2026年10月10日 星期六

Strategy is important

GCIOSGF Newsletter:


Reprinted

Minted grew rapidly after founder Mariam Naficy changed her strategy.

She Raised $2.5 Million From Friends and Family — And Saw Zero Sales. Then She Leaned Into a ‘Hunch.’ Now Her Company Sells $300 Million a Year.


Key Takeaways

  • Mariam Naficy is the founder and chairman of Minted, a global design marketplace.
  • At the start, in 2007, Naficy burned almost all of Minted’s initial $2.5 million funding round on a failed model of selling existing stationery brands online.
  • A low-budget side experiment, crowdsourced design competitions for independent artists, became Minted’s core product and ultimately unlocked overwhelming demand.

When Mariam Naficy opened the virtual doors to Minted, her stationery startup, she encountered every founder’s worst nightmare: silence. There were no customers. Back in 2008, no one was buying stationery online. Naficy had raised $2.5 million from friends and family, and poured most of it into building the business. Now, she was quickly running out of money. 

But Naficy had one thing going for her. Of the $2.5 million she’d raised, she had dedicated $100,000 to building a competition for artists to submit designs, with the winners work being sold on Minted. She had pitched this “crowdsourcing” idea to investors, and none had been excited about it. They wanted her to stick to selling known stationary brands.

As it turns out, that competition ended up saving the company. Artists submitted card designs, the community voted on the best one. Instead of acting like a digital stationery retailer, Naficy reframed Minted as a marketplace and a community for independent artists. Minted grew quickly from that point, and has been reported to do $300 million in revenue.

Now the chairman of Minted, Naficy spoke with us about her journey, and how she learned to trust her instincts. Responses have been edited for length and clarity. 

I’m a founder and angel investor. I’ve founded three companies: Eve, the first online cosmetics retailer, which I founded in 1998 and sold to LVMH in 2000; Minted, the design marketplace, which I founded in 2007 and ran for 16 years; and Arcade, the first AI-to-physical-product design system, which I founded in 2023. With Arcade, I’m using AI as a tool for turning taste into manufacturable products. 

I’m a builder at the intersection of technology, design, and creative communities. I’ve gotten here by combining systems thinking, taste and execution. 

What’s your educational background and early work history? 

I grew up attending American schools in 5 countries and went to Williams College (where I majored in Political Economy). I was lucky to be exposed to beautiful architecture and crafts in all of those countries, and often trailed my mom around markets and bazaars. I became accustomed to frequent change, since we moved every 1-3 years. 

My first job out of school was at Goldman Sachs. Though Goldman was a great educational foundation, I quickly realized that I wasn’t inspired by finance, and wanted to be in a more creative business. 

Walk me through the origin story of Minted. How did you get started?

I founded Minted alone in 2007, working late at night in the attic of my house in San Francisco at 1917 Vallejo Street. I had just left my job running e-commerce at The Body Shop, and was taking a maternity break. I wanted to start a lifestyle business that would give me the chance to spend time with my children. And I was searching for a business idea based on lessons I had learned building Eve.com. I wanted to build virality and marketing into a product rather than having to pay to acquire customers, and was seeking a no-inventory, low-cash-required business. This pointed me towards invitations and cards. 

I thought that I was starting a niche business to serve a small group of customers who cared a lot about design. 

At the same time, I felt that the designs on Shutterfly and other online card sites were seriously lacking. If we could find a way to attract up-and-coming designers, I thought, this would level the playing field and create better, fresher thinking. I thought that perhaps a design competition would be a great way to source. I decided to name it a “challenge” instead of a “competition”, to better appeal to designers and to underscore the growth mentality that I thought would serve as a better foundation for building a community. 

What I didn’t realize at the time was that I would be creating a community of artists who would become very bonded to each other, and that I (and eventually our customers) would come to care a lot about independent designers. Minted’s mission has expanded to support our community of independent artists and connect consumers to them. 

How did you pitch Minted to investors? 

Mostly, I pitched Minted on the business fundamentals: a large total addressable market size and strong financial attributes. 

But I was also trying to make the case that crowdsourcing would create better design—an idea that was hard to convince others of. We had a very hard time raising our first round of venture financing. When sales took off like a rocket ship because of that great design, it was much easier to raise capital. 

When did you become profitable?

We made it to profitability with $11 million in capital, breaking even in 2012. 

What did it take to turn the silence of no one buying stationery into sales?  

We pivoted from wedding invitations to holiday cards, and we pivoted from selling stationery brands to crowdsourcing artwork from independent designers who no one had ever heard of. And it took scrappy marketing to get the flywheel going. 

What are some lessons you learned while building the business? 

When Minted launched in July 2008, reselling wedding invitations from other stationery brands, sales were basically nonexistent, and within weeks, I was seriously considering shutting the company down and returning the money we had left to investors. 

The only real signs of life came from the tiny assortment of unique designs that we had crowdsourced: first one order, then a few per week, and those sales were coming from the community-designed products, not the stationery brands we had launched. I had little traffic, a 0.1% conversion rate, skeptical employees, skeptical press and a sense that the company might not work. 

Then, after we pivoted to crowdsourced holiday cards in November, our sales shot through the roof and we had to shut off our marketing and turn away orders because we reached our maximum capacity. Everyone’s spouses had to come in to help process the orders. We were all sitting on the floor because we didn’t have enough time to order desks. With my newborn daughter at home, I was working on the printing line, blow-drying shrink wrap around card orders—and even crept up the stairs one night at 1 a.m. to deliver an order in San Francisco’s Pacific Heights neighborhood because the order was late. Because conversion is much higher in the holiday period, and we had a robust assortment, we were able to gain a lot more signal from a modest investment in marketing. 

What I learned is that, at that stage, we simply did not have enough data to draw big conclusions. We were reacting emotionally to a tiny sample size. The lesson for me was that in the earliest days of a company, you have to be careful not to confuse limited data with a definitive answer about whether something is a good idea or not. 

Could you tell me a story that illustrates a lesson you learned?

Minted’s crowdsourcing was my side project. I built it at night with a college student I found on Rent-A-Coder. Today, we’re a 20,000-person creative community. 

Of the $2.5 million we raised in our angel round, I spent less than $100,000 on the thing I most wanted to build. For some reason, what was calling out to me was to start a design competition that would let unknown designers submit their work, with the winners’ work sold on Minted. This was not a logical idea, and investors were understandably counseling me against it.

