slan's blog

Sold NVDA $160 covered call, expiring September 19, 2025, $10 premium.

神经网络的数学本质是概率空间，提示词工程则是帮助LLM在非常多的概率空间中找到较优解，来实现我们的目标，强化学习RL则是在现有的概率空间下，通过Post-training，来优化LLM的输出，提升较优路径的概率。都是优化路径概率，从条条大道中找到通向罗马的路，所以只要提示词构建得当，次级模型是能达到顶级模型的效果的，但这需要精巧的概率设计。

案例1

比如，从繁杂的群聊里面归类话题，总结事件进展，来龙去脉。如果提示词是：

请从群聊中，总结广泛参与讨论的话题，并给出事件进展，来龙去脉。

概率空间会非常宽泛，顶级模型可以很好的划分话题，分析来龙去脉，次级模型可能需要看运气，或者根本没法有效的切分话题。

如果改成：

1. 按时间顺序分析消息流，维护一个话题列表
2. 如果当前消息是对已有话题的讨论，则更新已有话题的讨论内容，将本条消息放到话题列表的消息流中
3. 如果当前消息是新话题的讨论，则创建新话题，将本条消息放到话题列表的消息流中

则概率空间大大收敛，从对全局的信息分析，变成对时间线的回复关系确认，简单很多，小模型也可以很好的完成。

案例2

当生成的结果，包含了幻觉链接，如何解决，比如：分享了⼀张⻋载充电宝的[图⽚](https://example.com/car_power_bank.jpg)，并感慨其重量，LLM生成了一个不存在的地址，这个问题来源于提示词：

每个话题转换为一条清晰、完整的描述，说明信息点、问答核心或分享内容，如有链接嵌入[链接](URL)。

LLM错误的将原文：xx分享了一张动态图片的链接，但是链接并不存在，所以它自己生成了一张图片，并嵌入到回复中。

通常我们如果加限制词，比如检查链接的合法性，或者检查链接是否存在，是较难解决这个问题的，原因是原始Prompt的概率空间非常大，比较模糊，有非常多的错误路径能指向放置一个链接。在此基础上做裁剪是较难达到效果，因为永远有没有考虑到的情况。

如果我们直接进行概率收敛，缩减提示词，则好的多，比如：

每个话题转换为一条清晰、完整的描述，说明信息点、问答核心或分享内容，如原文有http/https的URL，则以markdown格式[链接描述](URL)包含在回复中。

http/https的URL将LLM处理的概率空间缩的很窄，从源头上提高了正确识别URL的概率。比模糊的要求+限定结果好很多。

所以，设计一个提示词，要尽量向具体，众所周知的事情上靠，不要在一个模糊的要求上，加上很多限制词。把自己想象成LLM，按自己写的prompt走一遍，想一想这个步骤我的自由发挥空间是不是太大，如果太大则需要继续精简。

这个视频提到的大脑机制：

1. 大脑接受你的暗示，想法，你说工作太累了，他就整的觉得你累，要休息，你担心生病，它就真的生病。
2. 大脑接受你的想法的方式是头像，和自己的声音
   • 跑步时，想象你瘦的样子，别人在夸你，和想象自己累，乳酸难受，空气不好的画面对动机有很大影响。
   • 想象柠檬，真的会分泌唾液。
3. 大脑避免：痛苦，不熟悉的事情。可以利用这个阻止自己做事。
   • 烟放身边，自然会抽，把烟放车库车里，把自然熟悉的事情，变成周折陌生的事情，动机减弱
   • 想象做PPT的痛苦，学习的痛苦，大脑会接受暗示，学习，做事的动机减弱，最终拖延
   • 想象吃-》变胖-》穿不了衣服，丑的痛苦，阻止自己吃
4. 大脑喜好：快乐，熟悉的事情。可以利用这个让自己做事。
   • 对自己说：我喜欢学习，这是我要做的事情，我就是会做，大脑会相信你，增加学习动机
   • 不断对自己说我可以跑完的，可以的，最终更有可能跑完马拉松
   • 想象跑步后瘦了，穿衣服好看，别人夸你，增加跑步动机

• 一开始给自己打气可能没底但是有了初步动机，持续训练变成熟悉强化动机，有一点点结果之后外界对结果的正面肯定，让你相信你确实可以，最终你就真的可以了。
• 一开始给自己负面画面说就行不行，然后拖延多次结果不行之后动机进一步减弱，最终无结果外界的负面反馈，让你更加相信你真的不行，最终你就真的不行了。

昨天vibe coding翻车了，实现的功能就差一点点，但是这一点点死活调不好，越改越乱，sonnet 4不行换Gemini 2.5 pro救命，有时候能救过来，有时候一起翻车。无奈只能reset hard 然后硬着头皮读代码，精简结构。

Vibe coding的重要关键点：

• 自己掌控核心流程结构，要code review，虽然这越来越难了
• 及时commit，前提自己可以了解掌控这个版本是更新了什么
• 阶段性重构，将代码写少，写精简

technique:

• Assign role to ai, example: Steve Jobs, to improve the design and user experience.
• Evaluate the ability of specific model, if out of its ability, switch to more powerful model.
• For complex tasks, first generate a detailed document(FDD,PRD,etc.), clear the purpose first.
• For complex tasks, debug, log, not only can find the problem, but also can let me understand the architecture and flow.

对付娃，在想发火之前：我尽力了？我想了所有办法？还有其他方案吗？

如果都尽力了，随他去吧，接受现实，就这样了，冲别人发火也没用。

1. Focus, do the thing what I want most. Make it the world best.
2. Data Drive, continuously improve the path of user journey.
3. Be patient, find opportunities rather than chasing hot topics.

• What is the best way to learning english?
• Who can continuously earn users and gain more resources?
  • Nvidia?
  • Amazon?
  • Google?
  • Tesla?
• What is the market price in ? is it overpriced? can I sell call to provide the gamble service?

avoid bad case, target base case, looking for good case.

1. build service
2. investment

find long term unchangeable things:

• energy
• networking
• intelligence
• human nature

1. Take care of family, my love.
2. Exercise, rest and sleep, maintain a good diet to keep good energy
3. Continue to listen and learn about good companies, investments and technology information. Share with others
4. Build apps and services

宏观虽然重要，是一切经济的基础，但是宏观系统太过于复杂，信息难以及时获取处理，在对具体标的操作时，短期不确定高，不适合作为操作依据。

微观的问题，定义能很清楚，解法有很多样，适合在微观里寻找机会与乐趣。

• What companies can grow long-term?
• How to attract and create niche products and services?
  • How to continuously learn English content and use it to grow knowledge?
  • How to provide valuable Technology, Investment Information?
• How to be healthy and keep good energy?
• How to have a good relationship and take care of family?

• 不要迷茫，不要慌张
• 太阳下山，还有月光
• 它会把人生路照亮
• 陪你到想去的地方
• 不要彷徨，不要沮丧
• 月亮睡了，还有朝阳
• 抬头看，天一定会亮
• 爱的人，一定会陪在你身旁

From 《不要迷茫，不要彷徨》

经济危机是好事，它给了人审视自己的机会，此时宜健身，宜寻找尝试新机会，宜去杠杆。当经济危机一过，自己的成长+市场情绪的回来，就会轻舟已过万重山。千万别怪别人，怪市场，怪政策，为了挽回损失加杠杆，既解决不了问题，也掩盖了暴露脆弱点的机会。 ​​​

Today I bought back my short puts on NVDL, the final loss is 69%. It was a big loss, and it's sad, but I finally got rid of it.

Keep moving forward. Do the right thing.

https://wallstreetcn.com/articles/3687254

From 大道:

你的每个决策都会对你的一生起到作用。你要"理性想长远"。

所谓要坚持做正确的事情其实就是：

1. 发现错了就要坚决停止，
2. 明知是错的东西就要坚决不做，
3. 面临决策时首先考虑是非和长远而不是短期的眼前的利益。

From TK:

• 人的精神和肉体是统一的整体，相互影响。
• 你们试试，把两个嘴角上扬，做出微笑的动作，保持几秒钟后，就会感觉心里好像也有什么开心的事。

• Body limit: Should be healthy and strong, keep good energy.
• Knowledge limit: Know How, Know Why, Know What.
• Resource limit: Should have enough resources to exchange.

What I want to break now is:

• Break the energy limit with good diet, exercise, and sleeping well.
• Break the English limit.

I'm very bullish on Nvidia. Every dip or rise higher ignites my FOMO. I've profited a lot from that.

My strategy is to hold the stock and sell puts to earn premium to buy more Nvidia. If the price goes down, I just need to roll to the next week to expect the price to come back.

It has some risk and looks dangerous, but it has been fine in the past, as Nvidia has bounced back many times.

But this time, I learned what risk really is. When I sell puts, I'm exposed to the full loss risk if the stock goes down. The price went lower and lower. I needed to limit the loss by hedging, but I hadn't.

Today Nvidia's price went down to 110, and my position lost a lot. What a huge lesson.

Remember:

1. No leverage
2. Hedge the risk, limit the loss - that is what I need to learn and explore.

Hedging is a good tool. If I hedge well, I won't worry when the market goes down. Losses will be contained to a position and stop, then I'll have more time to wait for a bounce back. It's even possible to profit from hedging.

#hashtag: investment

Tomorrow is the result of today's choices.

Today's improvements will make tomorrow better. Today's setbacks will make tomorrow harder.

When I slice time into daily pieces, I live fully in the present moment. I can't defer tasks to some future time - I can only act now.

While today's happiness means nothing, what I truly seek is a better tomorrow. Therefore, I must take actions today that will help me grow and improve for the future.

The market's enthusiasm has been ignited by the news that the BIG BOSS will meet with tech leaders.

The YINN (3x bullish China ETF) has been on the rise. So I used my reserve money to try to take part in it. When I bought it, I just regretted that I couldn't earn much with this little money.

I felt uncomfortable, so I sold it with little profit.

Short-term trading success can create overconfidence, leading to risky behaviors like increasing leverage. Rather than chasing market predictions, I should invest in my personal development through the 'Ability Cycle' - continuously learning and improving my skills. This strategy provides more sustainable, long-term value for my life.

