How to achieve such kind of mental status? The first step is to eliminating distractions. a. Make sure that you sit at a comfortable desk, the seating is ergonomic, there are right amount of light, b. make sure yourself is cleaned, well fed, hydrated and maybe even after some moderate amount of exercise c. close all the irrelevant windows on the computer. All those windows that are opened are related to the project that you want to work on.

The second step is to improve the focus by planning. This includes articulating the core scenario and algorithm, design the core test cases and write the spec. What is the primary scenarios that we want to enable for this? What is the core algorithm of the core scenario? Then you can start prototyping the core algorithm. Afterwards, you may want to design the core test cases for the core scenarios, particularly if you are into the test driven development (TDD). In the meanwhile, you can start drafting a design spec to document the core design decisions, such as selection of the framework that you are using, the pros and cons of the alternative designs, etc. All these prototype, core test cases, design spec will help you piece all these information together into your mind. With these planning efforts, you will be in a good shape of winning the inner game of coding. 

Internet companies creates free contents/services on Internet, they only receive payment when a transaction is performed or an advertisement is viewed or clicked. The ads income is proportional to viewer traffic, the transaction fee is often tied with both the traffic and the level of engagement of the users are. Typical Internet companies include google, facebook, amazon and yahoo. 

There are companies that are in between, those cloud computing companies, many of them selling software/service via Internet. These companies are also very profitable. Typical examples include AWS division of Amazon, and salesforce.com. 

Due to the differences in business requirements, the software engineering practices are also quite different among these companies. For example, for software companies, since many of those software were written in C, ensure the software does not have memory overflow problem is of paramount importance. As a result, these software companies are heavy in quality control, they have large number of QA or SDET or tester. They write tests of all kinds, run the software in all the possible ways to try hit as many code path as possible to make sure the software reaches the quality before releasing to the customer. 

Internet companies come from a different path. These Internet companies came from startup world, where the funding are often just enough to sustain to the next round fund raising. They need to get things done, and faster. Also, many companies of this category were founded at the time compiler / programming languages are a lot more mature than before. Hence, these companies typically have a much shorter release cycle, since the client view the website via Internet, the client do not have direct contact with the software behind, the quality control is often looser.

What is the value of peer review feedback system?

Without peer review feedbacks, some employees may not provide necessary support that their peers needed to complete the business priority of the organization. In order to get good peer review feedbacks, employees are motivated to collaborate with their peers. When the reviews are evidence based and for each piece of feedback, there is a constructive suggestion on improvements. 

Can peer reviews be abused?

Yes, peer reviews can be abused by unethical employees. People can lie about their peers' performance to gain an upper hand in the review. For example, a team has two members, A and B, they have a fixed budget for bonus. A is a high performer, instead of working hard, B may choose to lie about A's performance so that B may get a little bit more bonus without much extra work. 

How can peer review abuse be prevented? 

Just like book reviews on Amazon, buyer/seller reviews on Ebay, peer review should also allow the owner to leave comments on his review. In the above case, for example, when B leaves a comment "A's code is not well designed". A can leave a comment "The design documentation for feature X is located at Y, which has been reviewed by the team". This can largely mitigate unfair peer reviews. 

When does stack ranking system become unfair?

Stack ranking system can become unfair under a few circumstances. The first one is when the actual performance of the employees do not fall into the bell curve. Although statistically, at a large organisation, the performance of all the employees are likely fall into a bell curve. However, at a particular team, at a specific time, there is no guarantee that the performance of them fall into a the bell curve. They could all be performing well, or they could all be performing bad. When all the team members are performing well, they will feel they are under paid, thus this creates unfairness. When all the team members are performing poorly, they still get rewards like high performers at other teams, other high performing team will feel unfair. 

The second situation is when the employees are not competing against each other on productivity. When people feel pressure, depending on their culture, education background, they can adopt dramatically different behaviors, some of them can be counter productive. The feeling that some people on the team are not pulling their weight for the business also makes the rest of the team feeling unfair.  

In a well managed engineering organization, members should focus primarily on how to build the right features with good qualities. Milestones with clearly defined deliverable are achieved with high quality. The features implemented are well aligned with the business and organization priorities. Those members who deliver the right features are compensated according to their contribution to the business. Achieving this level of behavior consistency at a large organization is none trivial, the performance review system, the background of the people executing the strategy, the organization structure and the product itself all play a role in the outcome. 