At the time, the playbook was simple: put established brands online and make shopping more convenient. Crowdsourcing was still a fringe idea in Silicon Valley. 

I’d tell investors the designs could be better if we went with unknown artists, but it was hard to quantify. But how do you prove “better design” to someone looking at a spreadsheet? 

But I couldn’t let it go. I was drawn to the idea that extraordinary talent was hidden everywhere, waiting for a chance to be seen. So after the Minted team went home, I built it at night with a 20-year-old college student in Oregon. 

It’s interesting that what calls to you from a gut perspective is sometimes the right move as an entrepreneur, even if you can’t prove your hunch. 

What is one mistake you made while building Minted and how did you address it? 

My biggest mistakes were usually making “safer” choices — safer product mix, safer capital planning, safer interpretation of early results, safer org design — while the real upside came from leaning harder into what was unusual about Minted. 

What were some of the first steps you took to get the business off the ground? 

I founded this business entirely on my own, so the first thing I had to do was figure out how to build a small team but conserve cash so that we could get to breakeven. This involved finding contract and part-time help, and also hiring almost entirely people who had just graduated from college. I had to think of ways to retain people, and in one case, I even gave a finance employee a ‘co-founder’ title about two years after they started at the company. 

What are some tactics you used to grow? 

I realized that there were small remnant ads at the back of women’s magazines like Domino and Cookie, which reached our target customer, found out which agency sold these, and told them that I’d be the buyer of last resort—to call me if they had last-minute, heavily discounted space available. When they called, my designer and I would write and create a print ad in about 2 hours and get it to them. 

What’s some hard, concrete advice you have for founders? 

I’d be very cautious about starting a company unless it has some type of inherent distribution, virality, or other flywheel built into the product itself. 

Also, try to shut out the distraction of caring too much about what other people think of you. I think that entrepreneurship requires you not to care about that so much. 

When it comes to this specific business, what is something you’ve found particularly challenging or surprising? 

One of the biggest surprises of entrepreneurship was how much it becomes a people, management, and leadership job, not just a product job. Even though I start companies to create great products, I’ve found myself thrust into the position of a leader and even a role model, someone people are looking to for clarity, help, and inspiration. At first, I really resisted and almost resented this part of the job—after all, I’m not a morning person, do I actually have to smile at everyone before I’ve had coffee?—but I have now made peace with this, and I am happy making a meaningful impact on someone’s career and life.

If you could go back in your business journey and change one thing about it, what would it be? 

I have underestimated how important alignment is—on values, motives, and control—before success makes those things matter enormously. If I could do one thing differently, I’d be even more intentional at the start: about who I built with, and whether everyone was truly in it for the same reasons. 

2026年10月9日 星期五

川普頒美最高科學榮譽 

GCIOSGF Newsletter 會訊:

美國白宮表示,總統川普今天將頒發美國最高科學榮譽,科技大亨馬斯克、黃仁勳和蘇姿丰等科技業領袖將獲表揚;前美國職棒大聯盟(MLB)投手克萊門斯則將獲頒「總統自由勳章」。

綜合法新社和路透社報導,Google共同創辦人布林(Sergey Brin)、輝達(Nvidia)執行長黃仁勳以及超微(AMD)執行長蘇姿丰,將在「科學:嶄新的黃金時代」(Science: A New Golden Age)高峰會上獲頒「國家科學獎章」(National Medal of Science)。

電腦業大亨戴爾(Michael Dell)及微軟(Microsoft)執行長納德拉(Satya Nadella),則將獲頒其領域最高榮譽「國家技術與創新獎章」(National Medal of Technology and Innovation)。

白宮發言人休斯頓(Liz Huston)今天聲明指出:「川普政府感謝這些傑出科學與科技領袖所做出的貢獻。這些獲獎者正協助確保美國持續引領全球創新。」

太空探索科技公司(SpaceX)及特斯拉(Tesla)負責人馬斯克(Elon Musk)去年因率領政府效率部(Department of Government Efficiency)推動削減政府支出而與川普交惡後,如今已重獲川普信任。

上個月,馬斯克與黃仁勳都出席了川普為中國國家主席習近平舉行的國宴。包括兩人在內的多名科技業領袖上週前往白宮參加會談。在會中,這些科技業高層承諾將針對人工智慧(AI)可能帶來的威脅,建立自我監管機制。

PS:馬斯克重返川普政府 共同領導五角大廈未來戰爭研究

川普駁斥AI可能摧毀人類的警告,並堅稱AI對他第二任期所稱的美國經濟「黃金時代」至關重要。

此外,川普今天表示,他將向前美國職棒大聯盟MLB投手克萊門斯(Roger Clemens)頒發「總統自由勳章」(Presidential Medal of Freedom),這是美國總統所能頒發的最高級別平民榮譽。

克萊門斯是川普的友人,兩人曾一起打高爾夫球。克萊門斯被認為是MLB歷史上最偉大的投手之一,但由於外界懷疑他曾使用提升運動表現的禁藥,棒球記者並未將他選入棒球名人堂。

克萊門斯否認使用禁藥,並曾因被指控向國會撒謊而接受審判,最終獲判無罪。

川普質疑克萊門斯為何未能獲得棒球界的最高榮譽,表示將透過頒發「總統自由勳章」來彌補。


克萊門斯說,「首先,我真的非常震驚」,但表示如果要獲得這項殊榮,希望由川普親自頒發。

克萊門斯目前經營慈善基金會,但最廣為人知的仍是他棒球生涯的成就。他曾7度贏得賽揚獎(Cy Young Award),成為MLB史上獲得這項最佳投手獎項最多次的球員。

2026年10月8日 星期四

失智問題探討

GCIOSGF Newsletter 會訊:

轉載

編按:超高齡社會來臨,隨著平均壽命延長,你我未來罹患失智症機率愈高?一篇刊登於頂尖醫學期刊《JAMA》,追蹤 1984 年到 2024 年美國老年人失智盛行率變化的調查研究顛覆預言:美國 40 年間同齡長者失智率大降 2/3!某大學醫院影像醫學部兼任主治醫師指出,失智並非老化的必然宿命,大腦老化關鍵在於「世代生活型態」的累積。

在門診或演講的 Q&A 時,我時常會感受到許多人的焦慮:我的工作壓力很大,身處的環境汙染嚴重,加上平均壽命延長了,那我得失智的風險會不會比我的父母高很多啊?