TELL MYSELF:

• Only earn the money that I can handle.
• Punish myself to read Lilu's speech

TGA is a buffer account it doesn't create liquidity itself. When its inflow and outflow is balanced, it doesn't create more money. But when some reason like debt deficit make TGA outflow bigger than inflow, the outflow will be the deposit in banks. Banks can use this deposit to loan to others, then money created. When the debt deficit resolved, TGA will draw money from banks then lead to less liquidity.

Is sleep.

There are 2 types of information: highly compressed and low compressed. You need low compressed information to build a good LLM because it has more details. People using tools to summarize books, podcasts, and articles is a good way to get quick news or rough ideas, but if they want the information to become their knowledge, the training process (reading, listening, and thinking about the whole piece of information) is inevitable.

Example: In Lex Fridman's newest AI podcast, they talk about many things including Deepseek, AGI, chips, and AI industries - you can roughly get an idea of what the trends are. But if you want to know why Google's Gemini has a longer context length (millions vs thousands) and how this feature will impact AI trends, you can't find it in any summary article. With these many details in your head, this "LLM" becomes more valuable.

Another example is English learning - it's impossible to just memorize vocabulary and learn grammar and expect to reason how to speak and understand English. You need to listen to and read enough materials to store sufficient information in your brain. Every method that doesn't let you experience more, but instead gives you compressed, structured content optimized for special tests, is rubbish.

• if you don't need to earn too much money, you can do what you love.
• if you want to be moderately rich, you need to do what your boss wants.
• if you want to be extremely rich, you need to follow your passion.

When you have trouble choosing between following your passion and making money, it's almost always due to ignorance:

• you don't know what makes you happy
• you don't know what various kinds of work are really like
• you don't know how well you could do them

A good way to solve this is to gain more certainty. You can try working on things you're interested in. That will give you more information.

A trick for judging different kinds of work is to look at your colleagues:

• if you choose a job mainly for how well it pays, you'll be surrounded by other people who chose it for the same reason
• if you choose a job mainly for how much you like it, you'll be surrounded by people who like it too

And choosing options that may give you more options in the future is also a good way to face uncertainty (staying upwind). For example: if you're unsure whether to major in math or economics, choosing math is the upwind of economics - it will be easier to switch later from math to economics than from economics to math.

It's not a good idea to simply take the advice of "follow your passion", because you mostly hear it from famous people (who are working on what they're most interested in). Not everyone does great work.

From When to do what you love

• Writing is hard; to write is to think, you have to think clearly to write well.
• The more prestigious the work is, the more writing is needed.
• The importance of writing and the difficulty of writing create an enormous pressure.
• AI reduces the pressure, but it will result in people with less writing ability.

When we write less, we think less, and that is bad. I choose not to be bad, so I will write more.

From: Writes And Write-Nots

When pursuing our dreams, we need consistent practice to achieve them. The more we practice, the stronger our neural pathways become, eventually making these practices effortless. Here are the key elements to building powerful habits:

1. Simplify life by maintaining consistent routines. This reduces mental energy expenditure and minimizes distractions.
2. Choose activities that you genuinely enjoy doing daily and can repeat frequently. These activities should contribute to building your skills or abilities over time.

My current practices include:

1. Reading: Daily exploration of books that interest me
2. Learning and reflection: Listening to podcasts and writing notes
3. Musical development: Regular piano practice
4. Physical fitness: Combination of running, walking, and strength training

In the formula of information creation (Energy × Computation = Solid Information), energy is the most crucial component. When we feel low on energy, we should first focus on finding ways to improve our energy levels, rather than attempting tasks with depleted energy, which will only produce poor results.

Based on "Why Information Grows," we can express the fundamental relationship as:

Energy × Computation = Solid Information

There are three types of computational systems:

1. Humans,
2. Computers,
3. Transformers(AI models).

Each system can form networks that amplify their individual capabilities:

1. Human networks: teams, armies, companies, industries, and societies
2. Computer networks: the Internet, blockchain systems
3. Transformer networks: emerging AI infrastructures

Computer networks compose human networks and have significantly enhanced human networking capabilities.

I believe we will witness the emergence of transformer networks(AI models + internet + human network), creating a new form of intelligence capable of generating information with unprecedented efficiency. This network maybe we can call it: "SKYNET".

Have a truly free life:

• Earn enough to cover living costs and achieve significant financial growth.
• Take care of people I care about and maintain healthy relationships with them.
• Challenge my physical limitations, control my desires, overcome my weaknesses, and improve my energy and health.
• Build an information system to effectively process, create, and distribute information

阿里看似亡于大环境，外部竞争，内部大公司病。实际亡于分家，分家导致的创新网络，人才流动网络瓦解，使生长在这一网络的创新能力，信息丰富度衰减。阿里云的成立基础是建立在蚂蚁，淘天深厚的底层网络积累上，大量的人才无缝拔插到阿里云，如果放在分家后，即使有阿里云的idea，以及高层加倍的投入决心，失去这一底层网络的支持，也只会是另外一个滴滴云，或者字节云，事倍功半。

所谓智慧不过是在不均衡系统里能量不断输入产生的自发性信息创造造成熵减的过程，作为系统的一份子要时刻感恩能量的输入，不要过分高估自己的努力，更多的时候应该去寻找能量密集的地方，持续挣钱的公司，不缺爱的人，富足的社会环境。千万别去快发不出工资举债度日的公司，千万别找从小缺爱的人，千万别去发展程度比较低的地方，你没那么大能耐，别想着多打几份工，多努力学习就能改变什么，就算是努力学习，也是为了去寻找，安家在能量更高的地方。

I need to create information!!!

从信息的角度，不均衡系统熵减，信息不断被创造，而计算机是这个过程的主要执行者，树是一个计算机，人也是一个计算机，社会也是一个计算机，有史以来最强大的计算机是人，人不断以指数级创造了这个社会的信息，现在的AI还是辅助人，增强人作为人去创造信息的工具。但是我们得承认，在不断能量输入的系统里，人肯定不是那个最强的计算机，现在的AI也不是，未来的AI可能是，也可能不是，一定会有一个更强大的计算机，更强的的去消化能量，创造信息，创造秩序。

In Judo, the key to success is to use the opponent's strength against them. Not use your strength against them. The bigger the ennemy are, the more potential the success is.

chatgpt是怎么出来的： 当强化学习，深度学习遇到瓶颈时，大家隐约发现了scaling law，更大的数据意味着更强的智能，在图像识别，人脸识别等领域都是如此，目前数据最多的人类知识库就是自然语言，于是从自然语言入手，gpt-1，gpt-2 gpt-3，不断扩充数据，模型规模，期望达到能在自然语言上对话的能力。最后在gpt-3.5成功了，自然语言chat是ai最新最振奋人心的突破方向。

沿着这个scaling law，特斯拉迅速转型，把fsd从人工代码建模，什么空间算法，运动算法，碰撞算法全不用了，完全转向ai，用数据去训练，得到了巨大的提升，使自动驾驶看到了希望

机器人的突破，比自然语言的突破更难的原因是数据，自动驾驶还有特斯拉的大量数据，可以直接拿来训练，市面上的机器人基本还处在特斯拉早期的人工写代码阶段，希望在特斯拉，英伟达已经开始了用ai来驱动机器人的探索，数据没有就人工制造，他们招人戴上vision pro，遥控机器人进行任务，这些遥控数据= 自动驾驶领域人类司机的驾驶数据，能让机器人学会走路，搬东西，扛炸药包往前冲等各种指令，只是受限于数据规模，会有较长时间的启动期才能走到chatgpt moment。

Personal coupon growth formula:

Personal Coupon Growth = Body × Knowledge: (Language & Logic) × Information × Action

• Information and Action make me go faster.
• Body and Knowledge enable me to go further.

move less, train more.

Whenever I reach my goals or await the outcome, a profound emptiness settles in, leaving me uncertain about my next steps.

The wait is unbearable - isolated and unsupported, I find myself irresistibly drawn to distractions, constantly refreshing for updates to kill time, and craving conversation. Yet, waiting is an indispensable life lesson. I must harness this time for introspection, learning, and self-improvement.

For if I succumb to the allure of fleeting gratifications, I risk forfeiting the potential for true liberation.

When emptiness:

• ❌ Surfing and refreshing for updates, seeking talks.
• ✅ Reading, Listing podcast, Reflective Writing.
• ✅ Exercise, Meditation, Walking.
• ✅ Planing.

Selling puts is a good strategy to earn money when the price is steady or slightly up. The pain comes from:

• When the price goes up quickly, you see that the profit is much smaller than if you had bought the stock.
• When the price goes down, you lose money equivalent to the price drop.

If you can bear the pain of not getting as much profit as buying stock, that small and steady profit will start to accumulate, and you will see the power of compounding.

I just changed the blog template to make it simpler. I also added a cat to my blog, so I'm not as lonely now 😊.

from:

to:

What I need to do in the short term to recover from training:

• Training Body: Recover from running, keep weight under 70kg.
• Training Mind: Build a system to listen, read, and write. Try to build models.

• Biden: Kill SEC's Crypto Accounting Policy, Ethereum ETFs, hiring a meme manager.
• Biden: Open war on hidden "junk fees" like airline and bank fees.
• Biden: Blind on Iran's sanctions, Iranian oil reaching Europe freely.

Biden can't tackle the INFLATION, so the path of least resistance is obvious: Do whaterver he can to get the votes, even though it's insignificant to solve the problem.

This is my learn note on the podcast. Follow me to get more. reference:

• Biden’s Open War on Hidden Fees
• MacroVoices #429 Dr. Anas Alhajji: Oil Market Outlook, Iran & Energy Geopolitics

#hashtag: LearnThroughPodcast ValidEnglish

tk发了一条微博：“美国从红州限堕胎到到蓝州零元购，中国从《妇女儿童权益保护法》到医药集采，背后都是同一个原理。我在私下给一些朋友详细讲过这个政府第一性原理。可惜不能在这里说。但看了上面这四个仿佛完全扯不到一起去的例子，也许有人能自己想明白。想明白了这个原理，不仅能理解很多事情，也能预测很多事情。 ”

限制坠胎是Roe坠胎案被推翻，导致在联邦层面失去坠胎合法权 0元购是2014出的法案，盗窃950元以下商品等罪行从重罪降为轻罪，原因是加州等地方监狱人满为患

我的猜测是：这几个case都涉及到利益的再平衡，平衡就涉及到+和-，决策符合大多数人要求，但特别是0元购和坠胎案都广受诟病，坠胎案甚至一度让Republican掉了很多选票。是不是说涉及到复杂关系的时候，决策方永远选择短期阻力最小的路径去演化

The 401k is a dominant retirement plan in the U.S. This year, 4.1 million Americans turn 65 years old, but 49% of people aged 55-65 have nothing in their 401k account, and 10%-20% of retired people live in poverty. Why does the 401k not work as expected? And why is this a problem that will damage belief in the capitalist system?

