusses extensively the challenge of not only incentivizing managers and employees, but demonstrating that it is difficult to even fully understand what really incentivizes them (hint: it's typically different than what their superior thinks it is). He talks about human mis-judgement and its role in distorted intending incentives.
Munger says over 30+ years in business, this is the one area he continues to make judgement mistakes year after year, since this is such a difficult topic to get right. That doesn't mean you can't get it right or that proper incentives don't work however.
Check out the FedEx example in Munger's 1995 speech to Harvard.
The diagram of the Cost of a Petabyte is very interesting if it is true. It demonstrates the profitability of some SaaS models for selling what is a straight commodity.
However it seems in contradiction to the AWS, Rackspace model, which is a race to the bottom in that there are many competitors and they are selling a commodity (independent of the other high value services they are selling). There is some threshold of volume that is key in order to make money in that space.
AirBnb seems like a startup is getting significantly more success. Why this person considers them to be more innovative than many other startups however shows I think a bias towards selecting a winner after its already won. There are plenty of amazing, innovative startups in terms of their vision and technology that will ultimately fail. That's the nature of the business.
I think her problem is she just doesnt have the right deal flow, and doubt she has been part of or witnessed up close a game changer.
There are startups in the Valley that are working on microbial fuel cells, flow batteries, microwave radar, tumour selective drugs, computational photography, nano-controlled membranes for unprecedented efficient water purification, stream processing algorithms, optical computing, immersive UIs, gasification, robust computer vision, RF plasma lighting sources, novel financial transactions, high efficiency combustion engines, compressed air storage, green plastics, ML for genetic applications, next gen EDA for <25nm structures, and of course shoedazzle.
I think the issue is her dealflow and perspective. coming to SV for 6 months assuming you're going to see game changers when most VCs for 10 years dont is poor judgement.