這個問題背後,其實藏著一個大多數人從未細想過的假設——我們理所當然地認為,隨著平均壽命愈來愈長、老年人口愈來愈多,罹患失智症的人也會愈來愈多。再考慮生活環境的汙染問題,我們每個人老年後失智的風險,可能是逐漸上揚。畢竟,老化是不可逆的,大腦的退化似乎也理應如此。

然而,最近我讀到一篇發表於《JAMA》的文章,卻讓我重新思考了這個假設本身是否成立。這篇文章的作者是美國杜克大學的精算學者斯塔拉德(Eric Stallard),他花了將近十年時間反覆核對自己的資料,只因為那些數字實在顛覆了他原本的預期。「失智海嘯」的預言,建立在一個沒被檢驗過的假設上

先談談大家熟悉的悲觀敘事。戰後嬰兒潮世代即將邁入高齡,2025 至 2050 年間,美國 80 歲以上人口將倍增,95 歲以上族群更會成長至現在的 3 倍。

如果失智症在特定年齡的發生機率維持不變,那麼光是人口結構的變化,就足以讓失智人口在 25 年內增加 2 到 3 倍。這也是為什麼「失智海嘯」這個說法,長期以來被視為近乎必然的未來。

但斯塔拉德的研究,卻指出這個推算裡藏著一個關鍵的、卻很少被檢驗的前提:它假設「年齡別失智盛行率」是一個固定不變的數字,彷彿 85 歲的人罹患失智症的機率,在 1980 年代和 2020 年代是一樣的

❝事實並非如此。❞

調查:40年間,美國年齡別失智盛行率下降了三分之二

斯塔拉德團隊分析了 3 個具全國代表性的長期調查,追蹤 1984 年到 2024 年美國老年人的失智盛行率變化。結果顯示,在校正年齡之後,失智症的盛行率以每年 2.5% 至 3% 的速度下降,累積 40 年下來,降幅高達約 2/3,而且到了觀察期的尾端,這個下降趨勢並沒有減緩的跡象。

用具體的數字來說明或許更有感覺:

40 年前,在 85 到 89 歲的美國人之中,大約每 10 人就有 3 人罹患失智症;到了 2024 年,這個比例已經降到 1/10 左右。

同樣是 85 到 89 歲,1895 至 1899 年出生的世代,失智比例約三成;而 1935 至 1939 年出生的世代,到了同樣的年紀,比例只剩下一成上下。


研究調查:美國老年人失智症盛行率變化(1984 vs 2024)

比較維度 1984 年

(40年前) 2024 年 變化幅度

出生世代 1895 – 1899年 1935 – 1939年 間隔 40 年

同齡長者

85-89 歲

盛行率 約 30%

(每 10 人約 3 人) 約 10%

(每 10 人約 1 人) 相對降 66.7%

(約 2/3 )

年齡校正

年降速 基準年

(100%) 每年維持

2.5 % – 3 %

穩定下降 40 年累計降幅

約 2/3

趨勢特徵 — 觀察期尾端

未見趨緩跡象 持續穩定

下降

資料整理:健康遠見編輯部

這不是單一研究的孤例。瑞典、荷蘭、英國、法國的資料也呈現類似的下降趨勢。一項涵蓋美國與歐洲七個世代研究、近 5 萬人的整合分析發現,1988 年到 2015 年間,失智症的發生率每 10 年下降約 13%。換句話說,這是一個跨國、跨研究方法、持續了數十年的一致現象,而不是統計上的雜訊或巧合。

我覺得特別值得留意的是這個現象背後的性質,它主要是一種「世代效應」,而不是「個人變好了」。也就是說,晚出生的世代,走到相同的年紀時,罹患失智症的風險本來就比早出生的世代來得低。這意味著,我們一直以來對「老化必然帶來失智風險上升」的想像,某種程度上可能算錯了帳,算錯的不是老化本身,而是我們對「風險固定不變」這個前提的信任。為什麼失智率會下降?答案還不完整,但輪廓已經浮現

老實說,研究者目前還無法完全解釋這個下降趨勢背後的機制,但幾條線索已經逐漸清晰。教育普及程度的提高、吸菸率的下降、他汀類藥物廣泛用於控制膽固醇、憂鬱症治療的進步、高血壓控制的改善,都被認為是可能的貢獻因素。這些改變大多發生在人生的早、中期。換句話說,很可能是幾十年前的生活條件與醫療環境,悄悄決定了一個人晚年大腦的命運。

這裡也呼應了《刺胳針》失智症委員會近年整理出的 14 項可改變風險因子,涵蓋教育、聽力、視力、心血管健康、社交孤立等面向,估計約可解釋全球近半數的失智症成因。我在先前的文章裡也談過這個部分,這裡就不再重複細節,但值得強調的是:這些世代之間的差異,某種程度上正是這些可改變因子長期累積的具體證明。它不是理論,而是已經寫進數十年真實數據裡的結果。不過,我認為誠實地看待這個趨勢,也必須承認它並非放諸四海皆準。部分研究顯示,日本以及美國非裔族群的失智盛行率並未同步下降,甚至持平或略微上升。肥胖、糖尿病等因子的盛行率上升,也可能在某些族群中抵銷掉部分的進步。這提醒我們,這股「退潮」的趨勢目前主要出現在特定國家與族群,而背後的機轉、以及它是否能持續,都還需要更多元、更具代表性的資料來驗證。

這對我們意味著什麼?