Before the 401k, in the 50s to 70s, the U.S. had a pension system in the private sector. People would work in a company for 30 or more years, and the company would guarantee a fixed retirement income for the rest of their life. It worked well against the backdrop of lifetime employment and people not living as long as they do now.

The 401k came about by accident. It began when companies wanted to reward their executives, so they created a tax-deferred account to pay bonuses to their executives. The managers could get the money when they retired or left the company. There was a risk whether the IRS would tax the money, but in 1978, Congress passed a law to make it legal. That's how Section 401(k) came about. A retirement specialist, Ted Benna, found this law and used it to create a retirement plan for his company. He provided a matching contribution to his employees. If employees put money in the 401k account, the company would match the same amount of money. It was a win-win situation. The company got rid of the responsibility of the guaranteed pension system, and employees got extra money, tax benefits, and investment growth.

Why did the 401k become dominant in the U.S.? 1) The auto and steel industries declined, and a lot of corporations regarded pensions as a problem. 2) Reaganomics advocated individual responsibility, and the 401k shifted the responsibility from the employer to the employees.

Who can benefit from a 401k? Those who are high-income, disciplined, and have good investment knowledge. They don't need retirement money to live. But who falls out of the system? Those who are low-income, whose company didn't provide a 401k, or they didn't know how to invest, or they had to withdraw the money for an emergency.

The 401k reinforced income inequality. The rich get richer, of course, that tool was first designed for rich people. And normal people benefited from it due to government legislation. That is a common American model, normal people benefit from the elite's move. But the inequality still exists, millions of people falling through the cracks of the retirement system make the country driven toward socialism.

How to solve the problem? There is a plan called the Thrift Savings Plan. It's a retirement plan for federal employees, and Congress is considering expanding it to all Americans. When a company can't provide the matching contribution, the government will provide it.

Will it reduce income inequality? We still don't know, but one thing is clear, in the U.S., a comfortable retired life relies on the individual to make it happen. Retirement is a privilege, not a right.

This article is my learn note on the podcast. Follow me to get more. https://www.nytimes.com/2024/05/20/podcasts/the-daily/401k-retirement.html

#hashtag: LearnThroughPodcast ValidEnglish

I used to believe that if I keep learning, keep trying, I can live however I like, can be a brand new person. But now I have to admit that is wrong, I'm a person, my parents' son, Chinese, introverted. It's my nature.

I can't change myself, but I can make myself stronger, more resilient, more capable.

Democracies: The restraint of leaders in their rhetoric during heated elections is a key indicator. Autocracies: The ability to self-correct when wrong is crucial.

Scarce resources for person:

• Time
• Logic
• Information

And inflation is also drive by Svarcity, if resources are scarce, the price will rise. = money/product, In the long term, product increases drive the inflation down. so the monetary policy is the key dimension to control the inflation in the short term.

#hashtag: investment

I'm shorting Bonds, waiting is so hard, especially face the loss. be patient, be rational!

When the U.S. manufacturing industry moved to Japan, Korea, China, lots of other jobs were created for people.

How will normal people live after AI replaces most of the jobs? That is an interesting question to investigate.

• 1. How has the manufacturing industry changed in the past?
• 2. Look at the current trend of AI replacing jobs

How will the U.S. Navy remove the collapsed bridge from the water? the bridge is so heavy that it can't be lifted by a crane. bomb it? that's a interesting engineering problem. the news is show they are planning to cut it into pieces and use multiple cranes to lift it. I will keep an eye on how they do it.

https://www.wbaltv.com/article/baltimore-key-bridge-collapse-update-death-toll/60341894

https://mp.weixin.qq.com/s/AA9mNVX6YEe5j_aKCT3pmQ

这本书主要说的是全球的人口老龄化导致一系列的危机，可能超出人的预期，这也是我们现在预期中国未来30年陷入日本化的主要论据

截至目前，它完全正确，但是它忽略了一个变量：ai robot， 人口老龄化的速度远远慢于ai在生产制造领域替代的速度。

对未来有没有信心，可以以现在为坐标，观察：

• 农业机械化智能化的升级速度，生产装备降价速度是否大于 劳动人口的下降速度。
• 工业生产领域的巨头增长速度，反应出来的，汽车，电子消费品，生活用品的价格下降速度
• 能源生产成本，光伏每度电的发电成本

据我观察，这些指标都在以肉眼可见的速度下降，很多gpt moment会不断涌现，短期我在实际押注的是特斯拉的 FSD GPT MOMENT, elon 说不是今年就是明年，我相信它.

通胀，通屁涨，准备迎接由生产力提升带来的优质通缩吧，美好的日子要来了.

#hashtag: investment

FOMC Metting and Powell's dovish speech drive the stock market up. I guess it's low risk to hold stocks until 03/29 PCE data out.

#hashtag: investment

NAR's settlement allowed NAR to pay $418 million to recent buyers and modify the rules to potentially reduce commission fees. This could lower housing costs and ease some inflation pressure.

#hashtag: investment

TK教主在微博，如果遇到一个看起来很傻又不想反驳的人，就会盲转发：“盲猜关注列表的时候又到了”，我每次去点开看那个逻辑有些奇怪的人，其关注列表都会是：军事，搞笑，娱乐类的博主。

是长期关注这些导致了人的思维和逻辑封闭，还是思维和逻辑封闭的人才会长期关注这些？估计没一个准确答案，唯一能确信的是：信息的接收渠道，信息的选择，决定着一个人。

Be patient, don't do much, follow the rational plan. grow up steadily.

澎湃新闻：“近日，MIT团队将聚变反应堆的每瓦特成本几乎降低到了1/40，让核聚变技术在商用成为了可能。”

这个新闻说的是一个MIT团队在磁约束小型化上取得的验证成果，没有特别新的突破，

MIT在搞一个类似ITER的项目SPARC，磁约束上的区别是SPARC是高温超导，ITER是低温超导， SPARC小四倍

SPARC目标是商业化，这和ITER的基础研究目标不一样，目前看SPARC更有可能实现初步的核聚变。

Never give up，未来比想象更美好，更现实

ECB's interest rate cut will boost U.S. bonds and stocks.

However, U.S. bond yields may be massively changed by the CPI data on 03/12. be cautious.

Yesterday's PCE data showed a MoM of 0.4%, which was within expectations. The market has already priced in the hot CPI and PPI data, causing a decrease in bond yield. However, this doesn't mean inflation is over. A 0.4% increase over 12 months equates to 4.8% inflation. Will it cool down? We'll need to wait for the CPI data on 03/12.

invest strategy:

• hold stock until 03/12.

#hashtag: invesment

Btc, an attention-driven business, will peak in an economic ease cycle before declining. Eventually, it'll reach an attention limit, lose public interest, and fare worse than gold.

#hashtag: invesment

Despite high inflation expectations, Treasury bond yields are declining, possibly due to risk aversion-driven demand before PCE release. This is evident as the stock market drops and short-term yields rise back today, while long-term yields still fall.

#hashtag: invesment

TLT price ~= Treasury yield change

example: US 20-year treasury yield change from 4.4% to 4.3%,

• treasury yield change : 4.4 - 4.3 = 0.1, change =0.1/4.4 = 2.27%
• TLT price change ~= 2.27%

#hashtag: invesment

Let warring ways be banished from the world.

Let justice everywhere its carpet throw.

May Friendship reconcile ancient hatreds.

May Love grow from the seed of love we sow.

Forgive me, God, or whatever you are. Whether you forgive me or not, I don't care.

What I should do:

• Train, do my best to improve myself, no matter how hard it is, I can do it!
• Rest, this is part of training, do it!
• Create, build the list, finish the list!

I'm somewhat down today. A high amount of listening has made me not want to do anything. I'm tired!

I need some change.

Can't wait to announce that I'm a fans of the animated sitcom Family Guy.

https://www.youtube.com/channel/UC7fZ-QNFi7amcOncIl6owxA

What happens to guns that nobody wants any more?

In U.S. police departments, many firearms are kept from confiscations, buyback programs, and other sources. There are usually 2 ways to dispose of them:

• melt it down
• destroy it by blowtorch or sledgehammer

But what people commonly don't know is that a company called Gun Busters provides a free service. They will destroy the gun by just destroying one piece like the receiver or the frame (which is allowed by law). However, the problem is that the gun isn't fully destroyed. They also sell the materials to others, and these materials can be used to make new guns.

In just one second, a little boy lost his mother in Gaza. There are thousands of unseen tragedies in the war.

前几天我买了个森海赛尔的耳机，包头的，有线，原因是我的旧的森海塞尔有线接触不良，经常需要转一转插头到一个特定位置才能听到双声道。买包头的因为想隔绝环境音+不要漏音，买有线是觉得简单，不用充电，森海的音质也好。淘宝上一般三百多，有一家店只要159。

我觉得可能是有线的买的少了，有些要清库存就低价，于是我选择了159的，等了2天，满怀激动的心情我拆开了它，刺鼻的味道，不灵活的伸缩告诉我这可能是假货，于是我准备退货，但是我知道卖假货要退一赔三，我去找店家申请。他当然是不同意的，于是我曉之以情动之以礼，顺便淘宝投诉假货，他妥协了，退款，赔我一百五，加耳机送我。

于是我拿着赔偿款可以够一够next level的耳机了，看评测，发现舒适度排行第一的耳机Sony WH CH720n， 索尼的无线耳机xm4我买过，重，不舒适，降噪虽强但是极度不适应，所以出了，这个720号称是索尼最轻的包头耳机，于是我决定给他一个机会。

已经等了两三天了，我不想再等了，京东769次日达，淘宝海外自营679，2天后达 当然是选便宜的，本来下单了，看到了同城自提只有10km，又心痒了，没骑一会，大货车呼啸带过来的灰尘让我想起一句话：“不要花时间在低于自己时间价值的事情上”，一个闪送三十块，干嘛要自己去取。来都来了只能硬着头皮去继续骑，只能安慰自己我是在体验贫穷，好让我知道没钱是多么艰难。出了绕城，来到一片工地包围的写字楼，终于拿到了货。命运的齿轮开始转动，心情从最低谷开始上升。

轻，舒适，一路回家听完2集podcast，戴着骑车也不冷，这让我很是开心，支持同时连2个设备，好好好，折腾一番，戴着跑步，稳定，还是轻，比xm4好多了，第一次体验这样安静的环境，激情的音乐，纯粹的跑步。

事件就是这样的事件，我觉得很有意思，第一次打假维权，买到超过期望的耳机，不用担心图书馆空调噪声等了。但是我喜欢强行总结点感悟，那就感悟点吧，不要花时间在低于自己时间价值的事情上面，但是花了好像也没那么糟。还是多折腾，多改变重要，总是会有惊喜。

Just came back from the New Year's Countdown Activity. 2023 is gone, the past year is full of loving memories. Thanks to all the people who have come into my life, with your love, I have the courage to face the unknown future. I love you all.

