Jekyll2024-01-11T18:31:24+00:00https://www.christopherhwood.com/feed.xmlChris WoodChris Woodchristopherhwood@gmail.comRag Vs Big Context Windows2023-11-09T00:00:00+00:002023-11-09T00:00:00+00:00https://www.christopherhwood.com/RAG-vs-Big-Context-Windows<h2 id="is-smaller-more-focused-rag-still-the-way-to-win-at-scale">Is smaller, more focused RAG still the way to win at scale?</h2>
<p>At this point, most people who have been following generative AI discussions have probably heard of RAG (retrieval-augmented generation). For those who aren’t familiar, briefly it means you pull data from elsewhere and feed it through the LLM to get your final result. This is what we built at Earlyworm, a RAG system for turning online news and blogs into language learning material.</p>
<p>Recently there’s been some doubt about the future of RAG. Context windows are bigger now (gpt-4-turbo has 128k, Claude 100k). OpenAI is releasing their Assistant API that manages RAG for you. Is there value in building your own RAG for your AI app anymore?</p>
<p>While larger context or managed RAG might work for the lowing hanging fruit applications like summarizing a long PDF or single-threaded chat applications, I see a lot of value still in the custom RAG development.</p>
<p>AI apps have the following 4 distinguishing elements:</p>
<ul>
<li>
<p>Proprietary data (in training set and available for inserting to context)</p>
</li>
<li>
<p>Exactness of response</p>
</li>
<li>
<p>Cost</p>
</li>
<li>
<p>Speed</p>
</li>
</ul>
<p>Most startups aren’t going to have proprietary data, so that limits advantages to exactness of response, cost, and speed. In practically all of these, larger context windows and managed RAG lose out.</p>
<p><strong><em>Exactness of Response</em></strong>
Exactness of response is a moving target. Depending on who your user is, what time it is, what they are doing, etc. the definition of what is exact may differ. That means there’s no one size fits all generation technique. Unless the answer you’re looking for is a general summarization of a very long document, then bigger context windows are not going to give you a more exact response. Studies show that lazily packing irrelevant content into a large context window and relying on the LLM to figure it out performs worse than RAG.</p>
<p><strong><em>Cost</em></strong>
It goes without saying that as far as LLM service costs go, when you’re paying for tokens then sending bigger contexts is going to be more expensive than smaller contexts. Managed RAG might be cheaper for applications with less traffic, but I am not sure and initially skeptical that the savings scale better than hosting your own RAG.</p>
<p><strong><em>Speed</em></strong>
Sending more tokens means it takes longer to get a response. There are many studies that show even tens of milliseconds can influence retention and timespent at scale. For startups who want to differentiate by serving faster responses, using less tokens is going to be one of the easiest ways.</p>
<p>As an aside, one of my main workflows for ChatGPT is breaking a large project into tasks, and then smaller tasks until I can convert tasks to code. I get so much value from this workflow that I’m building some internal tools for Earlyworm to make it easier, if you’re interested in trying it out as well just send me an email.</p>
<p>What’s interesting to me is the recursive nature of this workflow. It seems like this is not only how individuals operate but also how larger organizations operate. Vice Presidents set high level, abstract goals like “Release a record breaking product.” Directors and senior managers turn this into more concrete tasks based on their focus areas - like “Limit major bugs to x”, “Respond to customer messages in x minutes”, “Drive x% of sales through strategic marketing initiatives”, etc. Then managers and team leads break these into projects for teams of 3-5, like “Limit bugs by improving automated testing”, “Limit customer messages with improved help center material and onboarding”, “Increase sales through strategic partnerships with influencers”.</p>
<p>Within each project, each individual contributors break the major tasks into a series of smaller tasks, and on and on until the tasks are easy enough to focus on and complete in just a day or two.</p>
<p>To me, this looks something like LLMs with large context windows figuring out general strategy, LLMs with a little more focused dataset and maybe slightly smaller context windows putting together a sub-strategy, and then progressively focused LLMs breaking these into projects, tasks, and subtasks until finally the tasks are converted into the final products - code, emails, social media posts, marketing materials, etc.</p>
<p>While completely automated “AI agents” are appealing in a technical and magical sense, I highly doubt we are able to create such sophisticated pieces of machinery from non-deterministic LLMs. For the time being, I still see the best AI products being those that have a human-in-the-loop to course correct and guide the AI.</p>
<hr />
<p>LLM-related reading from today, focused on how we get bigger context windows without needing to make new models. I don’t understand all of the math but with some work you can grasp the concepts and implications for people building on top of LLMs:</p>
<p>Tricks you can play to expand the context window of an LLM:</p>
<p>https://blog.gopenai.com/how-to-speed-up-llms-and-use-100k-context-window-all-tricks-in-one-place-ffd40577b4c</p>
<p>Good explainer on Rope: https://blog.eleuther.ai/rotary-embeddings/</p>
<p>A post about RoPE: https://medium.com/@ddxzzx/why-and-how-to-achieve-longer-context-windows-for-llms-5f76f8656ea9</p>
<p>Pinecone blogpost on why RAG isn’t dead: https://www.pinecone.io/blog/why-use-retrieval-instead-of-larger-context/</p>
<p>ALiBi: https://openreview.net/pdf?id=R8sQPpGCv0</p>Chris Woodchristopherhwood@gmail.comIs smaller, more focused RAG still the way to win at scale?Emerging Personalized Economy2023-05-03T00:00:00+00:002023-05-03T00:00:00+00:00https://www.christopherhwood.com/Emerging-Personalized-Economy<h1 id="implications-of-the-emerging-personalized-economy">Implications of the Emerging Personalized Economy</h1>
<p>New technologies are making it easier and cheaper to get products and experiences that are tailored just for you. These include things like AI, augmented and virtual reality, digital goods, and 3D printing. These changes can impact how we see the world, get news, and even communicate with others. This blog post will talk about how these changes affect our economy and society.</p>
<h3 id="economic-implications">Economic Implications</h3>
<p>These new technologies mean that custom products are becoming cheaper and easier to make. We’re already seeing this with services like Google and TikTok, which let niche businesses tailor their offerings to people worldwide, enabling businesses that weren’t previously possible. Now, 3D printing is making it cheaper to produce custom goods since inputs can be commoditized and digital designs will be abundant. Like digital designs for 3D printers, digital goods in AR/VR also have no marginal cost and will be available in mass.</p>
<p>In the future, it looks like everyone will be able to have personalized digital and 3D printed goods. This is so different from how the world works today, where centralized corporations like Ikea and McDonalds rule due to their lower costs. Instead of big companies making products for everyone, we might see more small businesses and artists making products for specific groups of people. This is a huge win for consumers who get both low cost and customization, and arguably a win for most laborers who may be able to work independently or in smaller organizations rather than massive corporations.</p>
<p>This change means workers will need to adapt. Instead of working for a big company, they might end up working for small businesses or even starting their own. The way we make and consume goods and services is changing, and this could have big effects on businesses, workers, and society.</p>
<h3 id="social-implications">Social Implications</h3>
<p>With more personalized goods and services, people will also start to see and share things in a new way. Before, you might have kept your opinions to yourself or shared them with a small group. But with new technology, you could start to express and reinforce your viewpoints even more through the things you consume.</p>
<p>This shift will impact our daily life, but also our culture and society. We will have more choices, and a more diverse group of people will be able to contribute to the market. This can lead to more innovation and a rich culture full of unique products.</p>
<p>However, there could be downsides. As we consume more personalized content, we might start to lose shared experiences. This could isolate us within our own echo chambers and lead to more polarization and fragmentation in society. We’ll need to actively seek out different perspectives and engage in open conversation.</p>