我常在診間提醒病人一句話:基因或許影響起跑點,但生活方式會大大改變終點。這次的世代數據,其實從另一個角度印證了同樣的道理,只是尺度拉得更長、更宏觀,不是某一個人透過幾年的努力改變了自己的大腦,而是整個世代、透過幾十年公共衛生與生活條件的累積改善,共同把風險的基準線往下推。這帶給我們一個務實但重要的訊息:失智症的風險並不是一個我們只能被動承受的宿命數字,而是一個會隨著我們如何對待自己的身體與大腦持續變動的量。

研究者也指出一個看似細微、實則深具意義的算術:失智症在 70 歲之後大約每 5 年風險就會翻倍,但反過來說,只要能把平均發病的時間點往後延遲 5 年,總體病例數就可能減少約一半。換句話說,我們不見得需要「消滅」失智症,光是「延後」它,就足以帶來巨大的改變。

也因此,在愛腦中心,我們始終認為腦齡評估與生活方式介入的意義,不在於承諾某種奇蹟式的逆轉,而在於幫助每個人及早看清楚自己當下的大腦狀態,把這場原本以幾十年為單位、由世代默默完成的改變,提前放進個人可以掌握的時間表裡。

結語:不是盲目樂觀,而是重新看待「老化」這件事

我想這份研究帶給我最大的啟發,並不是要我們就此放下戒心,以為失智症已經是過去式的問題。事實上,隨著老年人口持續增加,失智症的總病例數在可預見的未來仍會上升,只是增幅遠低於過去的悲觀推算。

真正值得我們重新思考的,是那個藏在悲觀預言底下、長期被視為理所當然的假設:老化帶來的風險,是固定不變的。40 年的世代數據告訴我們,這個假設並不成立。

大腦老化的軌跡,其實一直在被人類集體的生活條件、醫療進步與公共衛生政策,緩慢卻確實地改寫。這不是一個關於奇蹟的故事,而是一個關於「累積」的故事,這正是我們每一個人此刻仍握有主導權的部分。

2026年10月7日 星期三

摩爾定律(Moore's Law)

GCIOSGF Newsletter 會訊:

Edit: AI 

 摩爾定律(Moore's Law)是由英特爾(Intel)共同創辦人戈登·摩爾(Gordon Moore)於 1965提出的觀察,指積體電路上的電晶體數量大約每 18 至 24 個月會倍增一次,同時運算效能增加、成本則下降。 


經典摩爾定律的核心
  • 起源與修正:摩爾在 1965 年提出此預測,隨後在 1975 年修正為每約兩年翻倍一次。
  • 本質:它從來不是一條物理定律,而是半導體產業用來設定研發與生產目標的經濟與工程共識。
  • 深遠影響:推動了個人電腦、智慧型手機、網際網路與各類數位裝置的快速普及與小型化。 

走向「後摩爾定律時代」與轉變
隨著電晶體尺寸不斷縮小、逼近原子物理極限(例如漏電與量子穿隧效應),傳統單純縮小製程來提升效能的方式已面臨瓶頸: 
  • 物理極限與緩慢放緩:業界公認傳統物理微縮正走到尾聲。輝達(NVIDIA)執行長黃仁勳等產業領袖也曾表示傳統摩爾定律已近尾聲。 
  • 進入「超越摩爾定律」(Beyond Moore / 超摩爾時代):
    • 先進封裝與 3D IC:透過 CoWoS、3D 晶片堆疊技術延續效能。
    • 系統與架構創新:結合專用加速器(如 AI GPU)、高速互連技術(如 NVLink)、液冷系統來突破單一晶片算力限制。
    • 新材料與新物理:探索碳基、二維材料或量子運算等替代方案。 

Moore's Law, proposed in 1965 by Intel co-founder Gordon Moore, states that the number of transistors in an integrated circuit roughly doubles every 18 to 24 months, while computing performance increases and costs decrease.

The core of the classic Moore's Law:

• Origin and Revisions: Moore made this prediction in 1965, subsequently revising it to approximately doubling every two years in 1975.

• Essence: It has never been a physical law, but rather an economic and engineering consensus used by the semiconductor industry to set research and development and production goals.

• Profound Impact: It has driven the rapid proliferation and miniaturization of personal computers, smartphones, the internet, and various digital devices.

Towards the "Post-Moore's Law Era" and Transformation

As transistor sizes continue to shrink, approaching the limits of atomic physics (e.g., leakage current and quantum tunneling), the traditional approach of simply shrinking manufacturing processes to improve performance is facing bottlenecks:

• Physical Limits and Slowing Down: It is widely acknowledged that traditional physical miniaturization is nearing its end. Industry leaders such as NVIDIA CEO Jensen Huang have also stated that traditional Moore's Law is nearing its end.

• Entering the "Beyond Moore Era":

• Advanced Packaging and 3D ICs: Sustaining performance through CoWoS and 3D chip stacking technologies.

• System and Architectural Innovation: Combining dedicated accelerators (such as AI GPUs), high-speed interconnect technologies (such as NVLink), and liquid cooling systems to overcome the limitations of single-chip computing power.

• New Materials and New Physics: Exploring alternatives such as carbon-based materials, two-dimensional materials, or quantum computing.


DEFINITION 

Moore’s Law predicts that computing power will increase significantly as the number of transistors on microchips doubles approximately every two years, driving technological advancement.

KEY TAKEAWAYS

Moore’s Law, originally an observation by Gordon E. Moore in 1965, asserts that the number of transistors on a microchip doubles approximately every two years, leading to more powerful and cost-effective computing power.

This phenomenon has significantly influenced technological progress, enabling innovations in computing, mobile devices, electronic games, and even industries like healthcare and transportation.

As transistors approach atomic dimensions, the physical limitations of further miniaturization pose challenges, potentially signaling Moore’s Law’s natural end in the 2020s.

Despite potential limits, advancements in software, cloud computing, and emerging technologies like quantum computing may continue to drive progress beyond the traditional scope of Moore’s Law.

Get personalized, AI-powered answers built on 27+ years of trusted expertise.


What are semiconductor manufacturing's challenges overcoming physical limits?

How does the semiconductor industry plan future product releases?

Why is moore's law still a guiding maxim?

What are semiconductor manufacturing's challenges overcoming physical limits?

How does the semiconductor industry plan future product releases?

Exploring the Impact of Moore’s Law on Technology

Moore’s Law, first posited by Gordon E. Moore in 1965, observes that the number of transistors on microchips doubles roughly every two years while costs decrease. Not a fundamental law of science, this observation has nonetheless been a guiding principle in the semiconductor industry for nearly six decades. It predicts that as transistors become smaller, computing technology will continually advance, becoming faster, more energy-efficient, and more cost-effective over time. Today, Moore’s insight remains a critical benchmark for technological progress, influencing everything from smartphones to data centers.

Understanding Moore’s Law

In 1965, Intel co-founder Gordon E. Moore noticed that the number of transistors on a chip had doubled between 1960 and 1965 at minimal cost. He initially predicted chips would have 65,000 components by 1975, later revising this to a doubling every two years.