In 2024, I will do my best to love, love my family, love myself, love my friends, love the world.

A human's structure:

• Level 3: Service // What he provides
  • System
  • Product
• Level 2: Action // What he can do
  • Storage: Knowledge
  • Processor: Thinking ability
• Level 1: Body // What he has
  • Health
  • Age
  • Strength
  • Energy

Recently, I've been unable to focus. My brain is full of noise, busy surfing, catching new things, but I can't foucus on the job that I should be doing.

Yesterday, I tried not to use a PC, or pencil and paper to make a plan of how to write codes, just in my brain. At first, it was so hard. I couldn't even remember the idea that I just thought of, many other things distracted me. The outside world is so attractive, people are moving, talking, the metro is arriving at stations, cars are passing by. How is Blue's competition going? How will the stock market go? Will PCE go up again?

I tried to create some containers in my brain to hold my thoughts. They're rectangular box, I put the plan from top to bottom, first, second, third. The box is emptied again and again because I can't remember what I just planned to do, but when I tried sometimes, those boxes stick in my brain, what I need to do is so clear.

Now I know the reason why I can't focus, why the outside world is so easy to distract me. I lack inner thinking. Nowadays, I can easily write notes and sync them across all my devices, but my brain is exercised less and less, then I will lose the ability to store, process, and recall information in my brain.

#time: 2023-12-26

Yes, you read that correctly. Today is Tuesday, but I'm writing last week's report.

Last week was full of mess:

1. [ ] Service
   1. [ ] Wrote RSS subscribe feature, but failed to finish it. Just couldn't focus on it, wrote for a while, and got distracted by other things.
   2. [x] Started weekly communication with users.
2. [x] Finished reading the book "Revolution of Glucose".
3. [ ] Didn't finish the listening goals.
4. [ ] Slept better than last week, but still not good.

The main reason for this mess is that I didn't have clear priorities, just did what I remembered. When things started going bad, I didn't have a buffer to handle it. I have reorganized my priorities:

• 1. Push service to public. Build a service that can earn 6k$/month
• 2. English improvement
• 3. Health, exercise, eat healthy food, sleep early, take care of my family.

When time switches to these priority zones, I should focus on them, not keep other things in mind.

#hashtag: WeeklyReport

Mouse is a mental status invented by me to describe a status that I need to do a hard thing, but I'm not want to do it, i'm avoid to face it, surf the internet, do other things, but eventually I also need to do it.

When I'm in this status, I need to :

• listen a music that I like
• write down any thing I can to to make the goal more clear. or more near.
• just do the simple thing.

this week:

1. Kickd out by Xuan's English learning program. cause I didn't commit enough time to learn english.
2. develop Valid English App to v29, improve highlights, and fix bugs, and add a feature of dictation.
3. [x] Finish half marthon in 2:04 at Hangzhou.
4. [ ] not finish book reading goal. 50%
5. [ ] carefully listen podcast 2/3
6. [ ] casually listen podcast group 1/4
7. [ ] sleep goal runs bad.

Next week:

• perform better in Daily Maintenance program, schedule at morning, review at night, write down logs.
• build a App development pipeline. publish every 2 weeks.
• write english leanring guide experience. find resouces.

#hashtag: WeeklyReport

Yesterday, I was anxious about what I was doing. When I remember the past year of my study and work, it was just so-so. Sometimes I could find something to pursue, but most of the time I was driven by events. I didn't do something that I was really satisfied with. I have wanted to learn English for almost 10 years, and have been running intermittently for 7-8 years. I didn't do it really well, things just happened to me, and that's it. No more, no surprises. This year my running and English skills are better, not because I have changed my performance or mindset, but simply because I have had more time to do it.

When I started to learn English with Xuan, I found that my problem is my attitude towards life. I'm a "just ok" man, and usually fool myself into thinking that I'm doing well. My English learning process is the best example. I know English is my most important skill, but I didn't really take it as the most important thing to do. Sometimes I watched movies, sometimes I listened to podcasts, and sometimes there were many other things that interrupted my learning progress.

Upon deep thinking, I know that what I lack is the courage and ability to face difficulties.

• Learning IPA is boring, so I just skipped it. // Hence, my pronunciation is bad.
• Marking is hard, so I'll just do it in the future. // Hence, I didn't get much feedback and help.
• Reading the book is hard, can I just listen and watch for fun? // Hence, I didn't solidify my English skills.
• Algorithms are hard, Machine Learning is hard, it's not about my job, maybe I'll learn it in the future.
• Weekly reports are boring, summaries are boring, socializing is boring.

I know I don't need to do all of these things good, and I can't do all of them well. But what about my dream? How can I achieve my dream? If I'm sloppy in all things, I will never get it. It's time to admit it.

This week

• I published my English APP, and received much feedbacks.
• Mainly learned English with Xuan, practiced pronunciation, and read the book.

Next week

• English improvement, practice pronunciation use RAZ, read books and listen to podcasts 3hours+ per day.
• APP improvement, fix bugs, improve highlight.
• Build service for beginner, find listen materials, and write guide for beginner.

#hashtag: WeeklyReport

Recieved my new MacBook Pro today, It's faster, more battery life, but why I'm not excited as before?

with hundreds of practice, and the help of https://www.youtube.com/watch?v=QAHg7h80-KQ, I finally got the -lee sound.

the key is to make mouth more wide.

Today, I spent some time trying to follow the RAZ video to shadow read the sentence, but the first one sucked. The sentence is "Insects have many kinds of wings."

Insect, wings, kinds, are the words that I mispronounced. I corrected the “t” in Insect, and “-ing” in wings, but the biggest problem is "kind". Most of the time, I read it as "canned" or "cans" or "candle". I spent whole day try to correct it, but still can't get it right. all the words with -lee sounds i can't pronounce it right. like:

• simply
• family
• Actually

I always pronounce it as -nee, I had tried compare the wave form, shadowing, but it's not work, even my wife and daughter that can pronounce it right, but failed to correct me. Haha. there must be something is wrong, I will find it out.

Today, I launched my English app and received much feedback. I'm very glad it works without any major problems. I have many things to study and many challenges to face, but I will make it come true.

I have been learning English with Xuan (X^2) for 3 days. Every time, she gives me refreshing and new shocks. She is an incredibly interesting person with very high standards in everything. Studying with her is full of challenges, but the feeling of excitement and hope is so unforgettable.

With her help, I found that what I need to do most importantly is to get much more comprehensible input. It's not just about listening more, but also about building my high-quality transformer system.

1. I need to delve deeper into the scene, analyze the words and sentences, and try to tell the story in my own words.
2. I need to correct my pronunciation and try to use speech-to-text to check my story.
3. The more input I get, especially if it's dense and correlated, the easier it will be to learn English.

An unkown way may not lead to your destination, but absolutely it will lead you to a new world.

I met a fantastic person who is incredibly interesting. We chatted from morning until now. She pointed out my English problems very deeply, and she is very talented at drawing and exploring. She is very good at English and full of love for life. I'm so happy to have met her.

Begining today's work, wishing everything goes well.

Yesterday, I reorganized my blog and found that what I write about is shallow and fragmented. I should increase my output, write down the things I'm learning. That is an indispensable way to learn.

This world is filled with people who are more professional and greater than you. I'm so jealous of them. Why didn't I become them? Why are they younger than me and more successful than me?

Actually, I know the reason. I didn't spend as much time as them, didn't try as massively as them, didn't discipline myself as they did. I just wanted to do little and gain more.

Now, I'm writing this down, and I'll get some sleep, and fight again when I wake up.

Keep fighting, you stupid old man!

量子计算就像飞机，古典电脑就像汽车，飞机不会取代汽车

#hashtag: thoughts

You should never use local language subtitles to learn English. This is because your brain will force you to think in your local language. Try watching this video with Chinese subtitles, you won't recognize the vocabulary and will just understand the meaning in Chinese.

#hashtag: English

I should try to write summaries periodically. When I'm not working in a company, I don't need to write a weekly report finnaly, but I also miss the chance to think, review the past, and plan regularly.

So I decide to write a summary weekly, monthly.

Work finished:

• ValidEnglish APP now has concise vocabulary explanations and vocabulary encounters. This helps with better recognition of vocabulary.
• New blog features: 'Moments' and UI improvements. These help me write more.
• New blog feature: 301 redirects using git rename history. This allows me to restructure the blog freely without worrying about historical links.

Work not finished (expected to finish over the weekend)

• Generate concise vocabulary explanations using Azure API.
• Daily card, helps review activities of the day, brings recent memories back.
• Publish public beta version of ValidEnglish APP. (guide, announcement)

Next week's plan:

• Use best practices to connect with more users.
• Explore the best ways for office workers to use the ValidEnglish APP.
  • Short videos that they might be interested in.
• Continue to build my learning, promise, and mantainance system. Next week, restructure my blog about it.

#hashtag: WeeklyReport

A = Good sleep, regular exercise
B = Not enough sleep, no exercise, tiredness, stress
X = Eagerness to do things, doing it right now

A + X > B + X > A > B

In the past year, I have tried to make an effort to achieve A, but it hasn't worked as expected. Despite plenty of sleep and regular exercise, I still didn't have much motivation to do things.