For the first question, the answer is partially yes. It will be useful on some diseases, but not all. Infectious disease and classical surgery are very unlikely to benefit significantly from personal genomics. The infectious diseases are typically caused by the invasion of bacteria and/or virus. Most of the medicines of combating bacteria are the inhibitors of the enzymes inside the body of the bacteria. Understanding of the variation of the host(human) are very unlikely to help those inhibitors to work better. The classical surgery are also unlikely to benefit from the personal genomics, for example, tooth extraction, the detailed information of tooth growth probably won't help surgery much. Although the genome encodes how the tooth are suppose to grow, the environment will have a impact on how eventually the tooth are going to look like, thus personal genomics cannot replace dental X-ray.

What diseases that personal genomics will likely to be useful? I think complicated genetic diseases are likely to benefit from personal genomics most. Some complicated genetic diseases such as diabetes, rheumatoid arthritis, and cancer. These diseases are the results of faulty metabolic pathway, these pathways comprises multiple components, each of them can be faulty. Diagnostic method at each nucleic acid base pair level are likely to help to pinpoint the faulty parts of the pathway. Fault specific therapeutic strategies are likely to be more effective than a generic treatment.

However, there are still gaps between this vision and the current understanding of those diseases. To accurately identify the faulty part of the pathway, first we need to understand what constitute a healthy pathway, second, we need to understand what variations of the pathway are diseases related. Our knowledge of both questions are limited for most of the genetic diseases. Furthermore, variation specific treatment can be even further down the road after we understand the impact of each type of variation.

We can forsee that personal medicine for genetic diseases is a very tantalizing vision with very long way ahead. In the meanwhile, the technology to discover/store/simulated genetic variations will definitely help realize this vision faster, this is where computer will be extremely useful.

I would say desktop is losing a lot of its mind share due to the popularity of the tablets and smart phones, but there are two types of the computing are hard to be replaced in the near future. The first type of the applications is those applications that intensive in graphics, such as first person shooting games, simulation games, video editing, 3D CAD software, etc. The graphics cards on mobile phones are not going to get a lot more powerful in near future due to the consumption of the energy  and the battery life of the tablets. Those graphic intensive applications are not going to move to mobile. Moving to cloud is possible but that is going to take a while. 

The second category of the applications that desktop application are going to continuously dominant are those applications require specific ergonomic needs, for example,  Microsoft Word. Some writers need to write/edit documents for 6-8 hours per day, including frequent typing on the keyboard,  tablets are too small to type for a prolonged period of time. For example, Microsoft Visual Studio, software engineers need desktop and big screens to look at code, edit the code, run test them. This is also very unlikely to go mobile in the foreseeable future. 

For IaaS, there will be two areas of the innovations are arising, the first area is access control of IaaS. IaaS basically provide computing resources on demand to the customers, the billing management of the IaaS resources remains lackluster among the existing providers. Allowing flexible billing methods such as invoice and control of the billing according to organization structure are going to make the IaaS more usable to customers of various organizations. The second area of the innovation is the diversity of the VM types. Currently, most of the IaaS providers provide a few variations of CPU cores and memory size, the customer optimal virtual machine may be of a specific combination of CPU cores and memory size that are not in the pre-set, we can expect some cloud providers are going to provide more customization of the VMs, including CPU, memory and even GPU, sound card or bandwidth.

For PaaS, we are going to see improved consistency between in cloud and off cloud development experience. In the first generation of PaaS frameworks, developers write code on their own machine, and upload the code along with related resources to the cloud. "Hope" the same cloud runs on local machine will work well in the cloud. Increasingly, the discrepancy of the behavior of the code between on-premise and in cloud are driving cloud provider to provide strong in cloud debugging and tracing capabilities, the boundary between in cloud and on premise development are becoming smear over the time. 

For SaaS, in 2013, there are already very very successful SaaS providers such as Sales Force, github, these websites sell software experience to end customers via the Web. Customers no long need to purchase the software binaries to use them. We will experience continued growth of this trend, more and more enterprise software will adopt this path to grow because it simplifies the supply chain of software distribution and it makes the software itself platform agnostic. 