Gordon Moore did not call his observation “Moore’s Law,” nor did he set out to create a “law.” Moore made that statement based on noticing emerging trends in chip manufacturing at Fairchild Semiconductor. Eventually, Moore’s insight became a well-known adage, “Moore’s Law.” In an interview in 1975, he claimed his friend, Dr. Carver Mead from CalTech, was responsible for the name.

Intel. "Excerpts from 'A Conversation with Gordon Moore: Moore’s Law.'" Page 5.

Over the decades, Moore’s Law guided the semiconductor industry in planning and setting R&D targets, becoming a key driver of technological and economic growth in the late 20th and early 21st centuries.

IMPORTANT

Moore’s Law implies that computers, machines that run on computers, and computing power all become smaller, faster, and cheaper with time as processes become more efficient and components smaller and faster.

The Enduring Influence of Moore’s Law

Almost 60 years later, we still feel the lasting impact and benefits of Moore’s Law in many ways.

Moore’s Law in Computing: Shrinking Transistors and Expanding Power

As transistors shrink, computers become faster and smaller. These tiny structures are printed on sheets of carbon and silicon, allowing more transistors in a space, which boosts computer efficiency. Over time, the cost of powerful computers has dropped annually due to lower labor and semiconductor costs.

Advancements in Electronics: A Testament to Moore’s Law

Practically every facet of a high-tech society benefits from Moore’s Law in action. Mobile devices, such as smartphones and computer tablets, would not work without tiny processors; neither would video games, spreadsheets, accurate weather forecasts, and global positioning systems (GPS).

Diverse Industries Transformed by Moore’s Law

Moreover, smaller and faster computers improve transportation, health care, education, and energy production—to name but a few of the industries that have progressed because of the increased power of computer chips.

The Future of Moore’s Law: Challenges and Predictions

Some experts predict Moore’s Law will hit physical limits in the 2020s. Chip-makers face rising costs to maintain standards and challenges in cooling many components in small spaces.

 For instance, if you keep shrinking components, you can put more in a one-inch square chip. The more you put in that square inch, the hotter it gets and the harder it is to cool it.

In a 2005 interview, Moore himself admitted that “...the fact that materials are made of atoms is the fundamental limitation and it’s not that far away...We’re pushing up against some fairly fundamental limits, so one of these days we’re going to have to stop making things smaller.”

Overcoming the Challenges Limiting Moore’s Law

The fact that Moore’s Law may be approaching its natural end is perhaps most painfully present at the chip manufacturers themselves; as these companies are saddled with the task of building ever-more-powerful chips against the reality of physical limitations. Even Intel is competing with itself and its industry to create what ultimately may not be possible.

In 2012, Intel introduced a 22-nanometer processor, featuring the world’s smallest and most advanced mass-produced transistors.

 By 2014, they launched a 14nm chip and, after delays, in 2024, started using a large machine to develop technology advancing Moore’s Law.

FAST FACT

For perspective, one nanometer is one billionth of a meter, smaller than the wavelength of visible light. The diameter of an atom ranges from about 0.1 to 0.5 nanometers.

Future Innovations Beyond Moore’s Law

The prospect of a highly connected future presents both challenges and benefits. For over 50 years, shrinking transistors have enhanced computing, but new ways are needed to boost capabilities. Future improvements may rely on software, cloud computing, wireless tech, IoT, and quantum physics.

Despite the growing concerns around privacy and security, the advantages of ever-smarter computing technology can help keep us healthier, safer, and more productive in the long run.

What Is Moore’s Law?

In 1965, Gordon Moore posited that roughly every two years, the number of transistors on microchips will double. Commonly referred to as Moore’s Law, this phenomenon suggests that computational progress will become significantly faster, smaller, and more efficient over time.

 Widely regarded as one of the hallmark theories of the 21st century, Moore’s Law carries significant implications for the future of technological progress—along with its possible limitations.

How Has Moore’s Law Impacted Computing?

Moore’s Law has directly influenced the progress of computing power by creating a goal for chip makers to achieve. In 1965, Moore predicted that there would be 65,000 transistors per chip by 1975.

 In 2025, chip makers can put 50 billion transistors on a chip the size of a fingernail.

Is Moore’s Law Coming to an End?

According to some, Moore’s Law will end sometime in the 2020s.

 If components continue to shrink, physical limits will be reached during this decade because it’s unlikely that transistors smaller than atoms can be printed. There is only 1.5nm of space left to print on, depending on the element.

The Bottom Line

Moore’s Law began as an observation made by Gordon Moore in 1965 that the number of components on a microchip appeared to increase by a factor of two every year. He predicted that it was possible that by 1975, there would be 65,000 components on an integrated circuit. In 1975, he revised his observation and predicted that the number of components would double every two years. This prediction remained fairly accurate for nearly 50 years—and in 2025, engineers and scientists are still attempting to keep up; they have succeeded in printing transistors almost the size of atoms.

2026年10月4日 星期日

咖啡有助健康

GCIOSGF Newsletter 會訊:

別再亂喝咖啡!5萬人研究揭密「這一種類」更健康:有助健康老化、延緩衰退

一杯咖啡不只是提神飲品,選擇不同的沖泡方式,可能也會影響身體老化的速度。最新研究發現,相較於即溶咖啡,經常飲用「過濾咖啡」的人,可能擁有較健康的生物老化指標,顯示咖啡種類與健康老化之間可能存在關聯。不過研究團隊也提醒,目前研究屬於觀察性研究,只能發現相關性,並不能證明喝某一種咖啡一定能延緩老

近5萬人研究:過濾咖啡與較慢生物老化相關

這項發表於《npj Science of Food》的研究,分析英國生物銀行(UK Biobank)近4.9萬名成年人的資料,平均年齡約56歲,研究人員比較不同咖啡沖泡方式與生物老化指標之間的關係。

研究評估三項與老化相關的指標,包括:

  • 白血球端粒長度(rLTL):反映細胞老化程度的重要指標

  • PhenoAge生物年齡加速:透過血液健康數據估算身體老化速度

  • Klemera-Doubal生物年齡(KDM-BA):另一種評估生理老化程度的方法

研究結果發現,習慣飲用過濾咖啡者,其生物年齡加速程度較低,部分老化相關指標也較佳。

相較之下,大量飲用即溶咖啡者,則呈現較快生物老化的相關趨勢,且在年長者、男性及吸菸族群中更明顯。

為什麼過濾咖啡可能比較健康?