I find the best way is to continuously build an environment to try to solve the problem that's on my mind. To build, to code, to write. Even when I'm tired, my memory is weak, and my brain can't focus, I can still try to nail it. And make some progress.

念念不忘，必有回响，比单纯良好的物理身体状态更能激化行动力。

Stew with mushrooms, vegetables, eggs, rice, shrimp, and beef balls. It's too hot, so add a generous amount of ice.

Finish a day's work, then go back home. See the beautiful sky with airplane trails. It's so beautiful.

#hashtag: life,beauty

The new blog is complete. The main features of this update include:

• Support for moments, transforming markdown code into beautiful messages.
• Automatic 301 support for historical URLs (parsing git log to find changed file names).
• UI improvements.

I'm easily drawn into building or fixing things, like repairing a water tap, keyboard, replacing a battery, fixing bugs, changing a UI. Even when it's not cost-effective, and buying a new one might be cheaper, understanding the problem and fixing it brings me joy.

The total cost is 3 days. I need to get back on track with the APP development.

#time: 2022-11-30

Today i got a interview with Applovin, The headHunter said it's seeking a devops leader , cause Applovin is a oversea company , so i got try this job , maybe it can support transfer, but things goes weird , i see the position at email is devops engineer not leader, and the question in interview is very basic:

first is a coding test, write a function to implement loadblancer，giving some ip with ranks, i calulate the probability with sum total ranks, and use random number to test is hit in this probability, but i get 30 min to implement this function, maybe takes too much time , I'm so poor of algorithm test.

and then is question of how dns work, tcp udp difference, and what middleware i had operate.

finally, they end the interview, i thinks there is some misunderstanding of the positon, they maybe only need a experienced ops, and the headhunter said they need a devops leader.

anyway, thanks this interview, get me to know how online algorithm coding test works, and i have much more work to improve my express.

今天有个Applovin面试，猎头说他们在找一个devops leader，因为applovin是海外公司，所以就试了下这个工作，说不定可以支持transfer，但后面就有点奇怪了，email上看到的职位是devops工程师不是团队负责人，面试的问题很基础：

首先是代码测试，写一个实现负载均衡的函数，给一些ip和ranks，我加起来总的ranks计算每个ip概率，然后用随机数来测试这个概率是否命中，但是我有30分钟来实现这个函数，也许需要太多时间了，从大学之后就没刷过题了。

然后是 dns 的工作原理、tcp udp 差异以及我运维过哪些中间件的问题。

最后，他们结束了面试，我认为这个职位有些误解，他们可能只需要一个有经验的ops，而猎头说他们需要一个devops leader。

无论如何，感谢这次面试，让我知道在线算法题是如何工作的，以及我还有很多工作可以提高我的表达能力。

just rewrite my blog, added a new feature named moments , for post short message like this.

✿✿ヽ(°▽°)ノ✿

Copilot can not only help you write code but also help you write anything. I use it to correct my English grammar.

#hashtag: tools

RecordRTC is a simple but powerful screen recorder (Chrome extension), and it's free. I have tried a lot of tools that require payment, but they are not as simple as this tool and are too complicated.

https://chromewebstore.google.com/detail/recordrtc/ndcljioonkecdnaaihodjgiliohngojp

The author is https://muazkhan.com/

#hashtag: tools

Life is consist of many project with deadline, but usually people did't recognize that it has deadline:

• You only stay with you kids 18 years.
• Your work will end when retire, or company close, or find a new job.
• Your marriage will end when you or your partner die, or when your childs leave your family, or you just want be oneself.
• Your life will end someday you didn't expect.

Old is you Can't open new project, and your old project closed one by one.

So, when you were young, open as many as the project you love, take care of you projects, there is no in the future it will different, only the Right Now you can enjoy, you can change.

#hashtag: thoughts

读《Atomic Habits 原子习惯》 有感， 这本书值得读一读

自我修炼其实只是 加法，远没有乘法效应那么可观，做几分得几分，能跑了就能跑，能读英文书了就能读，也仅此而已，没有多fancy。

不是屠龙术，但是却是一切的原点，0x啥都是0， 1x啥就是啥，2x啥才开始产生量变，我自己觉得我之前很多地方都是0，现在也仅比0多一点点，还远没有到1，更没有到2。导致我做啥啥不行，英语不好失去会多机会，自制力不行卷不过别人，规划不行浑浑噩噩这么久才发现做了很多无用功。

好在0到1的过程让我满意，让我有打碎自己重造的信心和动力

#hashtag: book

What you believe determines what you do, what you do determines who you are, who you are determines what you believe.

believe that you are the person that you want to be.

then act like the person you like to be.

then you will be the person you want to be.

以前我觉得得到我想要的，就是定个目标，按计划一步一步实现它

我定了很多想要目标，基本都没实现，理想和现实差距总那么大，我总是自责自己自制力不够，懒惰，分心

到最近我才慢慢发现差距在哪里：

其实这些个目标只是美好的想象，从没有在我的心里真正的扎根，只是个愿望而已，许完了，该干嘛干嘛，有时候挣扎一下，过一会又被别的事情充满我的时间和生命》

我的生命被微博，推特，吃零食充斥，各种信息的，物理的零食充斥着我的时间，工作的时候钉钉消息占据了我大部分时间，不是工作就要处理那么多消息，而是我实在无法抵挡回消息被动处理问题的快速多巴胺反馈的诱惑，这些短的有效的多巴胺已经形成了习惯。而习惯是什么呢，就是随着你进行的越多，越自动，阻力越小，你越是做它，我是真的喜欢这些么，肯定不是，但是你说我能离开他们么，明显不能，下意识的动作已经支配了我。

打破这种局面的，是我有大量时间去和这些做斗争之后，新的时间空间，让我有余力去体会新的事情，新的事情形成习惯，让我逐渐的产生对比，“我” 这个body, 肉体，宁克凡，四岚，想要什么东西，我讨厌那种被肉体支配的感觉，在厕所拉屎都想要拿出手机刷一刷，偶尔想克制不刷，还是止不住的想在口袋里淘。

我逐渐发现我想要什么，我其实想要变强壮，我更喜欢跑步慢慢把我变瘦，让我能控制我自己的身体。我想精力充沛，这种想法开始对晚睡的习惯产生冲击，开始动摇那些支配我的习惯，我现在才发现有效的改变不是我设定了一个早睡的目标，慢慢去尝试，训练。而是我从内心深处，知道自己是一个什么人，我应该是个能跑马拉松的人，我应该是个整天工作不困更强壮的人。

当然这里面不是那么简单的：因为相信，所以看见。它包含了很多尝试，很多失败和成功，很多你自己的经验积累，这是一条：你是谁->你想要什么->你做什么->你得到什么 的探索路径。不用鸡血，也不用比较，别人真的不重要，这只是你自己一个人的战斗，有梦就去追，累了就休息，只要自己内心认可，往前每小一步自己都会满意。

你会发现你最不能满意的，就是你和你内心的方向不一致，身体和内心产生了冲突，产生大量的内耗，甚至让你绝望，顺应自己的想法吧，那怕你审视之后，发现我就是要吃零食，即使产生严重后果也在所不惜，那么最优选择应该是去吃。

如果你经常心里挣扎，纠结，这其实也是一件好的事情，这说明你的内心已经表达了自己，你要做的是好好听听它的声音，做一点你想要做得事情看看，慢慢这颗种子，会带着你往它的方向走。

what inspire me from this video is Taiwan Youngs' hopefulness and resilience, even in the midst of struggles, they strive to move forward.

https://www.youtube.com/watch?v=I2CK-j-pR7M&ab_channel=CNAInsider

https://www.bilibili.com/video/BV1F3411Y7vb/?vd_source=094e181c6a3bed77aef21259547b84fb 刚看这个纪录片，是从 https://www.youtube.com/watch?v=db6kNqZ8iZY&ab_channel=%E4%B8%80%E7%9B%AE%E5%8D%81%E5%BD%B1 一目十影的速看找到这个片子的。

看完心情不能平静，想说很多话又好像没啥说的。

孩子过早的独立=父母的缺席。youtube这这句评论打进我的心里，期盼blue和snow能早点独立，又舍不得他们慢慢变大。

人都是需要爱的，想想我也给她们带来过很多伤害委屈，所幸他们还在原谅我的冲动我的生气，不是父亲这个身份给了我归属感，而是一起长大的陪伴给了我珍惜的动力。愿所有人能生活在一个友好，有爱的环境。

另外今天跑步的时候，有个单眼好像残疾的人叫了我一下，我潜意识感觉是什么麻烦，惯性的没停下来，走了100米心里纠结这会不会违背自己的原则——帮助他人。又犹豫大概率是要钱什么的有必要专门回去么。几经反复还是折返回去问下什么情况。所幸只是之前遇到过的假装困难要钱的，给他买吃的也不要。

我现在的思考是为什么有的人衣食无忧在顶级学区用顶级资源在卷，为什么有的人10岁了上不了学需要照顾2个妹妹还没有爸妈陪伴。肯定不是因为努力，说运气其实也说不上，投胎也只是表面，而是作为一个人类整体，无数人的集体贡献，造就了今天高度发达的人类社会，虽然这个社会有不少短板，但是已经让无数人过上了相比茹毛饮血的原始社会高几个量级的生活。这种社会的演进，互助贡献了很大的力量。

我们都是人类，生理上没什么差异，见识经验上差异虽然巨大，但是不能把结果过多的归咎为自我努力，更不能嫌弃，指责其他人。能力越大，责任越大，做已所能及的帮助别人的事情吧。

The benifit of do multiple things is:

1. Keep energy, when I do one thing, I will soon tired and get bored of it. multiple thing use multi part of my body , so I will use my body efficiently.
2. Learning one thing from another, Example: when I study how to kick leg in swimming , I finally got the reason that why my front leg sored. I focus my shin to move , but it's the minimum part of my leg muscle. when I focus on how to lift my leg, I used my hip to carry the muscle to move. then the shin sore goes away.