<p>Also, with fewer large employers, benefits like healthcare could be disrupted. We’ll need to find new ways to provide these services and keep our communities together.</p>
<h3 id="opportunities-for-startups">Opportunities for Startups</h3>
<p>Startups can help by creating platforms that allow artists and small businesses to sell their goods and services. They can also help design and produce these items.</p>
<p>To deal with the issues of fewer large employers, startups can create innovative solutions to replace traditional jobs. For example, they could partner with communities to provide healthcare or education, or they could create platforms for freelance work with benefits like insurance and retirement savings.</p>
<p>Startups can also help address the potential downsides of a more personalized economy. They could create platforms for open dialogue and exposure to diverse viewpoints, or tools to combat isolation and fragmentation from personalized media.</p>
<h3 id="investment-ideas">Investment Ideas</h3>
<p>Investors might want to look at companies that are creating personalized experiences using AI, AR/VR, and 3D printing. They could also look at companies that are disrupting traditional businesses and providing digital design repositories and markets. There is also a need for companies building tools for creators and ways for people to maintain social connections.</p>
<p>Investors should also consider startups that offer solutions for education disruptions as degrees from top universities may matter less when competition for jobs at large-scale companies decreases. Additionally, startups that offer innovative solutions for pooled resources like healthcare and other benefits will be in high demand.</p>
<h3 id="conclusion">Conclusion</h3>
<p>Personalization technologies are changing our economy and society. More and more, we’re seeing personalized goods that are cheap and easy to make. This is creating a more diverse and customer-focused economy. It’s also changing how we see and share things. But we need to be careful about losing shared experiences and we need to find new ways to support workers. It’s important that we keep having open conversations and seeking out different perspectives to make sure our society stays cohesive and inclusive.</p>Chris Woodchristopherhwood@gmail.comImplications of the Emerging Personalized EconomyThe Difficulty Of Becoming A Manager2018-01-07T00:00:00+00:002018-01-07T00:00:00+00:00https://www.christopherhwood.com/The-Difficulty-of-Becoming-a-Manager<p>Nearly every new manager makes the same mistakes.</p>
<p>She’ll try to compete with those whom she’s supposed to be managing and try to continue being the team superstar. She’ll horde the difficult work because others on her team are “not capable” or “mess things up.” She’ll get jealous when others compliment one of her team members’ work, because she should be the one receiving the compliment. And she’ll get stressed beyond belief as she tries to do the entire team’s work by herself because “she’s the one who’s held responsible now.”</p>
<p>Getting promoted to manager is a weird thing. It’s actually completely different from whichever position you were promoted. It’s its own beast. You used to be a doer, and you got all those dopamine boosts from doing things well and being complimented. But now… now you just get told whenever things are going wrong. And most of the time it’s not even something you did wrong, it was those dastardly team members you’re supposed to be managing.</p>
<p>Whenever something is wrong, your first instinct is to jump in and fix it yourself. Because you were informed of the problem, you assume it’s yours to fix. That’s how it used to be. And you used to just fix it and move on. But not anymore. You’re going to hear about every mistake that every one of your team members made. If you try to fix all of them you’re going to quickly find yourself overwhelmed beyond belief. At first maybe it will feel good, you got promoted and promotions are supposed to be hard, so this is just how it feels, right? Wrong. You’re going to burn yourself out faster than you know it and you’re going to stunt your entire teams growth in the process and quite possibly put a big dent in your team’s chemistry if you don’t fix this and soon.</p>
<p>Your job used to be to write code / write drafts / other entry level work. Now your job is to strategize and work on the logic that holds the code / paper / more complex thing together, to check your team members’ work, and to learn how to work with all of those personalities that make up your team.</p>
<p>I know, you get it. I thought I got it too. I too read things like this, I was even lucky enough to have my supervisor point out to me that I was doing too much and not delegating enough. But it wasn’t until almost a year later when I saw someone else going through the same struggles that it finally hit me.</p>
<p>After my supervisor recommended I delegate more, I took his advice to heart. I started letting other people take the bigger, more critical projects. I would help them to plan the work, then do other things while they wrote the code. Once it was done I would check it, but I wasn’t good at giving constructive feedback. I was afraid of hurting people’s feelings and wasn’t sure that my feedback was even 100% correct. Well, it turns out most of the time most people aren’t 100% correct, and even if they are they are probably still not sure of it. But by kicking the can down the road and not giving better feedback, I helped accumulate a lot of technical debt and failed to help my team members grow. Later, when some of my team members interviewed at other companies, they were bewildered at what was expected of someone with their level of work experience - and that was mostly my fault for failing to push them because I wanted to be nice.</p>
<p>Another problem I faced was feelings of competition. I used to be in the team and now I was leading the team. In the past I received a lot of praise for my work, that’s how I managed to get promoted. But now when I saw my team members receiving praise it weirdly bit at me. I wanted that praise and to not receive it made me feel like I was slacking. Besides, I felt that I was better at the work they were doing than they were - why were they getting praised for subpar work?</p>
<p>This whole mindset is completely backwards. Once I became a manager, I was no longer doing my previous job. I was still in close proximity, which makes the edges of responsibility less clear cut, but in having these jealous and competitive feelings it only proved that I hadn’t realized yet what my new job was. I should’ve been delighted that my team members were receiving praise. I should have hopped on the bandwagon and praised them as well. In fact, I should have been the first to dish out praise. My team members getting praised meant that I was doing my job. They were growing and getting better. They were filling the gap that was opened when I moved out, just like I was filling the gap ahead of me, and so forth. If things keep going well, then maybe one day one of my team members will be set to get promoted to my position. And that will be great news, because that means I’ve done my job well, and it’s probably time for me to get promoted as well!</p>
<p>So, if you just became a manager or are about to become a manager, remember to work with your team, not against them. Remember that you have a new job, and it’s time to let someone else do your old job. And remember that you are all working together for the same goal, not competing!</p>Chris Woodchristopherhwood@gmail.comNearly every new manager makes the same mistakes.Why China2017-11-05T00:00:00+00:002017-11-05T00:00:00+00:00https://www.christopherhwood.com/Why-China<p>I originally came to China as a way to continue my Chinese language study and pursue a career in consulting, government work, or perhaps an NGO. After getting involved in a startup competition at the Hopkins-Nanjing Center and interning at a startup involved in the ChinaAccelerator program in Shanghai, I changed directions and started working on building a skill set that would let me reach my goals of joining a startup and hopefully one day building my own startup.</p>
<p>In August, 2015 I joined my first startup in Shanghai, 学银钱包 or Stubank. Stubank was one in a new wave of fintech companies working to make digital transactions real. This was around the time of the iPhone 6 release, and the implications of phones that could be secured with finger prints was just being realized by the market. Although Alipay had already been in existence for some time, Wechat Pay was still a relatively new thing. I recall my first few days in the office using cash to pay for my Ele.me fast food deliveries, but after discovering how to bind my bank card to Wechat Pay, online services became so much more convenient. Stubank was out to get a piece of this market, but taking (what I thought at the time to be) a fairly unconventional route of targeting the college student market.</p>