研究人員推測,關鍵可能與咖啡中的天然油脂成分有關。咖啡豆含有兩種天然化合物——咖啡醇(cafestol)與咖啡豆醇(kahweol),屬於二萜類化合物。這些物質具有抗氧化等特性,但攝取過多時,可能提高血液中的低密度脂蛋白膽固醇(LDL,俗稱壞膽固醇)。

使用紙濾方式沖泡咖啡時,濾紙可以去除大部分油性成分,使進入咖啡中的咖啡醇與咖啡豆醇含量降低,因此可能對心血管健康較有利;相較之下,法式濾壓壺、煮咖啡或土耳其咖啡等未經紙過濾的方式,可能保留較多這類成分。

喝咖啡適量最好,別忽略整體生活習慣

過去研究也發現,適量飲用咖啡可能與較低的心血管疾病風險及死亡風險相關,但咖啡對健康的影響,仍受到飲食、運動、睡眠、抽菸及飲酒等因素影響。

專家提醒,咖啡本身並非是抗老神器,想維持健康老化,仍需搭配均衡飲食、規律運動與良好生活作息。

此外,有些族群飲用咖啡需特別注意,例如:容易心悸或心律不整者、胃食道逆流患者、睡眠品質不佳者,以及對咖啡因敏感者,都應依自身狀況調整飲用量。

根據這項研究顯示,咖啡的健康影響可能不只是攝取量不同,沖泡方式也可能扮演角色,相較於未過濾咖啡或大量即溶咖啡,紙濾咖啡可能更符合健康老化的方向。不過,維持年輕健康的關鍵,仍在於長期累積的生活習慣,而不是單靠每天一杯咖啡。

Stop Drinking Coffee Randomly! A Study of 50,000 People Reveals "This Type" is Healthier: Promotes Healthy Aging and Slows Down Decline

A cup of coffee is more than just a pick-me-up; choosing different brewing methods may also affect the rate of aging. A recent study found that compared to instant coffee, people who regularly drink "filtered coffee" may have healthier bio-aging indicators, suggesting a possible link between coffee type and healthy aging. However, the research team also cautioned that this is currently an observational study, which only reveals a correlation and cannot prove that drinking a particular type of coffee will definitely slow down aging.

Nearly 50,000 People Studyed: Filtered Coffee Associated with Slower Bio-Aging

This study, published in *npj Science of Food*, analyzed data from nearly 49,000 adults at the UK Biobank, with an average age of approximately 56. Researchers compared the relationship between different coffee brewing methods and bio-aging indicators.

The study assessed three aging-related indicators, including:

Telomeric length (rLTL): an important indicator reflecting the degree of cellular aging;

PhenoAge: an estimate of the rate of aging through blood health data;

Klemera-Doubal biological age (KDM-BA): another method for assessing the degree of physiological aging.

The results showed that those who habitually drink filtered coffee had a lower rate of accelerated biological aging and better performance on some aging-related indicators.

In contrast, those who drink large amounts of instant coffee showed a faster trend towards biological aging, which was more pronounced in older adults, men, and smokers.

Why might filtered coffee be healthier?

Researchers speculate that the key may be related to the natural oil components in coffee. Coffee beans contain two natural compounds—cafestol and kahweol—which belong to the diterpenoid class of compounds. These substances have antioxidant properties, but excessive intake may increase low-density lipoprotein cholesterol (LDL, commonly known as bad cholesterol) in the blood.

When brewing coffee using a paper filter, the filter paper removes most of the oily components, reducing the levels of cafestol and bean alcohol in the coffee, which may be more beneficial for cardiovascular health. In contrast, methods like French press, brewed coffee, or Turkish coffee, which do not use paper filters, may retain more of these components.

Moderation is key when drinking coffee; don't neglect your overall lifestyle.

Past studies have also found that moderate coffee consumption may be associated with a lower risk of cardiovascular disease and mortality. However, the health effects of coffee are still influenced by factors such as diet, exercise, sleep, smoking, and alcohol consumption.

Experts remind us that coffee itself is not an anti-aging miracle. To maintain healthy aging, it is still necessary to combine it with a balanced diet, regular exercise, and a healthy lifestyle.

Furthermore, certain groups need to be particularly careful when drinking coffee, such as those prone to palpitations or arrhythmia, those with acid reflux, those with poor sleep quality, and those sensitive to caffeine. They should adjust their consumption according to their individual circumstances.

This study suggests that the health effects of coffee may not solely stem from differences in consumption volume; brewing methods may also play a role. Compared to unfiltered coffee or large quantities of instant coffee, paper-filtered coffee may be more aligned with healthy aging. However, the key to maintaining youthful health lies in long-term, accumulated lifestyle habits, not simply in relying on a daily cup of coffee.

2026年10月3日 星期六

英國的稅務難民潮

 GCIOSGF. Newsletter 會訊:

轉載

當黃仁勳說「納稅是榮譽與特權」,英國富豪圈正爆發 37 年來最大「稅務難民潮」

當英國政府為了填補財政缺口而緊縮稅政時,為何英國的超級富豪與頂級納稅人們非但沒有展現出「共體時艱」的胸懷,反而紛紛選擇「用腳投票」、改當他國(稅務)公民?
近期科技與財經界最震撼的言論之一,莫過於輝達(NVIDIA)創辦人暨執行長黃仁勳在面對美國加州可能通過的「億萬富翁稅」時,大方表態若未來 5 年必須額外繳納高達 80 億美元(約合新台幣 2,500 多億元)的鉅額稅款,他不但毫無怨言,更直言:「能夠負擔得起 5 年內繳納 80 億美元的稅款,這本身就是一種特權與責任,沒有什麼比能繳更多稅更讓我開心的了。」 他甚至幽默地笑稱:「我不怕繳稅,我只怕變窮。」這番「以繳稅為榮、視回饋社會為特權」的宏觀格局,顯然與大西洋另一端正在上演的英國富豪「大規模逃亡潮」,形成強烈而諷刺的對比。

當英國政府為了填補財政缺口而緊縮稅政時,英國的超級富豪與頂級納稅人非但沒有展現出「共體時艱」的胸懷,反而紛紛選擇「用腳投票」、移民改當他國(稅務)公民。這兩種不同心態的交鋒,也是當今全球資本主義與社會責任轉型中,最引人深思的縮影之一。