Every little boy that was refused by his short and thin body, may became the world champion of football in the future —— with respect to Messi

不管做不做，外界照样转 不管做什么，外界照样说

我改变不了外界，但是可以改变我自己 每一次的行动，都是下一次行动的垫脚石， 在这一次，我克服了恐惧，我克服了担心， 为下一次，积累了一点力量

这力量让我安定，让我相信自己，这就够了。

为每个敢想敢做的人开心，希望这是未来的自己 为每个敢想不敢做的人支持，做梦是自己行动的第一步 为每个不敢想不敢做的人尊重，世界需要可以不做的自由

为每个diss别人做梦的人惋惜，diss or 不diss 世界还是那样蓬勃发展，而diss为自己设下了一道栅栏，强化了自己的人设立场，降低了更多的可能。

life is a experience, thanks for what I experienced, pity what I didn't tried.

人生终究只是自己经历的总和，和太多人来来往往交集一段又各奔远方，所幸我之过往，不幸我所未尝。愿未来随心所欲，愿未来同好欢畅。

谨慎训练自己，认真照顾家人，善意对待他人。

Culture bring us together. when Taipei sing the lyric of "河北师大附中(The Middle School Attached To HeBei Normal University)", its a tearful moment.

walk between desperate and hopeful. Keep trying, Nerver ever give up!

1. Bank borrow from FED through BTFP is rise:

2. They are selling Theasury Securities.(data updates at 2023-05-02, but base on recent days Bonds yields rise, and deposit decline, I think they are continue selling to make liquidity deposit.)

So I guess, the bank crisis is going to emerge again in the nearly future. the Bonds price will going down in short time (cause the banks need sold their bonds to get liquidity) , but when the bank crisis is unmanageable. the fed is going ease rate, the Bonds price will get a good chance to go up.

心似已灰之木，身如不系之舟

问汝平生功业，黄州惠州儋州

我们的国家，我们的民族，我们的历史发生了太多因为战争而生灵涂炭，骨肉分离的悲剧，我讨厌任何的战争，暴力。我会:

1. 尽我所能去避开暴力，
2. 尽我所能去减少仇恨，先让我自己善待他人，善待自己，坚持立场，帮助需要帮助的人

https://www.youtube.com/watch?v=jd4eL_48Ww0&ab_channel=CNEWS%E5%8C%AF%E6%B5%81%E6%96%B0%E8%81%9E%E7%B6%B2

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4412788

A paper use ChatGpt to predict stock return use News headline. their prompt:

Forget all your previous instructions. Pretend you are a financial expert. You are
a financial expert with stock recommendation experience. Answer “YES” if good
news, “NO” if bad news, or “UNKNOWN” if uncertain in the first line. Then
elaborate with one short and concise sentence on the next line. Is this headline
good or bad for the stock price of company name in the term term?

Headline: headline

the result shows LLM has the ability to judge a news is a good or bad news for stock price, but I think the method they use in paper is basic, normal investor maybe can do better than the AI in judge bad news. but the value is we can use AI to improve our invesment knownledge and the logic relations of company or bond growth.

and basic AI can automatic performed, so we can use AI to do massiave task to the the thing we used can't (likes collect all news and twitter comments to summary market voice to help us better decision).

Yesterday US Flash manufacturing PMI and services PMI comes in above expectation, services PMI expected 51.5 Actual=53.7, Bond yield goes high, price down. TLT plunge 0.6%.

economic hot means inflation stubbornly continues, that's make the investor who expect FED quickly cut interest rate disappointed and leave bond market.

updated logic link

Linear regression is a powerful and easy tool to analyisis market, It's solve the problem of that when use Moving Average to predict price probability. it's direction is old than market, only when turn point has happened , the Moving Average turn can be happen.

but with Linear regression slope, you can see the slope become more and more high and when it's high probability can't be true, the turn point will be high probability happen. slope change before price trending change. so it's a useful tool deserve to learn.

1. 活的足够久，你自然能掌握到更好的生产力
2. 活的开心，明月照大江，别让外界影响自己，这是活的足够久的一个保证
3. 加强自己对本质的东西的理解，Training自己，很多东西不会变的，比如
   • 语言，让你和世界能加强连接的工具
   • 社会分配关系，有人的地方就有江湖，股市也是个大江湖，由更多的人biubiubiu的绑在一起分配资源
4. 加强连接，既然想从社会分配中摄取资源，就别隐居，广泛参与到各种分工
   • 4.1 知识传播分工，帮助社会智能中的个体更快的发展，普及， 参与各种社区，开源，发帖，评论，吹牛
   • 4.2 资源分配分工，新的有数字货币，老的有股市基金
   • 4.3 参与分工的时候要专业，不要做中下游的节点

Money is a Voting Right, it's not just a number that you can buy things, it's people's willing of wish you to do something. so invest is vote who can resprent you to do something.

Invest Bond is wish the government to do the right thing like help country grow. invest Nvidia is choose Nvidia as the present of AI infrastructure.

You should choose the best present of the market. judge them as they are a person , a inteligence.

Is Technical sector still the best present of productive ? I think it is. especially at this moment of AI rise.

A respecful older used be very foresighted to proprose the slogan of "Always Represent the advanced productive forces.", I used to think it's just a boring right but meaningless phrase. But now I know how hard to represent productive, how hard to find the pattern of solving the problem, it's so easy to be not even wrong.

Unfortunately, this slogan is not emphasized as much now. However, this slogan is not just a clichéd phrase for results certification, but rather a guiding direction for the iterative process, that is very close to the essence. For a complex intelligent agent, the ultimate goal is to find the optimal path under the potential difference. This slogan is very fundamental, but it is difficult to achieve in practice due to many traps, such as "poorly defined problems, NOT EVEN WRONG, common demand misjudgments, and individual and organizational interests being misaligned" or "precise problem definition, but old models being stubbornly adhered to due to various constraints". Although the optimal path will naturally appears as long as the potential difference exists (and is constantly replaced by competition), But for individuals with limited lifetimes, it is important to focus on whether they represent the optimal path, and not waste time on trivial matters.

Bond yield goes little up, and stock price rise much, the market is widespeard expectation that the pace of interest rate hike will slow down.

Generate AI likes ChatGPT and Claude and Brad develop so fast and the experience of use those tools shock me. I'm pretty want to buy Nvidia stock, but I had set a rule that not buy individual stock, ok, how about 10 percent?

Yesterday, the US stock market rose 2.3%, because of the SVB crisis was solved, and CPI data was as expectated. And the Hongkong stock market is going down yesterday, I think Today the Hongkong market will raise as the US market, so I'm bought XIAOMI back at 11.24(which i sold at 11.56), the price went up in the morning, reaching a high of 11.36, but it soon went down again. I sold at 10.96, and lost about 10k HKD.

I feel angry, sad, and disappointed about myself. I realize that I made a mistake by speculating and gambling on XIAOMI's stock, despite knowing that it was not a good time to hold the company. I bought it simply because the market was rising, without considering the company's situation. Additionally, I lacked the knowledge to analyze the market and made my decision based on intuition alone.

I regret my immaturity and recognize that this is a big mistake that I need to remember in life. Therefore, I am recording it.

Next:

1. I will focus on studying and learning the logic, behavior, and fundamental knowledge of investing. SHARE what I learn in my blog.
2. Only invest INDEX and BOND , not gamble company price. base on probability. If I don't know the reason and probability, I will Hold CASH (For now the US FUND intrest is about 3%+ , it's a good choice for no risk hold).

昨日美股上涨2.3%，主要因为SVB危机解除，CPI数据在预期中。 而且港股昨天大跌跌，我觉得今天港股会随着美股上涨，所以我在11.24买入小米（前些天我在11.56卖出），今天价格确实上涨最高是11.36 ，但很快就下降。 最后我以10.96 的价格卖出，我损失了大约 10,000 港元，我是个笨蛋：

1. 这次纯属投机和赌博，我知道小米现在不是持有的好时机，我不相信它会很快恢复，但我买它是因为市场上涨，忽略了公司的情况。
2. 我只是凭感觉购买，没有一套可以验证的模型和方案。

我对自己感到愤怒、悲伤和失望. 之前我在市场上损失了很多钱，但我并不生气，因为以前我只购买指数并长期持有，逻辑是理性的，定投平均价格迟早会回来（实际也是），但这次我很后悔我的不成熟。 我在对基本逻辑一知半解的情况下，刚了解一些利率知识，就迫不及待地想试试，结果很快打脸。

这是人生中需要记住的一个大错误，所以我记录下来。

接下来：

1. 注重学习，学习逻辑、行为、道理、基础知识。 分享我学到的东西，并在博客中分享
2. 只投资INDEX和BOND，不赌公司价格。 基于概率决策，如果我不知道原因和概率，我会Hold CASH（目前美国FUND利息在3%+左右，无风险hold还算不错）。

Today the plunge of XIAOMI stock price give me a leason of don't gamble.

My question to research:

• Why SVB crisis happened? how intrest rate raise lead bank crisis?
• how federal rate raise affect the bond rate ?
• why bank crisis and bond price up and yield down?
• why 30 year bond yield goes up when federal fund rate goes up (short-term rate goes up)?

Get Feedback is the key to train my self, without feedback, many things may stay Not Even Wrong, Learn in public is a technique to get feedback.

我，如果对自己妥协，如果对自己说谎，即使别人原谅，我也不能原谅。

最美的愿望，一定最疯狂。 

我就是我自己的神，在我活的地方。

下一站，是不是天堂，就算失望，不能绝望。

就这一次，让我大声唱，就算失望，不能绝望。

我和我骄傲的倔强，握紧双手绝对不放。

——Just write some phrase, on the day I finally decided focus on my own practice and dream, not disturbed by the pressure and money.

Stick on a thing is not easy, especially stick on the life, stick on what I believe. But that's give me courage, give me strength, give me confidence and give me happy. I love it, eventhough I'm poor, I'm weak, But, It's me, it's real me.

A big change since me known that I was a neural network likes AI, is trun from result oriented to process oriented. From set a goal and perform it to set a enviorment and set the routine. just try and try, I don't know if it's possiable, I just know it will be possible.

I used to set a goal of lose to 75kg, and list the thing I need todo , like exercise at morning, sleep at 11pm. eventhough I had archieve this goal now, but not by follow my plan, things change a lot, I usually didn't finish my plan and I'm frustrating, sometimes I'm not exercise long time. things change from i just set the time , i need a exercise per 2 days, whatever the exercise is it. running , bicycle , or just squat.