<p>The idea behind Stubank was that Alipay and Wechat Pay were still largely used for paying for online services, and the opportunity existed to bring that convenience of paying via app offline. Of course Stubank wasn’t the only company with this idea, and the race was on to see who could capture the most market share before this window of opportunity closed. Whoever could win would have massive amounts of consumer data, from which many other products could be born.</p>
<p>Targeting college students seems counter-intuitive at first, but on further analysis makes sense. College students have no source of income, and thus spend less money than adults. So why would a startup mobile wallet company go after this no-income market?</p>
<ol>
<li>College students spend in much smaller amounts than adults. $2 here for lunch, $0.50 there for a soft drink, etc. They would be quick to see the advantage of abandoning a pocket full of loose change for a mobile wallet app.</li>
<li>A larger percentage of young people are early-adopters.</li>
<li>Lifetime value of a college student is greater than that of adults.</li>
</ol>
<p>To get money into the hands of our consumers, we had built a micro-loan platform into our application. Students could take out loans up to 10,000 RMB (~ $1,500) after a review from our credit team. The long term goal of the company was to collect data on consumption behaviors and credit worthiness of our customers and use that data to build a credit-rating system that could be used by third parties.</p>
<p>During my interview, I thought the idea sounded very unique and innovative. Only after joining the company did I realize that there were already hundreds of companies in the market. Stubank only employed 20 people (9 developers, 2 product managers, 2 designers, 2 credit analysts, 5 marketers), with marketing making up 25% of the team. The way we marketed our product was by sending our marketing team to local colleges with all kinds of company swag. Every college would have one or two days at the beginning of each semester when they would allow companies to come and pitch products to their students. The pictures that were shared in our company Wechat group showed dozens of similar micro-loan companies almost identical to ours at every college’s event. Since the concept was still fairly new, competition was essentially centered on who had the coolest product or the best rates - depending on how economically savvy the student you asked was.</p>
<p>Growth hacking was crazy. We built ways to import contact lists, share passcodes that would give away cash to new users, and even built a wrapper on top of Taobao’s website to enable quick purchases to Taobao via our app and using our loans as payment. Our marketing team was visiting 2-3 colleges per week in and around Shanghai. Each college we visited would result in 2,000-3,000 new signups in our app. We were pushing app updates once every two weeks, and constantly being told we were moving too slowly.</p>
<p>The Taobao wrapper was hands down the coolest growth hack I’ve ever worked on. After entering the Taobao purchase page, the user would be prompted to login to her Taobao account via the Taobao website login page. The app would intercept the Taobao server’s reply and check to see whether a captcha was needed or if the login was successful. After login in, we would push a new view controller onto the stack, keeping the webpage running in the background. Through a set of javascript function written by me, but downloaded from our server, we would open the users shopping cart. From there, we would use another javascript function to scrape and parse the shopping cart data and feed it into a table view inside of the current view controller. From there, the user could select which items they wished to purchase and enter a checkout screen. Once the user confirmed the payment amount and the terms of the micro loan that they wanted to take out to cover the purchase, she entered into one last webview to re-enter her password and confirm the purchase on Taobao. Outside of logging in, the whole process was carried out via native view controllers, with a webview constantly running in the back, controlled by a series of javascript functions. It was a really clever function that no one else had implemented.</p>
<p>Unfortunately, after only a few months at Stubank the company closed down after our founder sold the company to a competitor citing our slow growth as a reason for not continuing on. I’m forever thankful to him for giving me the opportunity to enter into the local startup ecosystem and for the faith he had in my development ability. The experience I had working at Stubank confirmed my desire to work in startups, and confirmed my opinion that the Chinese market is the future of mobile consumer business.</p>Chris Woodchristopherhwood@gmail.comI originally came to China as a way to continue my Chinese language study and pursue a career in consulting, government work, or perhaps an NGO. After getting involved in a startup competition at the Hopkins-Nanjing Center and interning at a startup involved in the ChinaAccelerator program in Shanghai, I changed directions and started working on building a skill set that would let me reach my goals of joining a startup and hopefully one day building my own startup.20172017-01-03T00:00:00+00:002017-01-03T00:00:00+00:00https://www.christopherhwood.com/2017<p>This is a follow up to the post I wrote last year, <a href="https://christopherhwood.com/2016">2016</a>.</p>
<p>I started 2016 having worked at Niuwa (牛娃互联网金融) for about one and a half months following the merger of Stubank and Niuwa. Because of the depth of talent at Niuwa and since I entered via a merger, I was relatively low on the iOS developer totem pole (we had 5 developers for one app). Although I was able to learn a lot from the company’s use of Agile and their good development practices, I also realized that more likely than not it would take me a great deal of time to work my way up the iOS developer totem pole. In late January I left Niuwa and joined DayDayCook (日日煮) as their first technical hire after the CTO.</p>
<p><img src="https://cdn-images-1.medium.com/max/2560/1*0rgYadIe0rL8F0XrffR1yw.jpeg" alt="" /></p>
<p><a href="https://www.daydaycook.com.cn/daydaycook/website/EN/index.html">DayDayCook</a> is a company that creates recipes, videos, and other content. The company distributes this content through various video platforms both in and outside of China as well as through our award-winning iOS and Android apps. Our CEO, Norma Chu, was born in Hong Kong, raised in the US, and returned to Hong Kong after graduating college. She founded DayDayCook in Hong Kong in 2012, and expanded into mainland China in 2015. Our app was first released on the App Store and various Chinese Android app markets in December, 2015.</p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*G7vZM2a7C8QxQuB5Dd-kVg.jpeg" alt="" /></p>
<p>Prior to my joining DayDayCook, the app development had been outsourced in order to quickly get the app up and running. From January to May I worked alongside our outsource company in developing, testing, and debugging the App. I also worked alongside our angel investors and product managers to design the Apple TV and Apple Watch MVP versions.</p>
<p>In March our company announced a <a href="http://www.scmp.com/tech/start-ups/article/1927081/day-day-cook-how-hong-kong-hobbyist-chef-plans-ride-e-commerce">$5 million dollar A Round of investment</a> led by 500 Startups and MFund and we began hiring more talent for the app and our content development team.</p>
<p>Starting in May, another iOS developer colleague and I began work on a complete redesign of the iOS app in preparation for a global relaunch. We soon added a third iOS developer and worked day and night through May and June to prepare the new iPhone, iPad, Apple Watch, and Apple TV apps for a global launch in three different languages (Simplified Chinese, Traditional Chinese, and English). Once we finally finished development and testing and submitted the app, we were rewarded with a week on the front page of the App Store in mainland China, Hong Kong, and Taiwan. In Hong Kong, our app rose to the #3 spot on the Free Download ranks.</p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*lIUSZTwoSqqHN0hf8ghRRw.jpeg" alt="" /></p>
<p>Following our promotion in July, I was promoted to iOS Team Leader. This was something that was almost completely unfathomable to me - a self-taught developer without significant experience. I’m tremendously grateful for the opportunity and have been working very hard to continue to improve my knowledge of iOS development, people management, and leadership.</p>