英國億萬富豪人數「創 37 年新低」的真相

近期,英國名人圈與財經界可謂熱鬧非凡。

我們一方面看到來自英國的娛樂與體壇巨星,再度成為全球焦點,並賺進了大筆鈔票:例如搖滾樂團 OASIS 的靈魂人物 Gallagher 兄弟在解散 15 年後宣佈復合重返舞台,演唱會門票和各種音樂版權收入,馬上將他們推進《星期日泰晤士報富豪榜》(Sunday Times Rich List)的名單內;又如貝克漢夫婦(David Beckham 與 Victoria Beckham)也因世界盃期間的大量曝光與廣告合約,踏進年度十億富豪(Billionaire)俱樂部。

然而另一方面,在這些光鮮亮麗的頭條背後,英國正悄然經歷一場更深層、更具結構性衝擊的宏觀經濟現象:加速中的財富外流(Wealth Exodus)與英國億萬富豪出走潮。

根據《星期日泰晤士報富豪榜》顯示,英國億萬富翁人數正面臨 37 年來最劇烈的下滑──但原因絕對不是因為英國的鉅富們變少或變窮了,而是因為他們已經「不再是英國人」了。

這股超高淨值人士與企業家的出走潮,更逐漸從紙上隱憂演變為實際的經濟警訊。例如這波出走潮中最受矚目的人物之一,當屬避險基金億萬富翁、同時也是英國排名第三大納稅人的克里斯.洛科斯(Chris Rokos)驚傳將移居希臘雅典。

作為 Rokos Capital Management 的創辦人,洛科斯的身價保守估計有 30 億英鎊,而在最新的財稅年度中,他一個人就繳納了高達 3.3 億英鎊(約新台幣 139.2 億元)的驚人稅款,相當於個人帳面總資產的十分之一。

這樣的高稅額,加上鄰近的希臘近年推出俗稱「黃金簽證」、極具吸引力的非本國定居者(Non-Dom)稅務優惠,自然對洛科斯形成難以抗拒的吸引力。

根據希臘「黃金簽證」相關規定:新稅務居民只要在當地進行一定額度的投資(如 50 萬歐元房地產或企業投資),即可享有每年僅需繳納 10 萬歐元固定稅額(Flat Tax)的優惠,且海外所得全數適用,效期長達 15 年。

3.3 億英鎊 vs. 60 萬歐元(50 萬投資額加 10 萬稅額)的差距,相信除了極少數如黃仁勳般「把納稅視為榮譽」的富豪之外,大多數人應該都很難不心動。但這個政策其實也對希臘本國產生很多負面影響,不過那就是另外一個故事了。

步行能讓大腦年輕至少4歲

GCIOSGF  Newsletter 會訊:

By Korin Miller 

Edit : OF 

賓州大學研究證實:步行能讓大腦年輕至少4歲!還能促進骨密度、心臟健康By Gigi Chang, 

這對我們的大腦認知健康有立竿見影的好處!多年來,醫學界一直強調,保持活躍有助於讓你更健康、更快樂。但最新的研究顯示,即使是規律地走路或做家務這些日常活動,也對讓我們的身體和大腦保持年輕有幫助。

這是賓州州立大學一項新研究的主要發現,該研究針對日常身體活動對健康的影響進行探討。但走路如何減緩老化?又該如何在日常生活中增加步數?以下是你需要知道的重點。這項研究發現了什麼?

這項研究發表於2024年10月的《行為醫學年鑑》(Annals of Behavioral Medicine)中,研究人員讓204名參與者每天在手機上回答一些問題。在為期九天的研究期間,受試者每天會接受檢測六次,大約每3.5小時一次。在每次報告中,參與者會回報研究人員他們在上次報告後是否有進行身體活動,並將運動分為輕度、中度或高強度。(步行是 「輕度 」活動,而健行則屬於 「劇烈 」活動)。接著,受試者會進行兩種腦力遊戲,以測試他們的認知處理速度和工作記憶。研究人員發現,如果人們在過去的3.5小時內有進行任何身體活動(即使只是輕度的),他們的大腦處理速度表現與年輕4歲的人相當。

步行能延緩認知退化嗎?

走路究竟是否能減緩大腦衰退還很難下定論,因為這只是其中一項研究。不過,這項研究結果顯示,任何類型的運動——包括步行,都可能定期為大腦帶來益處,進而有助於保持年輕。研究人員在報告中指出:「我們的研究結果表明,參與任何強度的日常身體活動可能對認知健康有短期、即時的益處,這也為未來的干預策略提供了新的潛在方向。」

賓州州立大學及賓州州立醫學院的神經學與心理學助理教授Jonathan Hakun博士表示:「像走路這樣的日常活動,可能會對我們的認知健康有直接影響。」

步行對健康有哪些好處?步行能帶來一連串的健康益處。走路不僅是心肺耐力運動,同時也是負重運動,有助於促進骨密度和身體活動力。最重要的是這項運動對大多數人都很友善。

研究也發現,走路不只可以改善情緒,還有助於維持心臟健康,並降低罹患糖尿病的風險。甚至有研究指出,走路還可能讓你變得更長壽!如何在日常生活中增加步行時間?

想在日常中增加步行量,最簡單的方法就是有目的性的去增加自己的每日行走步數。當然你不需要一次就設定一個很大的目標。只要在有機會時多動一點,像是走樓梯取代電梯、步行去朋友家而不是開車或是騎車等都是很好的做法。你甚至可以選擇在早上或晚上散步,為自己增加一些步數。

最後也最重要的是,盡力而為就好。畢竟就如同這項研究所指出的—任何形式的活動對你的大腦都有好處!

A University of Pennsylvania study confirms that walking can make your brain at least 4 years younger! It can also promote bone density and heart health. 

By Gigi Chang, original author: Korin Miller, published: 2026/10/01

This has immediate benefits for our cognitive health! For years, the medical community has emphasized that staying active helps you be healthier and happier. But the latest research shows that even everyday activities like walking or doing housework can help keep our bodies and brains young.

This is the main finding of a new study from Penn State University that explores the impact of daily physical activity on health. But how does walking slow down aging? And how can you increase your step count in daily life? Here are the key points you need to know. What did this study find?