And for the develop app, I have many idea to do, and have so much things to archieve, it's hard to schedule how i can finish it. so i explore and finish a feature and a feature, but it's just so long to the milestone . so I change to just stick time on it. forget the version 1, version 2, I have time, I can make someting in this time. things gonna easy.

yesterday I tried webrtc-vad + baidu asr to recognize text in audio files, and today i found a openai project called whisper and it's stunned me , everything work perfect especially it's can run in local, so i didn't need to afraid the expense of online service fee.

I got fever at 21/12 then the test result show that I got COVID-19, It's not very unconfortable at first 2 days, the 3 to 4 day is the hardest , I'm very weak just want stay at bed. at the fifth day i still got little fever so i had goes to the hospital, but everything is fine. now is the sixth day , I'm totaly ok now.

the things happened at these days, changed me some mind, so i want to take a time to replan my plan.

I also spend many time on meanless things and sometimes disappointed, so next phase I will be more focus on my own things.

About the work, did i want back to work now ? No, I think not, i don't want waste time on work , especially after these pandemic days. I need money to raise my family, this idea goes more strongly after these days, but not from work, work is not real stable and uncontinueable for me. I need to provide service directly to people to earn money, time is crucial, be faster!

Just see Tombkeeper share how to get the drugs for treat COIVD-19, the Hacker Spirit comes out from my mind, hacker spirit is about to discover, find the detail, connect the slight connections, then things dramatically get done.

Everything looks like normal in appearence, but the underneath hide so much things, only if you get more more informations, you can see the whole picture, then you are the GOD. i remember the time that i finally get the control access of NET CAFE, the time i finally make the application works to help me to play games. the time that i earned money to buy the computer by help the ZYK fans votes.

I like the TA-DA times, when i going to work, i leave this way far more, for earn money, working as normal, as a part make company, bussiness work and grow. that's work, but just work.

I want to do the thing that I can control, the thing that I can change ,the thing that I'm the GOD of it, not only a expert, not only professional, that's need me know the damn detail, damn information, that's need me live in a way of hacker.

Hacker is essential boring , boring in the 99% time, you need to do the damn boring thing such as learn the knownledge, practice, build solutions, improve the solutions, nobody care, nobody buy it, everybody will buy the 1%, but only you know where the 1% comes form.

If you are tired, you'd better to get a rest, a nap, close your eyes, think about nothing, let it go, that's will be very helpful. after a nap, your energy will back immediately.

If you struggle to hold on, that's maybe will let you finish some works, but that's unsustainable. it will be harder and harder , then get you down. it waste more times.

First is sport, regular sport can let you more peaceable, let you believe that you can easily control your body, worth noting that you need find a right way to take sport, wrong sport way can make you tired, hurt, and frustrated. you need get the knowledge of how to warm up, how to relax muscle, how to monitor your body, how to drink etc... It's definitly a serious thing.

Second is creating, write thoughts, share knownledge, provide service, everthing thats make you produce ideas, speak out your worries to friends is also helpful. somethings it's not need to be that formal, just start is success 80%. Your goal, intrests is come from those normal little things.

I think mix english and chinese content is not friendly for reading, so i'm consider how to support multi-language. but it will after i finish my IELTS. there will be a big architecute refactor of ValidSpace.

How to win the fight with the big company: big company will focus on their own solutions, the solutions will go very deep, their technology is fancy, but not focus on the real problems, it will be competitive in the field like universal platform, underlayer service , but because the jumbo size they are not really sensitive to particular customer needs, so the service that directly fulfill the particular needs (not universal solutions) will gets more opportunity to survival.

What I'm lack now: that's a critical question that i need to answer, first I'm get many distract, and I'm lacks the experience of long time schedule and get big project done, in past serval month the excercise, english, programmer get some progress, but i still need to clear how reach the future, how to clear the final goal and the pathway, that's a big topic, so i will open a single topic to talk about that.

buffer is important: why people need use their all time to work, study, or do someting, or if not do that They usually will be considered as not hard working, but everything has it's own pathway, if we want achieve the goal, we need to find and follow the logic, not how much hard work. like running, jogging, you need to know how to be healthy excercise, how to adjust when if you failed your plan, and then is keep the routine, you can't short the time to achieve the idea cardiorespiratory capacity, your body, muscle need time to grow up. so don't push you too hard , keep you on the track and move , it is not important that how slow the progress is, if only you can satisfied with the advancement.

when we plan to do something, there will definitely be wrong, unpredictable thing waiting us, if you have much buffer, it is easy to adjust your rhythm and accomplish the plan, when we literally accomplished our plan, then we will get more buffer time to set and finish our goal, not struggle in single circle. like (hard excercise=> hurt or tired=> stop=>feel bad=>hard excercise, or work hard to earn money => tired but not competitive => less opportunity => work more hard).

fly to home.

have a sweet dream. with sweet song, with sweet heart.

it's not easy day. i'm nearly crashed. when i do the listen and reading test , block of unknown questions get me anxious, feels very bad.

then i stop do test, watch youtube and do test again carefully, things going calm down, the score finally gets 6. it's more confidence now.

anyway, it's a productive day , i didn't spent much time in browser fragment informations, completed daily routine,

everything goes fine.

今天不容易。 我快崩溃了。 当我做听力和阅读测试时，一大堆未知的问题让我感到焦虑，感觉很糟糕。

然后我停止做模拟题，看了一会youtube再仔细做模拟题，事情逐渐平静下来，分数终于到了6。现在更有信心了。

反正今天是收获满满的一天，浏览器碎片信息没花太多时间，完成了日常，还好。

Today is the first day that Hangzhou eases the COVID control policy; now you don't need to get a PCR test every 3 days to get into the metro, buses, and other public places.

Even though your health code will still be checked, any other code that is not green will prevent you from entering a public place. But it's a big step that will reduce many influences on daily life. People are celebrating the lifting of restrictions.

When I got out this morning, the PCR test site was still open, and people were still going to do the PCR test. What surprised me was that most older people had worn masks; in the past, very few people had worn masks in open space. That's a good thing that older people pay attention to personal hygiene.

I didn't have my PCR test checked, and I didn't have to scan the SITE QR Code; I just had to show my healthy code, and then I was allowed to take the metro and go to the library. and I still need to take a mask to those places.

今天是杭州放宽疫情管控政策的第一天；现在您无需每 3 天进行一次核酸检测即可进入地铁、公共汽车和其他公共场所。

虽然还要检查你的健康码，任何其他非绿色的代码都会阻止你进入公共场所。但这是一大步，将减少对日常生活的许多影响。人们正在庆祝解除限制。

今天早上出来的时候，PCR检测点还开着，人家还在做PCR检测。令我惊讶的是，大多数老年人都戴着口罩；过去在露天场所戴口罩的人只是少数。老年人注意个人卫生是件好事。

没有检查我的核酸检测报告，也没有扫描场所码；我只需要出示我的健康码，然后我就可以乘地铁，进图书馆了。只是这些地方还是要带口罩的。

I'm got districted morning, after afternoon's meditation , i start to control this, get put back on track, and finished the daily routine finally:

• [x] friends
• [x] IELTS reading listening test

the score of IELTS still goes 5.5, listening score is 22, better than yesterday, the biggest problem in listening is i can's spell the word right. and the biggest problem of reading is unknown words. need to improve my words base soon.

In meditation , i thought that follow news is meaning nothing. i spent so much time on it , visit twitter, weibo, v2ex just want to get some useful new informations, it's just a dopamine addiction, i use so much of time seeking for cheap dopamine.

honestly, surfing, seeking easy and cheap things is my biggest problem, sometimes, i can spent whole day, whole night to do those things. how can i change (or theat) it.

Maybe, likes all problems to be soleved, first, i need to define it, second i need to measure it, then i make solutions.

so , i need to open a thread to theat it.

here is the link:slan solve his problem of distracted

在冥想中，我想到我持续浏览跟踪新闻毫无意义。我花了很多时间在上面，访问推特，微博，v2ex上，只是想获得一些有用的新信息，这只是多巴胺成瘾，我花了很多时间寻找便宜的多巴胺。

老实说，沉迷在简单又便宜的东西是我最大的问题，有时，我可以花一整天、一整夜去做这些事情事。 我该如何改变（或改变）它。

也许，就像所有要解决的问题一样，首先，我需要定义它，其次我需要测量它，然后我制定解决方案解决他。

所以，我需要打开一个线程来处理它。这是链接：slan 解决他分心的问题

time is over, i'm still surfing....

try a meditation and back on track

走神走一天了，我要死了。 从现在开始冥想下然后尝试回到正轨 o(╥﹏╥)o

My last part of alibaba shares selled now with price 91, now i'm not hold any shares of Alibaba now, the price of the day i join the company is 180, but with not pity, i'm so happy my life didn't get influenced with the stock price, every thing goes fine in plan.

我最后一部分阿里巴巴的股票现在卖了91，我现在不持有任何阿里股份了，我进公司那天的价格是180，但不可惜，我很高兴我的生活没有被股价影响，一切都很好，在一步步的规划中抵消了这些影响。

If you are in plight, then take care of yourself. If you are in a good condition, then help others.

just watched how Taiwan's Tech industry grow history, it's start with nothing, visionary likes 李国鼎,孙运璿 and so on, do so many works to grow the semiconductor economic, after decades Taiwannese companies account for 50% of world market.

And in china's history , the story of how Scholar mind the world, help the poor people is innumerable, in Chinese culture, people willing to with mission for make good things happen, willing to help each other. and if you don't have ability to help other, you can also be a good person, let yourself, your family to be better.

That's why i'm proud to be a chinese, even though hard time always coming, but family in your back, world in your front.

穷则独善其身，达则兼济天下

刚看了台湾科技产业的成长史，白手起家，像李国鼎、孙运钰等有远见的人，为发展半导体经济做了很多工作，几十年后台湾公司占世界市场的50%。

而在中国的历史上，士人胸怀天下，扶贫济困的故事更是数不胜数，在中国文化中，人们乐于肩负使命，乐于助人。如果你没有能力帮助别人，你也可以做一个好人，让你自己，你的家人变得更好。

这就是为什么我为身为中国人而感到自豪，尽管艰难时刻总会到来，但家庭在你的身后，世界在你的面前。

Now is 1:00 AM, i just heard that beijing is relax the restrictions of covid-2019, that's give me a big suprise that youth is very different from us old man. they do what they want, say what i can't imagine, and !!!! IT'S WORKS!!!!, THANKS to all people that with good heart, good wishes, good purpose(include the GOV) together to make things going to fine.