<p>The end of the year was quite eventful as well, although certainly less stressful than the mid-year global relaunch period. In September, we released our iMessage app simultaneously with the release of iOS 10 and were again promoted to the front page of the App Store for a brief period. In October we raised a <a href="http://www.marketing-interactive.com/alibaba-entrepreneurs-fund-picks-daydaycook-grana-and-nosh/">$5 million A+ round of investment from Alibaba</a>. And we ended the year with the <a href="https://developer.apple.com/app-store/best-of-2016/developers/">App Store naming us one of the top ten apps in Hong Kong and Macau</a>.</p>
<p><img src="https://cdn-images-1.medium.com/max/5116/1*wKWiyPd0rdS6mNkzp1TAQQ.png" alt="" /></p>
<p>I was incredibly lucky to land at such a wonderful company as DayDayCook, and even luckier to join the company at such a perfect time in its development. I have again been lucky to be around great team members and management, and again have learned so much about not only app development, but also how to create a well-performing team and how to manage personnel and incentives.</p>
<p>Of course, one can never have everything. Because of the incredible success of the DayDayCook apps and company as a whole, I have been super busy with work and side projects have taken a backseat. Many of my 2016 goals went unaccomplished, but I also succeeded in many unanticipated ways. After all is said and done, goals are just a concrete way to express personal growth intentions. I may not have achieved all of my goals, but my personal growth this past year has far exceeded my expectations.</p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*hIOxjYvY_efv-ISZhhzIwg.jpeg" alt="" /></p>
<p>My goals from 2016 were:</p>
<ol>
<li>
<p>Get a stable version of my independently developed app, Earlyworm, on the App Store (failed, I have since given up on the idea of Earlyworm)</p>
</li>
<li>
<p>For at least one month have 1,000 unique users on Earlyworm (failed)</p>
</li>
<li>
<p>Learn Python (and maybe R) and start a data visualization blog (learned Python while studying <a href="https://www.coursera.org/learn/ml-foundations/home/welcome">Machine Learning on Coursera</a>, but no data viz blog)</p>
</li>
<li>
<p>Learn more about China’s fintech and startup ecosystem and publish learnings on my personal website and Medium (although I did learn a significant amount about China’s startup ecosystem, I haven’t yet published any of my learnings)</p>
</li>
<li>
<p>Learn basic web front end and make 2–3 small web apps to share on Product Hunt (failed…completely failed)</p>
</li>
<li>
<p>Continue to seek out innovative individuals in Shanghai to collaborate with on side projects - specifically looking to have one bigger side project that provides low margin services to the local community (failed, joined the Shanghai Global Shapers Hub but did not have enough personal time and did not align with my expectations)</p>
</li>
<li>
<p>Begin work on next big project, targeting either something in financial technology — potentially stock trading, or something related to media and dynamic content forms (failed, pondered a few side project ideas but never found anything intriguing enough to warrant significant time contribution)</p>
</li>
</ol>
<p>My goals for 2017 are as follows:</p>
<ol>
<li>
<p>Improve leadership and management ability. Read one management related book per month and share insights via Medium or personal website.</p>
</li>
<li>
<p>Create an easily scalable side project that can bring in passive income on scale of $5-10K/year.</p>
</li>
<li>
<p>Get better at investing, continue to study Warren Buffett and Charlie Munger and share insights on stock market and investing strategy through personal website or Medium. Measurables: 20% yoy increase in value of stock portfolio.</p>
</li>
<li>
<p>Create one low margin side-project to benefit the community.</p>
</li>
<li>
<p>Learn OpenGL ES and make a photo filter app.</p>
</li>
<li>
<p>Learn Natural Language Processing using Python, Pandas, and Numpy and create a chatbot.</p>
</li>
<li>
<p>Workout, eat healthier, and lower body fat percentage by 7 percentage points.</p>
</li>
</ol>
<p>Every year I try to read a lot of books, last year I read the fifteen listed below:</p>
<ul>
<li>
<p><a href="https://www.amazon.com/gp/product/1591847087/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1591847087&linkCode=as2&tag=christopherhw-20&linkId=2e5ccda4c5f8b145f0dcccf0ab911984">Hatching Twitter</a> - Nick Bilton</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0316219282/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0316219282&linkId=e7fcdb6c630c1cecb0c587e83756480e">The Everything Store, Jeff Bezos and the Age of Amazon</a> - Brad Stone</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0062316095/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0062316095&linkId=3aa00aeb8d9bf110b7517810bb51d755">Sapiens, A Brief History of Humankind</a> - Yuvah Noah Harari</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/1416596585/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=1416596585&linkId=ab958b0e88afce499b87c90f22cb5799">In the Plex</a> - Steven Levy</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0062237500/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0062237500&linkId=fdf89890bab38ee3ea40c8cd4eaf66a2">Epic Measures</a> - Jeremy N. Smith</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/1591845327/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=1591845327&linkId=8e9f7532542ecce56dc522023e94461f">Leaders Eat Last</a> - Simon Sinek</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0062273205/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0062273205&linkId=b63199cb52dc6ae51b63cb04c39681f4">The Hard Thing About Hard Things</a> - Ben Horowitz</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0446310786/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0446310786&linkId=3c41b81ad91958b36621b2126ce34375">To Kill a Mockingbird</a> - Harper Lee</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0143120530/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0143120530&linkId=1d44684a1f4960aa3934279553a11037">Moonwalking with Einstein</a> - Joshua Foer</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0743226720/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0743226720&linkId=d4f59373545462f0f3090dd03e3b0f15">1776</a> - David McCullough</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0679729453/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0679729453&linkId=95ccc84fd59b6206ffba2cb9aab96151">The Years of Lyndon Johnson, The Path To Power</a> - Robert A. Caro</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0385349947/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0385349947&linkId=a259c7c16068cc893b88abbf60f75908">Lean In</a> - Sheryl Sandberg</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/080701429X/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=080701429X&linkId=ba9cb2f98599131509574565107d77f2">Man’s Search for Meaning</a> - Viktor E. Frankl</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/1250037751/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=1250037751&linkId=95cd0ac3fb069274e3ab87c996381cad">Mr. Penumbra’s 24 Hour Bookstore</a> - Robin Sloan</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0393007707/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0393007707&linkId=d18d123d7d8c7ac4ad5fc7fa9dc9845a">Group Psychology and the Analysis of the Ego</a> - Sigmund Freud</p>
</li>
</ul>
<p>I’m currently reading a few books, with a goal of doubling my total from last year and reaching 30 books for the year this year:</p>
<ul>
<li>
<p><a href="https://www.amazon.com/gp/product/039333810X/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=039333810X&linkId=e76723429ec279c0a8925a8fae030a13">The Elegant Universe</a> - Brian Greene (on and off reading)</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0553384619/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0553384619&linkId=43a38101229c33c65ed48e5f9d3cabfa">The Snowball, Warren Buffett and the Business of Life</a> - Alice Schroeder (dedicated reading)</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/1578645018/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=1578645018&linkId=611d4c04b42939cf39b9a8635cf55d0d">Poor Charlie’s Almanack</a> - Peter D. Kaufman (dedicated reading)</p>
</li>
<li>
<p><a href="http://www.berkshirehathaway.com/letters/letters.html">All of Warren Buffett’s Letters to Shareholders</a> (on and off reading)</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/0387402721/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=0387402721&linkId=9c569f048b6eafb7acc04b73b4d9b8ac">All of Statistics</a> - Larry Wasserman (study)</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/067145711X/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=067145711X&linkId=872a0cea17c2dbd54aff64d6d5fdb202">The Great Bridge</a> - David McCullough (Audible)</p>
</li>
<li>
<p><a href="https://www.amazon.com/gp/product/B009PTQ8MQ/ref=as_li_tl?ie=UTF8&tag=christopherhw-20&camp=1789&creative=9325&linkCode=as2&creativeASIN=B009PTQ8MQ&linkId=2aaa9f1c114f132c174a7b84dd072db7">檀香刑</a> - 莫言 (on and off reading)</p>
</li>
</ul>
<p>Here’s to a great 2017 - Chris</p>
