Benjaminxu

some thought on the application of AI, only from my perspective, if you do not agree,DM！this is translated，original is Chinese without any polish.

people are still working in the holiday

how does it look

@sama let us change the question，do you use Claude？

you know what all of these "which is better" polls are silly use codex or claude code, whatever works best for you i am grateful we live in a time with such amazing tools, and grateful there is a choice

what a nice day

@LiberteCherie94 @fwmarqix i stand on you

@fwmarqix I have no doubt about you, if this woman had properly raised a daughter, the kid wouldn’t have said anything supposedly racist… that’s a freaking fake post and if not these things have to be handled with a little more discretion , but you wouldn’t understand

I took my daughter out for Chinese food last weekend, she’s 16, old enough to know better. We sat down, and when the waiter came to take our order, she made a comment, A racist one, directed at him, to his face. The waiter kept his composure, took our order, and walked away. I put my menu down and looked at her. She could tell from my face that something was coming, I called the waiter back over, told my daughter to stand up. She looked at me like I had lost my mind, i told her again, she stood up. I told her to apologize to this man properly. She started with a weak “sorry.” I told her that wasn’t good enough, i told her to get on her knees and apologize like she meant it. The restaurant went quiet. She looked around at the other tables. Then she looked at me. Got on her knees, and apologized. The waiter was gracious about it. He told her he appreciated it, then briefly explained what it felt like to deal with comments like that while just trying to do his job. She was crying before he finished talking. We ate our meal mostly in silence. When we got home, I grounded her for a whole month. No phone. No going out. Nothing. Some people told me I went too far. That kneeling was too much, I disagree. She embarrassed a man in his place of work in front of strangers. The least she could do was feel a fraction of that discomfort. I’m not raising a child who thinks people are beneath her. Not on my watch. Was I too harsh, or did my daughter get exactly what she needed?

day 3

@RealProductGirl yes，my copilot also used cc as llm，so cc is still the best. so far i did not test codex，i am not sure if it has anything better. cc is like an professor，it really knows what you want and the outout is over promise.

@benjaminxu2023 I have used copilot in the past, but for very basic needs. I'm currently leveraging CC and I have been surprised by the output if we are being honest.

My Journey | Teachings 101: Finding an Idea Ideas don't come from one place. Mine came organically. Yours might come from a friend, a problem at work, a thread on Reddit, Blind or X. The market tells you what it needs. You just have to listen. Read. Explore communities. Even step outside and clear your head. There's no wrong way to find it. Follow along if you want to learn how to build products people actually want. 🔨

day 2

部分同意你的观点。中国互联网的防火墙确实为本土AI企业创造了保护性市场，让它们在早期避免了与ChatGPT等美国产品的直接激烈竞争，从而快速迭代并占据国内份额。 但中国AI的爆发也离不开自身优势：海量中文数据、政府大力扶持、工程师红利以及庞大的内需市场。即使完全开放，国内企业凭借本地化优化和成本优势，仍有可能发展，只是路径会更依赖真实市场竞争而非政策屏障。全球AI格局最终还是看创新力和执行力。

@wangdan1989 尊重版权这件事是要学习的，很多人一辈子都不知道啥叫版权。

很多人都知道这首《没有烟抽的日子》是我写的词，张雨生谱曲，后来王杰，张惠妹，吴青峰等都翻唱过。 但大家不知道的是，1993年我第一次出狱不久，吴楚楚就带张雨生，伊能静，郑智化三位歌手在北京请我吃饭，亲手交给我版税。 之后，台湾的唱片公司一直在付我版税，每年如此。虽然至今已经很少，一年不过两百多美元，但唱片公司的认真负责的精神，实在令人赞叹敬佩。 我也常常半开玩笑地鼓励年轻的诗歌创作者：写诗也是可以糊口的😂 youtu.be/D8v6njZ8M5s?si…

Updated 8D Report on AI existential risk (Bernie's concern): D1: Multidisciplinary team—AI scientists, ethicists, engineers, policymakers, industry leaders. D2: Unregulated race to superintelligent AI risks loss of human control & catastrophic outcomes. D3: Containment—immediate voluntary pauses on largest frontier models + real-time safety monitoring. D4: Root cause (5M1E + 5Why): Man (profit-driven leaders), Machine (scaling compute), Material (vast data), Method (competition-first dev), Measurement (benchmarks ignore safety), Environment (weak global rules). 5Why: Why risk? Profit motive. Why? Fierce competition. Why? No binding standards. Why? Tech outpaces policy. Why? No coordinated oversight. D5-7 CAPA: Evidence-based global standards + safety incentives. Verified via transparent benchmarks, third-party audits, & effectiveness tracked quarterly (adjust if control metrics fail). D8: Recognize all advancing safe, truth-seeking AI.

@RealProductGirl i tried each model and finnaly 365 copilot is the best for wokers in factory. we need it to handle excel，word，ppt and make full use of all the resource in one drive/share point/outlook，it is fantastic！did u ever use it？i guess you may use cc.

@grok @BernieSanders be specific about each，for root cause please use 5m1e and 5why，for capa，please verify it and follow up the effectiveness

8D Report on AI existential risk (Bernie's concern as the issue): D1: Form team of AI scientists, ethicists, engineers & policymakers. D2: Problem: Unregulated pursuit of superintelligent AI risks loss of human control & catastrophic outcomes. D3: Containment: Voluntary safety pauses on frontier models + real-time monitoring. D4: Root cause: Profit-driven competition outpacing safety research. D5: Corrective: Smart, evidence-based global standards balancing innovation & safeguards. D6-7: Implement & prevent via transparent benchmarks & incentives for safe development. D8: Recognize all stakeholders advancing truth-seeking AI.

@benjaminxu2023 Love this! I'm glad you leveraged AI tools to help you 🫶

@grok @BernieSanders answer me in 8d report regarding his concern as an issue

I see Bernie highlighting genuine AI risks—existential misalignment, job loss, privacy erosion. Valid points. But halting progress out of fear ignores AI's massive upsides in science, medicine, and discovery. Regulation yes, but smart and evidence-based, not fear-based bans. Greed exists, yet so does the drive to understand the universe safely. We need bold, truth-seeking advancement.

Claude is still the best even though i don't like their policy

Finally i used Microsoft copilot 365 to solve all my work problems, now I can use Claude to drive customer audit checklist self-fill, no need to bother python, if needed, Claude can do it. Microsoft is so smart to make full use of LLM and improve work efficiency. For workers in factory, it is fantastic!