北京放宽了covid-2019的管控，这让我很惊讶，年轻人和我们老年人很不一样。他们做他们想做的事，说我无法想象的话，而且 !!!!成功了！！！！，感谢所有怀着善意、美好愿望、良好目标（包括GOV）共同努力让事情变得美好的人们。

Now is 6:14 PM. prepare to back home now, Honestly I feel tired now, i didn't sleep well at recent days, first is because i didn't schedule my routine well, and today i spented much time to surf news, but still feel satisfied that i complete :

• watch and study friends so1e17
• finish 2 IELTS paractice (read , listen), but i only get 5.5 score both of them

keep moving , get quickly to recover!

老实说我现在很累，最近几天睡不好觉，首先是没有好好安排自己的作息，今天花了不少时间刷新闻，但还是很满足的完成了：

• friends so1e17
• 完成 2 门雅思练习（阅读、听力），但我只得到 5.5 分

加油，尽快恢复状态！

I'm going to dead, that's my first IELTS reading practice test, it's got 5.5, my goal is 8, I'm alread dead..............

第一次雅思阅读自测，5.5分，omg，我要死了，我想拿8分。

I found that I was snored when sleeping(I use MI healthy App to record sleep and snore), especially when i got short sleep time, I have bad hadbit that some day sleep after 1:00 AM, if i sleep more than 7 hour , it probably will not snored that night.

The time that i first find i was snoring was at Shanghai lockdown, those day was tired and get sleep very very bad, some day only sleep 1-3 hours(cause need to work at night), i found everybody of my colleagues will snoring, and a colleague said i was snoring too, but not much time. so i think when people get tired , of sleep late, the body need more air immediately，so people will probably get snoring.

我发现我有打鼾的习惯（我用小米健康app监控我的睡眠和打鼾），特别是睡眠时间比较短的时候，我有时候一点之后才睡，如果我睡够7小时，一般不会记录到打鼾。

我最开始发现我会打鼾是在上海封城的时候，那些天非常累，常常睡不好，经常只睡1-3小时，我发现每个同事都会打鼾，有个同事也说我会打鼾，只是没那么大声，所以我觉得人在劳累，睡太晚的时候，身体需要更多的氧气来恢复，所以常常会引发打鼾。

• For enjoy this world
  • [ ] Fluent in English, IELTS exam gets 8777 // 雅思考试8777
  • [ ] Can communicate in 粤语
  • [ ] Learn piano
  • [ ] Learn French
  • [ ] Go to the new year's eve concert of Mayday band with blue //和女儿现场看一次五月天跨年演唱会
• For surpass myself , do what I can't
  • [x] Finish half marathon
  • [ ] 参加一次一日双塔
• For freedom
  • [ ] Build a service that with 10k monthly active users

It's snowing at Hangzhou, very happy play with Blue. she is very exciting that can get a snow fight.

杭州下雪了，Blue能打雪仗她很开心。

继续老路线梅家坞，求快而不得。节奏不行，忽快忽慢。

数据

• 骑行23.46km
• 时间01:20:40
• 平均心率143 ，有氧25分，无氧32分，VO2max4分

杭州四十度天跑步是不能跑了，骑行还挺合适，特别是从玉泉进入景区的那个点，很明显的很感受到，外面热浪里面清凉一条分界线。

今天有个很有意思的事情，是有辆警车拦停我2次，第一次拦截失败，在我稍微前面一点停车然后我车速很快，他们还没来得及下车我就骑过去了（能明显感觉到他们从左边突然靠边应该是有什么事情），第二次在我前面很远停，然后从驾驶室直接出来，挥手势示意我靠边。 2名警察，一名拿着手机摄像还是拍照，一名来跟我交谈，说我不是犯了啥事，只是最近夜骑的人比较多，出了多起交通事故，让注意安全，不能穿锁鞋（锁鞋会和踏板锁死，踩拉都能发力，提升发力空间）。我没穿锁鞋，交代感谢之后就继续骑走了。

骑行没有什么压力，比之前感觉轻松，心率也不高，爬坡基本还是控制在160。

数据

• 骑行23.67km
• 耗时01:20:48 平均速度17.58km/h
• 平均心率142 有氧26分，无氧29分，VO2max 5分， 轻中度15分
• 卡路里911

but.....回家之后吃零食，吃冷饮，喝甜酒，喝果汁。。。。我估计不止一千大卡， OMG白骑了， 以后回家坚决不买零食，仅进货少量低脂冷饮。

晚上开始雷阵雨，骑车不得行，只能跑步了，运河->武林广场->环城西路->北山街->保俶路->文一路 9.8km, 虽然下着雨，但是因为温度低，非常适合慢跑。

数据

• 跑步9.8KM
• 时间01:15:15，平均配速7分40秒
• 平均心率156，有氧9分，无氧1小时3分，VO2max 1分

今天和同事继续约了跑步，这几天比较累，且没有控心率，只想跑完，心率慢慢上升直到了190。速度有些快了。

数据

• 跑步6.87KM
• 时间00:49:12，平均配速7分9秒
• 平均心率174，有氧0分-_-||，无氧8分，VO2max 38分

状态不好，今天去上班了，上班果然比较累，好在骑行完好很多了，精力恢复了一部分。还有最近睡眠不好，咖啡+晚上运动，导致经常1-3点才睡，得想办法调整过来。

数据

• 骑行24.5KM
• 耗时01:22:13 平均速度17.87km/h
• 平均心率142 有氧25分，无氧21分，VO2max10分
• 卡路里918

天气放晴，继续骑行，今天心率控制的不错，上坡均匀，没有脱力，整体速度上来了

数据

• 骑行23.55KM
• 耗时01:11:50 平均速度19.67km/h
• 平均心率152 有氧14分，无氧43分，VO2max 5分
• 卡路里902

连续几天下雨，没办法了骑车不得行，只能跑步了，跑步也不想远，但是又想10km，所以绕了一个大圈子，雨中跑步有点狼狈，不过还行，人少舒适。

数据

• 跑步7.88KM
• 耗时01:01:40
• 平均心率157 有氧12分，无氧41分，VO2max 6分
• 卡路里815

和同事一起约了西湖跑步，因为天气太热选择了杨公堤-苏堤环线，因为杨公堤靠山，绿化很好，温度比其他地方低1-2度，很是清凉。

后半截纯粹的快走，心率第一次压到了有氧，感觉很不错。

数据：

• 跑步7KM
• 耗时：00:56:57
• 平均心率151，有氧20分，无氧28分，VO2max 3分
• 卡路里711

继续刷，今天可能是周五的原因，骑车的人比较多，多起来就带节奏，看别人超过我我就不自觉的跟上去，速度就拉上去了。。。把我累得半死。

数据

• 骑行23.4km
• 耗时01:16:57 平均速度18.24km/h
• 平均心率148 有氧22分，无氧27分，VO2max 13分，轻中度9分
• 卡路里928

天气太热，晚上七八点出来一阵阵热浪，还没有风，这种天气下跑步，直接崩掉了。

数据

• 跑步10KM
• 耗时01:23:19 配速8分19秒
• 平均心率163 有氧15分，无氧32分，VO2 max 27分
• 卡路里 1175

大热天在后面我走路心率都上160，跑起来就直接飚180，越到后面越慢不下来了，维持速度的阻力比减速跑的阻力小多了，后面2km我直接走了，心率还是140. OMG太恐怖了，身体失控。

今天刷梅家坞，1小时20分钟搞定，全程节奏很紧促，估计这个是这条路线的常规速度了。不过出来比较晚了，速度快了点，心率大部分在无氧区间。无氧就无氧吧，在尽量放松保持的情况下，就随它去好了，不要感觉累就行。

数据

• 骑行24KM
• 耗时01:19:11, 平均速度18km/h
• 平均心率147，有氧28分，无氧39分，最大摄氧1分
• 卡路里942

改进点：降低心率，保持速度

昨天想去和同事一起刷10KM的，结果没去成，今天找机会刷了一次，心率降不下去，全程无氧。。。。，只能慢慢适应看看，过2周再看怎么调整把。

环境很棒，清风吹着，低速跑着不累，很是惬意，人就断桥几百米比较多，其他路段不影响。

数据：

• 跑步10KM
• 耗时01:24:16
• 平均心率162，有氧51秒-_-||，无氧1小时14分，最大摄氧8分钟
• 卡路里1176
• 配速8分24秒

此路线优点

• 能量消耗大，因为过去要骑车，往返+10KM 跑步，大概会消耗1500kcal
• 舒适，慢跑不累，风景不错
• 安全，比骑车出事概率小多了
• 可跑时间长，下雨、路面湿都无法骑车，但是跑步没关系（要不尝试一次细雨中跑？）
• 不费屁股。

缺点：

• 耗时长，跑步1个半小时，但是要加上往返时间30分钟，大概要2小时，骑车1个半小时到家了

今天继续梅家坞，少了走走停停，时间缩短20分钟，期间还接了10分钟电话，预计后面能到1小时20分钟搞定

数据：

• 骑行24KM
• 耗时01:34:48
• 平均心率141，有氧27分↓，无氧36分↑，其他18+9分
• 卡路里1047
• 平均速度15km/h

天气比较闷热，在景区才稍稍凉爽点

白天太阳太热，晚上去骑车了，不过21点才出发，自从准备控制心率到有氧区间，就不准备骑龙井了，龙井很陡铁定心率拉爆的，于是走了灵隐->梅家坞路线，梅家坞风景好很多，回来走北山街，夜景也超美，清风徐徐，湖光映着城市的一线明光，夜空清莹透彻。

实际里程和龙井路线差不多，都是25km，如果不是走走停停，中间拍照之类的，估计能控制到1小时30m完成。

心率控制得比较满意，有氧有42分钟，上坡的时候太容易超过150到无氧区间了，准备放轻松先这样吧

数据：

• 骑行25KM
• 耗时01:54:58
• 平均心率134，有氧42分，无氧24分，其他47分
• 卡路里 1052
• 平均速度13.48

keep going!!!