<p><img src="https://cdn-images-1.medium.com/max/2560/1*Fe2wT7PlzfQTjfMK4n0Sug.jpeg" alt="" /></p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comThis is a follow up to the post I wrote last year, 2016.20162015-12-30T00:00:00+00:002015-12-30T00:00:00+00:00https://www.christopherhwood.com/2016<h2 id="goodbye-2015">Goodbye 2015</h2>
<p>Going into the beginning of 2015 I had set out a few vague goals for the year. These mainly revolved around me finding my way into a technology company, preferably a startup. At the time, I had already spent about four months slowly teaching myself iOS programming with Swift.* I had a plan to continue programming and eventually making an app, and I thought this could help me to transition into a job at a technology company.</p>
<p>My goals were more or less:</p>
<ol>
<li>
<p>Get better at coding</p>
</li>
<li>
<p>Make an app</p>
</li>
<li>
<p>Get a job doing coding, marketing, or statistical analysis in a tech startup</p>
</li>
</ol>
<p>I continued learning Swift and in February had an idea for an app that I thought could provide me challenges beginning with user interface and continuing all the way to big data analysis and natural language processing — two long term goals of mine. I started work on the app, and quickly realized I had no idea what I was doing and this is not something that would be easy.</p>
<p>After about two and a half months of making a lot of mistakes, I decided to completely scrap my heap of spaghetti code (if you can even call it that) and start all over from the top. My job at the time was with American Councils for International Education, and I was stationed in Shanghai but due to go home at the beginning of June. I somehow got the idea that instead of just aiming for a job at a tech startup, I should shoot for a job at a Chinese tech startup to make use of my six years of Chinese study. In order to prove my coding ability, I would need to have at least a demo of my app available for the interviews. I began devoting almost all of my spare time to developing a minimum viable version of my app.</p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*XKKQCWpDQoq7iWNPjzDgVg.jpeg" alt="Stubank had the best mascot" /><em>Stubank had the best mascot</em></p>
<p>Around May 15, I began applying to as many jobs as I could through China’s <a href="http://51job.com">51job.com</a> and an amazing startup job website, <a href="http://lagou.com">lagou.com</a>. I had interviews with five or six companies ranging from Alibaba’s international marketing division to <a href="http://www.moyuntv.com">a tech startup in Chengdu doing an interesting T2O (Television to Offline) marketing app</a>. I was lucky enough to interview with a Shanghai financial tech company named <a href="http://stubank.com">Stubank</a> (学银钱包) two days before I left for the US. One week after returning home, I received a phone call from the CEO of Stubank offering me a full time position as their second iOS developer and 23rd employee.</p>
<p><img src="https://cdn-images-1.medium.com/max/6528/1*-Y8s4Gs1GeLhHWUDSHkmXA.jpeg" alt="Stubank coworkers and I gearing up for a water balloon battle" /><em>Stubank coworkers and I gearing up for a water balloon battle</em></p>
<p><img src="https://cdn-images-1.medium.com/max/6528/1*LBwzhP7RS4s5zYEXLXshQQ.jpeg" alt="A late night debate between our head of QA (holding marker) and CTO (right) about the fastest way to multiply large numbers mentally" /><em>A late night debate between our head of QA (holding marker) and CTO (right) about the fastest way to multiply large numbers mentally</em></p>
<p>I had managed to create a semi-stable first version of my own app, <a href="http://earlyworm.co">Earlyworm</a>**, that I demoed during my interview. Since the app was working moderately okay, I decided to test it out on the App Store. I very quickly realized the first version was not as stable as I had hoped and had a lot of major bugs. I began spending my mornings, nights, and weekends working to improve the stability of the app and move it closer to my vision of how it should work. After about a month and a half of this, I resubmitted to the App Store and had my app rejected for stability issues on iPad — something I had forgotten to check. At this point, I was not happy with the functionality of my app, and it had a serious bug that was preventing it from being publishable, so I again decided to scrap the code and start from the top.</p>
<p>This time, I decided I would use my own backend instead of <a href="http://parse.com">Parse</a>’s backend as a service. The only problem was I didn’t know how to build a backend. So in July I started teaching myself Node.js during my spare time. Stubank was on the up and up, we were hiring new backend developers and releasing a new update every two to three weeks. I had a hand in nearly all aspects of the app’s iOS version and was getting amazing experience and facing lots of challenges.</p>
<p>Four months later our CEO called a meeting to inform all thirty of us that we were moving too slowly and the competition was getting more and more fierce. This was already pretty well known — we were working on making a micro loan application for Chinese college students and the space was heating up. <a href="http://www.wsj.com/articles/alibaba-unit-backs-student-microloan-startup-1439285775">Alibaba</a> and <a href="http://www.theverge.com/2015/1/5/7493265/webank-tencent-online-private-bank-opens-china">JingDong</a> had invested in competitors in August and <a href="http://www.jinr.com/zixun/article/474">other</a> <a href="http://tech.hexun.com/2015-01-07/172173027.html">competitors</a> were also beginning to pull in A & B rounds of <a href="http://www.rongzi.com/gl/detail/5724.html">funding</a>. Little did we know however that this was our last meeting as a company. We soon realized we were being merged with a larger startup called <a href="http://niuwap2p.com">Niuwa (牛娃互联网金融)</a> and we were all due to interview starting with the technology department that afternoon.</p>
<p>That afternoon I printed out my resume, hopped in a cab with some of my coworkers and headed to the downtown offices of Niuwa. We arrived at the super awesome Xintiandi offices around 3 PM and headed to the top floor for an interview. In my interview, I realized the company uses manual reference counting and Objective C, with neither of which I was very experienced.*** I did not expect to be invited to show up for work on Monday, but later that night I received a call with that very invite. That weekend I studied up on manual reference counting (which is not as complicated as I expected), and on Monday I showed up for the first day of work at Niuwa.</p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*zK1bIatjb4aHY-YHRjhluQ.jpeg" alt="The office kind of resembles something out of the Matrix" /><em>The office kind of resembles something out of the Matrix</em></p>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*EXsbCrjxPW2-ql4A0rmvgQ.jpeg" alt="吴开喜 - My Chinese name, and iOS 开发工程师 - iOS developer engineer" /><em>吴开喜 - My Chinese name, and iOS 开发工程师 - iOS developer engineer</em></p>
<p>Although I loved working at Stubank, I also love Niuwa — both for very different reasons. Stubank’s organizational structure was super flat, and developers had a lot more freedom and input with both how we wrote the code and how the app itself looked and worked. This was a lot of fun, but admittedly I was not experienced enough at that time for it to work well. The app seemingly always had a handful of senseless bugs and the code was quickly growing unwieldy. At Niuwa, the company is structured much more like I would expect a larger company to be, but the management is amazing and I am learning as much about writing good code as I am about how a company should be run.</p>
<p>Earlyworm 2.0 is nearing completion, and is much more in line with what I had originally envisioned. However, I’m currently held up on figuring out how to get my Docker containers up and running on Amazon Web Services and can only hope that I get all issues resolved and will be able to relaunch the app by the end of January.</p>
<p>Which leads me to my goals for 2016.</p>
<ol>
<li>
<p>Get a stable and working version of Earlyworm on the App Store</p>
</li>
<li>
<p>For at least one month have 1,000 active monthly users on Earlyworm</p>
</li>
<li>
<p>Learn Python (and maybe R) and start a data visualization blog</p>
</li>
<li>
<p>Learn more about China’s financial technology sector and startup ecosystem in general and share what I learn on Medium and <a href="http://christopherhwood.com">my website</a></p>
</li>
<li>
<p>Learn basic web front end and make 2–3 small web apps to share on Product Hunt</p>
</li>
<li>
<p>Continue to seek out innovative individuals in Shanghai to collaborate with on side projects (specifically looking to have one bigger side project that provides low margin services to the local community — something a company probably wouldn’t want to tackle)</p>
</li>
<li>