2014 is going to be another great year of cloud computing. 

The salary for the physicians are different. 

In the United States, a physician can make 150K to 300K  depending on his/her speciality. That is about 3 to 6 times household income in US. In China, a physician is typically paid with 40K RMB per year, which translates into 7K  USD per year. The average household income at China is about 10k per year, thus you can see doctors in China just make the average income when their American colleagues are making a much higher salary even if we factor the geographic differences into the consideration.

The education system for the physicians are different.

In the United states, a physician needs to take six to seven years training on top of his undergraduate education before starting his own practice. The cost of the medical training is high which often incur significant student loan for students with limited financial support from the family.  

In China, the medical training is much shorter, a MD can be obtained after five years training after high school. It is three year basic science training with two years clinical practice. This may also contribute to the lower wages that physicians have in China. 

The business models are different.

Comparing with China, there are more physicians practicing in private in US. This is due to culture and historical reason. In Mao's era, private practice is considered a small business which does not fit into the ideology of communism party. Physicians typically work in the large public hospital and considered government employees. In the United States, many physicians work as individual practitioners after obtaining his/her medical license. Even large private hospitals are popular. 

The pharmacies are different 

In US, the pharmacies are typically independent from the hospital. In China, the pharmacies are a part of the hospital and drive a significant portion of the revenue of the hospital. 

The workflows are different

In China, the workflow of patient administration and discharge are different. Chinese hospitals are pretty much like government agencies, when you go to hospital, you feel like to to getting a drivers' license at the United States. You go to a waiting line to get a number and when it is your turn, then you go to your physician's room to get care. 

In the United States, in most of the private clinics, you got a receptionist who let you wait without a number or expected waiting time. Depending on your luck, you will likely first get the attention of a nurse who is going to measure your vital signs regardless of your condition. The nurse will take notes of the patients' condition. After some serious waiting, the physician will appear, finalize the diagnostics and prescribe procedures and medications. 

All these differences contribute to the difference of the cost of the care and the quality of the care. These also elude some cues on potential options to reduce the cost of care at US via technology. 

a. reduce the cost of medical education using technology and combined program, expand the labor force of medical doctor. 

Medical education require hospital facilities which incurs the expense of medical education, by using distant learning, virtual reality laboratory, the cost of medical education can be reduced to a factor of the current price. We should also provide the option for those students who are determined to be medical doctor to start their medical training earlier. Instead of starting medical school after four year college, medical education can be combined with undergraduate education as eight year BS/MD degrees.  More medical doctors should be trained so that high quality of care can be served for more people.

b. IT based workflow optimization to reduce operation cost.  

The waiting time at US is much longer, and the information gathered by the nurses are redundant, there are just too many front desk clerks at US hospital than necessary. All these can be streamlined using technology. Patient can simply just checkin at the front desk of the hospital using a kiosk, after admitted to the waiting room, most of the patient can measure/report their vitals using computers by themselves, afterwards, they can interact with the physicians directly. Content management software can be employed to facilitate physician/patient communication so that during those downtime, the patients can get information that are pertaining to his/her condition, thus the physician can spend less time talking. The improved workflow can reduce the cost. 

c. Improve the liquidity of the healthcare market using online review system. 

Yelp was originally designed to be a healthcare review system, however, over the time, the more popular use case is probably restaurant. It lacks some healthcare specific features such as the procedures performed by the physician, the rating of each procedure, the price of the procedure, etc. How many times have you run into a doctor who wants to do some procedure that is not right for you?

The net is healthcare IT can be very very useful in reducing the cost of the care and at the same time, may also poise some great business opportunity. 

The biggest problem of these record system is the fragmentation. Quite a few of these systems were designed prior to the dot com boom, thus the interoperability of the EMR systems are terrible if not horrible. If you switch from one hospital to the other, you are likely to lose most of your record if you do not explicitly request a transfer. HealthVault is presumably in the position to defrag the EMR systems, however, the progress has been lackluster since its inception. The reason is in order to integrate with HealthVault, EMR providers have to take the action to perform the integration, however, the customers of the EMR systems typically do not have a financial incentive of sharing the healthcare data.