<p>Begin work on next big project, targeting either something in financial technology — potentially stock trading, or something related to media and dynamic content forms.</p>
</li>
</ol>
<p><img src="https://cdn-images-1.medium.com/max/2000/1*zqVTN3XNOUUaFUiD9Kjk5w.jpeg" alt="More Stubank Swag" /><em>More Stubank Swag</em></p>
<ul>
<li>Prior to taking on learning Swift, I had extremely little knowledge of code. I had taken the Javascript course on <a href="http://Codecademy.com">Codecademy</a> and written a few scripts to run in Google Spreadsheets in a previous internship. I learned Swift mainly through the Stanford Course on iTunes U, <a href="http://bitfountain.io">Bitfountain.io</a>, <a href="http://stackoverflow.com">stackoverflow</a>, and blogs like <a href="http://natashatherobot.com">natashatherobot.com</a>.</li>
</ul>
<p>** <a href="http://Earlyworm.co">Earlyworm</a> is a combined Read It Later and Flashcard app for Chinese learners. It uses real life articles from sites like <a href="http://36kr.com">36Kr.com</a> and <a href="http://Caijing.com.cn">Caijing.com.cn</a>, and generates flashcards plus handles your deck automatically using spaced repetition. It’s amazingly simple and requires no work on the reader’s part besides simply reading. It should appear on the App Store in early 2016.</p>
<p>*** When I joined Stubank the source code was entirely written in Objective C, but the company encouraged me to write all new code in Swift. Thus I was able to continue to practice with Swift 2.0 while also learning Objective C while maintaining the code base.</p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comGoodbye 20152 Seconds2015-05-01T00:00:00+00:002015-05-01T00:00:00+00:00https://www.christopherhwood.com/2-Seconds<h3 id="the-difference-in-knowing-and-not-knowing">The difference in knowing and not knowing</h3>
<p><img src="https://cdn-images-1.medium.com/max/5000/1*Puy8rUI5IckIweBMOBni8g.jpeg" alt="" /></p>
<p>Don’t read the below word. I mean look at it, but don’t read it. Don’t think about what it means, don’t try to think of it’s definition. Just look at it as some ink on a page. Maybe it’s artwork.</p>
<blockquote>
<h2 id="amazing">Amazing</h2>
</blockquote>
<p>Oh no! You read it didn’t you. You probably read it before you even got to the break in the page. It’s an appropriate word to pick for this challenge, because the truly <em>amazing</em> thing is that you read and processed that word in less than 100 milliseconds. <em>Less than 1/10 of a second</em>!</p>
<p>Yes. 100 milliseconds to understand a written word is pretty awesome. But it’s also super important. Your brain can only hold new information in its working memory (aka short term memory) for up to <strong>two seconds</strong>. That’s quite the time-window…no pressure, brain!</p>
<p>What does this mean for you the reader? Firstly, if you come across an unfamiliar word you’re not going to be able to identify it in 100 milliseconds, or probably even two seconds.</p>
<p>[Luckily you have some help from the surrounding context, but that’s a topic for another day. ]</p>
<p>If you get tripped up on a word and the context is of no help (think second language learners), then after two seconds the word floats out of your memory and you have to revert your attention back to it. This is why reading in a foreign language is so slow and painful.</p>
<blockquote>
<h2 id="when-a-reader-slowly-analyzes-a-word-into-component-sounds-and-blends-them-a-great-deal-of-capacity-is-consumed-with-relatively-little-left-over-for-comprehension-of-the-word-let-alone-understanding-the-overall-meaning-of-the-sentence-containing-the-word-and-the-paragraph-containing-the-sentence-in-contrast-automatic-word-recognition-ie-recognizing-a-word-as-a-sight-word-consumes-very-little-capacity-and-thus-frees-short-term-capacity-for-the-task-of-comprehending-the-word-and-integrating-the-meaning-of-the-word-with-the-overall-meaning-of-the-sentence-paragraph-and-text">When a reader slowly analyzes a word into component sounds and blends them, a great deal of capacity is consumed, with relatively little left over for comprehension of the word, let alone understanding the overall meaning of the sentence containing the word and the paragraph containing the sentence. In contrast, automatic word recognition (i.e. recognizing a word as a sight word) consumes very little capacity and thus frees short-term capacity for the task of comprehending the word and integrating the meaning of the word with the overall meaning of the sentence, paragraph, and text.</h2>
<h2 id="--p-68"><a href="http://www.amazon.com/Reading-Instruction-That-Works-Third/dp/1593852282/ref=sr_1_1?ie=UTF8&qid=1430490358&sr=8-1&keywords=pressley+m+reading+instruction+that+works">- p. 68</a></h2>
</blockquote>
<p>The worst part is, comprehending a word (or in fancy scientific language: lexical access/lexical processing) is only a “low level” working memory process. Lexical access, combined with syntactic parsing and semantic proposition formation make up the three low level processes. These processes can best be thought of as similar to the fuel and engine of a car. Without them, the rest of the car is not much use.</p>
<p>The rest of the car, the really awesome parts like turbo, air conditioning, and leather seats, are equivalent to our working memory’s “high level” processes. These four processes are listed below:</p>
<p>• Text model of comprehension
• Situation model of reader interpretation
• Background knowledge use and inferencing
• Executive control processes</p>
<p>These are the processes that let you understand what this one word means in the context of its sentence, paragraph, and entire document. It ties in whatever background information you already know, and lets you build off of that. Basically it’s your critical thinking.</p>
<p>Why have you never had to worry about analyzing the way you read? Because fluent first language readers can recognize 98-100% of the words they encounter. Fluent readers not only can recognize a word in under 1/10 of a second, but they don’t even realize they do it and it can’t be stopped.</p>
<p>That’s a powerful advantage for fluent readers. But what about second language learners? For them, reading is not such a piece of cake. Depending on their vocabulary base, they may need to frequently stop reading to look up words, causing them to stop their high level processing and making reading incredibly inefficient.</p>
<p>Are you a second language learner? Then you’re probably wondering how you can combat this. How can you get on something of a more level playing field with those fluent readers? Below I have listed a few activities you can practice.</p>
<blockquote>
<h2 id="flashcards-have-a-terrible-reputationmccullough-1955but-its-undeserved--flashcards-can-foster-automaticity-by-helping-children-read-words-accurately-and-quickly-critics-argue-that-flashcards-only-teach-children-to-bark-at-print-and-do-not-contribute-to-the-bottom-line-of-reading-which-is-comprehension-but-recent-research-suggests-the-opposite-that-teaching-children-to-read-words-faster-can-improve-reading-comprehension-dramatically">Flashcards have a terrible reputation(McCullough, 1955)But it’s undeserved. … Flashcards can foster automaticity by helping children read words accurately and quickly. Critics argue that flashcards only teach children to ‘bark’ at print and do not contribute to the bottom line of reading, which is comprehension. But recent research suggests the opposite, that teaching children to read words faster can improve reading comprehension dramatically.</h2>
<h2 id="--p-37"><a href="http://www.jstor.org/discover/10.2307/20202039?uid=3737408&uid=2&uid=4&sid=21106662102493">- p. 37</a></h2>
</blockquote>
<ol>
<li>
<p>Word and Phrase Recognition Exercises:
• Think flashcards and exercises where you try to pick out the correctly spelled/written word from a cluster of similar, but incorrectly written words</p>
</li>
<li>
<p>Timed Semantic Connection Exercises:
• Exercises that make you examine the relationship between words like flower & rose, or collocations like basket & ball or read & book</p>
</li>
<li>
<p>Lexical Access Fluency Exercises:
• Make a list of words on the left side of the paper and a list of definitions on the right side. Match each word to the definition.</p>
</li>
</ol>
<p>If you’re beginning learning a foreign language, it may be better to simply focus on broadening your vocabulary. Some suggestions are below:</p>
<ol>
<li>
<p>Use a systematic procedure to select words to study. Example: You are reading an article and you want to pick some words to study from it. Each word can be sorted in two ways:
• Helpful to understand the article or not
• Helpful outside of the article being read or not
You should focus on words to which you answer yes to both questions first, then depending on your personal goals decide which of the two requirements is more important and focus on words that only fit that requirement as well.</p>
</li>
<li>
<p>Concept-of-Definition Map. Pick a word, and map out answers to the four questions below:
• What is it?
• Examples?
• Non-examples?
• What is it like?
These maps will help you to store the word in your memory.</p>
</li>
<li>
<p>Try to place yourself in a vocabulary rich environment. Some good ideas I’ve heard include putting sticky notes on all appliances in your house and labeling them in your foreign language. Hang out with friends who speak your foreign language. Watch foreign television shows, listen to foreign music, read a foreign newspaper. Or best yet, travel to the foreign country!</p>
</li>
</ol>
<p>All research for this piece was drawn from <a href="http://www.amazon.com/Teaching-Researching-Reading-Applied-Linguistics-ebook/dp/B00GXBEU8I/ref=tmm_kin_title_0?_encoding=UTF8&sr=&qid=">Teaching and Researching Reading by William Grabe and Fredricka L. Stoller</a>. If you’re interested in learning more about how your brain works while reading I recommend you check out their work.</p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comThe difference in knowing and not knowingNewspapers Vs Free Market2014-09-20T00:00:00+00:002014-09-20T00:00:00+00:00https://www.christopherhwood.com/Newspapers-vs-Free-Market<p>I woke up this morning to a story on Medium from Kristi Culpepper. It is linked here, but in summary it details the growing lack of journalistic quality due to the Internet. While Ms. Culpepper makes a good point that a growing amount of journalism is poorly done and poorly sourced, she fails to mention that the overall amount of journalism and news content being produced has also grown with the advent of the Internet. While today we have Zero Hedge, a Twitter account she picks out for being especially bad, in the days of print we had the National Enquirer and still have the NY Post. It always has and always will be up to the reader’s discretion as to what articles he or she decides to read. This is the beauty of the freedom of the press.</p>
<p>Her article was inspired by a <a href="http://www.brookings.edu/research/essays/2014/bad-news?cid=00900006020080101US0001-2#">much longer piece</a> written by Robert G. Kaiser of the Washington Post. Mr. Kaiser’s work offers a much more perplexing issue: the important role of America news’ old guard as a government watchdog. Today’s flat media structure, in which anyone can publish online and gather an audience, opens us to many previously marginalized viewpoints and educates the populace for next to nothing. While this is a spectacular phenomenon, it does limit profits of large news organizations; profits that Mr. Kaiser argues allow these large outlets to pursue groundbreaking stories.</p>
<p>I have little doubt that today’s emerging media titans – Twitter, Politico, Vice, Vox, BuzzFeed, Reddit, etc. will be able to fill this gap. After all, a single man, not a large news corporation, wrote one of the most influential pieces of journalism in the history of the United States – Common Sense by Thomas Paine. He distributed to 500,000 people when the primary form of transportation was horseback. Furthermore, we see countries like China and Iran that do not have freedom of the press continue to censor their Internet and build firewalls to keep out western media like Twitter, Facebook, and Google. To me it is clear that our society has made a tremendous advancement in the adoption of these new technological media outlets. They have removed power of the press from the hands of a few, and shared it with billions, and in doing so have given potentially given millions of Thomas Paines ink and a quill.</p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comI woke up this morning to a story on Medium from Kristi Culpepper. It is linked here, but in summary it details the growing lack of journalistic quality due to the Internet. While Ms. Culpepper makes a good point that a growing amount of journalism is poorly done and poorly sourced, she fails to mention that the overall amount of journalism and news content being produced has also grown with the advent of the Internet. While today we have Zero Hedge, a Twitter account she picks out for being especially bad, in the days of print we had the National Enquirer and still have the NY Post. It always has and always will be up to the reader’s discretion as to what articles he or she decides to read. This is the beauty of the freedom of the press.Buzzfeed Is Only The Beginning2014-08-19T00:00:00+00:002014-08-19T00:00:00+00:00https://www.christopherhwood.com/BuzzFeed-Is-Only-the-Beginning<h3 id="buzzfeed-just-raised-50-million">BuzzFeed just raised $50 million</h3>
<p>and is valued at $850 million. Andreessen Horowitz, on of the most respected VC firms made the investment and has predicted a changing media landscape that <a href="http://a16z.com/2014/02/25/future-of-news-business/">could grow 10x-100x</a> in the next five years.</p>
<p>News has always been received in a quick, glancing kind of way. Newspapers were published daily or weekly, sold cheaply, and intended to be skimmed and dismissed. Now, these publications have moved to digital, but have not transformed with the new medium. It’s near impossible to read a newspaper article on a cell phone while boarding the subway or struggling for a seat on the bus.</p>
<p><img src="https://cdn-images-1.medium.com/max/6400/1*6szhs2m2WczXCV0CfsVXkg.jpeg" alt="" /></p>
<p>More and more people spend their time watching television shows or playing games while in transport. Reading a digital magazine or newspaper article is too time consuming and too inconvenient to win the battle for attention. Daily televised news broadcasts are too dull and boring to compete with the addictive enjoyment of playing a game or watching a TV show.</p>
<p>In order to survive, news needs to transform into something more interesting. News itself is inherently more interesting than a fictional television show or game, the current problem is it is just not displayed as such. People should be naturally inclined to learn about what is happening in the world around them.</p>
<p>Another problem facing current news forms is the focus on the negative. Anything that drives customers to feel bad will not last long. On the other hand, anything that is too cheery (Upworthy) faces the problem of seeming like a scam or too fake. It is important to focus most on making interesting content, not content that serves any psychological or political purpose.</p>
<p>The news media is in a unique position to harness amazing profit growth over the next decade. With maturing social sharing sites like Facebook, Twitter, and Youtube as well as more up and coming sites like Pinterest, Instagram, and Snapchat, the industry now has a solid and growing foundation on which to build. Just as newspapers required mature printing and distribution capabilities and television news required mature television distributors and a mature television market, the new age media is beginning a period of rapid growth.</p>
<p>Furthermore, with the advent of crypto currency like Bitcoin, profitability is becoming less and less of an issue. With Bitcoin, users can pay fractions of a dollar or even cents per webpage view. Surely a $5/month or $60/year price tag is frightening for users, and the SaaS format makes media companies’ revenue predictions very difficult. However, it is doubtful that the same user would opt to not pay 5 cents to read an article documenting the growth of the US police state through an interactive graph and animated video following the issues in Ferguson, Missouri.</p>
<p>Financially speaking, this ability to charge per page view will create a basic expectation for article revenues and allow a more stable pricing structure to emerge. Investments and loans will be easier to obtain when future income can be predicted more easily. Hiring of editors and contributors will be easier as well.</p>
<p>Secondly, just as magazines and newspapers sported amazing, high quality advertisements in their heyday, I believe we can propel the online advertising market forward into a second advertising golden age by only accepting the most high quality advertisements from the best brands. I see a future where pay per click is disregarded by finer luxury brands and advertisements can once again be judged by their beauty and sophistication rather than their trickery and ability to garner clicks. Indeed, having a distinguished reading audience view a well-created digital ad, whether it is clicked or not, will undoubtedly improve the company’s image.</p>
<p>Lastly, I see a change in how contributors and editors will function. As news media moves from purely written or filmed accounts and into a more interactive and engaging format that can be accessed through mobile devices (namely smart phones, google glass, etc.), I see contributors creating content more and more by means of coding. No longer will journalists be able to only rely upon specialized knowledge and a solid grasp of language. They will now need all of these skills, but also the ability to code and truly “develop” a story. Editors will need to be supreme coders and most likely have a solid sense of marketing technique as well in order to help stories to reach their highest potential of being shared.</p>
<p>We need not fear a future where news media is dead and news media is overtaken by Twitter feeds and random bloggers. Surely these will still linger, and many will offer high levels of utility. However, I see our current stage of news media as a temporary fix for a disrupted industry that has yet to formulate an answer. Just as Marc Andreessen said, look for this to change over the next five years and for phenomenal growth in the industry.</p>
<p>Other relevant articles:</p>
<p>Chris Dixon + Jonah Peretti on BuzzFeed’s Success
<a href="http://cdixon.org/2012/07/24/buzzfeeds-strategy/">http://cdixon.org/2012/07/24/buzzfeeds-strategy/</a></p>
<p>The Argument for BuzzFeed Disrupting the News Media Industry
<a href="http://valleywag.gawker.com/here-s-what-buzzfeed-s-new-investors-really-think-about-1619389022">http://valleywag.gawker.com/here-s-what-buzzfeed-s-new-investors-really-think-about-1619389022</a></p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comBuzzFeed just raised $50 millionUber + China = Fail2014-02-20T00:00:00+00:002014-02-20T00:00:00+00:00https://www.christopherhwood.com/Uber-+-China-=-Fail<h3 id="why-ubers-market-entry-strategy-was-doomed-before-it-began">Why Uber’s market entry strategy was doomed before it began</h3>
<p>If you’ve followed Uber’s rise from Silicon Valley to what is now essentially world-wide, you’ve certainly noticed a few bumps in the road (forgive the pun).</p>
<p>Uber has battled what seems to be the same complaints in France, DC, New York, Boston, Chicago, and even San Francisco.</p>
<p>These complaints are largely based around the unjustness of Uber’s business model. Uber requires a bare minimum of its drivers in the US (background check, insurance, etc.), but even less in China (driver’s license, proof of incorporation, proof of car rental license, proof of public liability insurance). Chinese taxi companies are required to undergo expensive licensing procedures and government limits on the amount of taxi drivers per city. Furthermore, Boston cabbies’ case against Uber exposed that Uber “partners”* park their car outside of the city, where it is significantly cheaper than what is available for the local taxis.</p>
<p>Other complaints stem from consumer rights’ advocates. It has already been seen that China is ready to defend its citizens against unjust business practices, as several Chinese cities have taken stands against the likes of Kuaidi Dache and Didi Dache for using a tipping system to create a competitive atmosphere that incentivizes taxi drivers to bypass those without the money to tip the almost double rates, do not have a smartphone, or who otherwise are simply just not using these taxi booking apps.</p>
<p>While Kuaidi Dache and Didi Dache have seen blowback from the incorporation of their tipping systems, Uber has yet to encounter much resistance.</p>
<blockquote>
<p>In fact, Allen Penn, Uber’s head of Asia reported to the Wall Street Journal that Uber “has had ‘minimal contact’ with the Chinese government, and that the company had a ‘cordial conversation’ with Shanghai authorities when Uber entered the city.”</p>
</blockquote>
<p><strong>Mr. Penn and Uber may not want to get too comfortable.</strong></p>
<p>Although Uber works hard on both its legal page and driver signup page to distinguish itself as a “request tool” and not a “transportation carrier,” according to China’s “Anti-Unfair Competition Law” 《中华人民共和国反不正当竞争法》, Uber may already be in violation of the law. If not, they are awfully close.</p>
<p>China’s “Anti-Unfair Competition Law” states in the eleventh item that it is illegal to make pushing out competition your primary goal. Furthermore, if the offender is found guilty, he will be forced to pay back all of the profits earned during the time period in question and may also face a fine of $10,000 to $30,000.</p>
<p>Certainly, Uber’s current position as a luxury cab does not necessarily interfere with the entire taxi market, although following the lowering of base prices from 60 RMB to 30 RMB, local cabbies may become more worried (base cab rate in Shanghai is currently around 10 RMB).</p>
<p>One aspect that should certainly add to their concern is the fact that Uber drivers are only required to have a bare minimum amount of documentation, while all other taxi drivers have been required to pay expensive licensing fees and be lucky enough to avoid the government cut on the number of local taxis.</p>
<p>Assuming that Uber follows the same path of development in China as it has in all other markets, where it first establishes itself as a luxury brand, then expands to incorporate the entire market by adding local cabs and UberX drivers, I would not place a bet on Uber lasting long in the Chinese market.</p>
<p>Whereas Uber has thus far been able to bully American politicians into accepting its new business model, China’s leaders do not have to listen to its citizens as closely as the elected officials in Washington. Also, China’s courts and officials have always tended to protect homegrown firms rather than foreign firms, and I don’t see local governments throwing Kuaidi Dache and Didi Dache under the bus to protect American Uber.</p>
<p>*Uber refers to their drivers in Shanghai as partners. From examining the application process, it appears that a taxi company owner (with an already established company, hence the proof of incorporation) can apply himself, then set his entire company up with Uber. This should be of slight concern for their Shanghai user base as Uber, and possibly the government, does not know the individual drivers’ backgrounds.</p>
<p>Follow me on Twitter: <a href="https://twitter.com/C_H_Wood">@c_h_wood</a></p>Chris Woodchristopherhwood@gmail.comWhy Uber’s market entry strategy was doomed before it began