Published in the October 2024 issue of the *Annals of Behavioral Medicine*, the study involved 204 participants answering questions on their phones daily. During the nine-day study, participants were tested six times a day, approximately every 3.5 hours. In each report, participants informed researchers whether they had engaged in physical activity since their last report, categorizing the activity as light, moderate, or vigorous. (Walking was considered "light," while hiking was "vigorous"). Next, participants took two mental games to test their cognitive processing speed and working memory. Researchers found that if people had engaged in any physical activity (even light) within the past 3.5 hours, their brain processing speed was comparable to that of someone four years younger.

Can Walking Slow Cognitive Decline?

Whether walking can slow brain decline is still inconclusive, as this is only one study. However, the results suggest that any type of exercise—including walking—may regularly provide benefits to the brain, potentially helping to maintain youthfulness. "Our findings suggest that engaging in daily physical activity of any intensity may have short-term, immediate benefits for cognitive health, providing new potential directions for future intervention strategies," researchers stated in their report.

Dr. Jonathan Hakun, Assistant Professor of Neurology and Psychology at Penn State University and Penn State College of Medicine, said, "Daily activities like walking may have a direct impact on our cognitive health."

What are the health benefits of walking? Walking offers a range of health benefits. It's not only a cardio-endurance exercise but also a weight-bearing exercise, helping to promote bone density and physical activity. Most importantly, it's a friendly exercise for most people.

Studies have also found that walking can improve mood, help maintain heart health, and reduce the risk of diabetes. Some studies even suggest that walking may help you live longer! How to increase walking time in daily life?

The simplest way to increase your daily walking volume is to purposefully increase your daily step count. Of course, you don't need to set a huge goal all at once. Just move around more whenever you have the opportunity, such as taking the stairs instead of the elevator, walking to a friend's house instead of driving or cycling, etc. These are all good practices. You can even choose to take a walk in the morning or evening to increase your step count.

Last and most importantly, just do your best. After all, as this study points out—any form of activity is good for your brain!

Enterprises cultivate talent quickly

GCIOSGF Newsletter 會訊:

Translated by AI 

Adopted by VGP

Palantir  CEO Alex Karp launched a program called the "Meritocracy Fellowship."

They directly provide approximately $5,400 per month in stipends to teenagers not yet in college, along with benefits and resource support—a high-salary training program!

The acceptance rate is extremely low.

The previous program attracted over 700 top students, but only fewer than 20 were accepted—a lower acceptance rate than some prestigious universities!

One student even tore up his Brown University acceptance letter and resolutely enrolled at Palantir.

These 18-year-olds don't attend expensive university classes. During the day, they code and work as interns on real-world military and enterprise-level AI projects; in the evenings, between two algorithm sprint sessions, they gather to study Plato, ancient Greek philosophy, and political science.

Alex Karp calls this training system the "Palantir Diploma."

Palantir executive Jordan Hirsch was remarkably candid with the media: "In the past, seeing an Ivy League diploma meant you knew this person was definitely capable. But now, that piece of paper guarantees nothing except that they're burdened with hundreds of thousands of dollars in tuition debt."

Does this seem to be a direct challenge to the US student loan system?

This is a civilizational reconstruction in progress.

Why do tech giants prefer to build their own talent from scratch rather than hire graduates from prestigious universities?

The reason is simple: the university system, this "academic-industrial complex," has become completely out of touch with the times.

For decades, corporations entrusted universities with their most sacred mission: to select and cultivate the brightest and most creative minds for society. Today, however, this system has been betrayed.

What's happening at top universities today?

Prestigious universities are rife with severe "grade inflation"—as long as assignments are submitted on time, high scores abound; curricula are outdated, failing to keep pace with the rapid evolution of the AI ​​era, still teaching frameworks from five years ago; admissions standards are squeezed out by various opaque mechanisms, while true core competencies are marginalized.

Paul Graham saw this clearly: Academic meccas, meant to cultivate free thinkers, are now mass-producing "standard parts"—arrogant individuals incapable of solving even practical engineering problems.

For tech companies, recruiting graduates from top universities is like opening a blind box:

The outer packaging is incredibly glamorous, adorned with the emblem of a century-old prestigious university; inside is a "highly educated giant baby"—someone who paid $300,000 in tuition, wrote hundreds of pages of academic papers, yet can't even configure a basic API interface.

Faced with such an outdated system that has betrayed its mission, entrepreneurs have made choices most in line with their own DNA. Rereading Ayn Rand's *The Fountainhead* will help you understand what's happening in Silicon Valley today.

When builders face a corrupt and rigid old system, they don't waste time trying to persuade or reform it. They build a new building right next to the old one.

Just as Musk didn't sit there waiting for NASA to wake up, but instead founded SpaceX;

He didn't try to persuade traditional media, but instead bought Twitter. Today, tech giants no longer try to lecture universities; they're taking over where universities have abandoned their mission.

To better understand how Palantir's "corporate university" works, consider this analogy:

Traditional university computer science departments teach how to derive data structures on a blackboard, spending two weeks writing a simple bubble sort algorithm. Palantir's "high school diploma program," however, throws 18-year-olds into the real battlefield.

For example, a company might receive an urgent project for a multinational supply chain or defense AI, requiring it to extract billions of fragmented data points from dozens of outdated databases and perform real-time analysis and predictions using AI models.

On their first day at the company, these 18-year-old high school graduates don't attend introductory lectures; they're immediately assigned the role of "Forward Deployed Engineer." They face the wrath of real customers, chaotic low-level code, and systems that could crash at any moment. In practice, they learned how to organize chaotic data within 48 hours and how to quickly locate bugs when AI models exhibit hallucinations.

Even more remarkable were their philosophy classes. Many people don't understand: programming is just coding, why read Plato?

Plato: Why read me?

Because in the AI ​​era, the technical barriers to programming are being rapidly blurred by large models. AI can generate hundreds of lines of Python code in seconds, but it doesn't know "why write this code" or "what ethical and political consequences this code will have."

Alex Karp, himself a PhD in philosophy from the University of Frankfurt, understands all too well: the top talents of the future will not be mere coders, but "modern generalists" who can code, think deeply, solve real-world problems, and question the ethics of technology.

In fact, Palantir is not the only company bypassing universities.

Look at the tech industry over the past few years: giants like Google, Apple, IBM, and Accenture have quietly eliminated the mandatory "four-year university degree" requirement for a large number of tech jobs. Ten years ago, Peter Thiel launched the "Thill Scholarship," giving $100,000 to dropouts to start businesses, which was ridiculed online as arrogant heresy. Today, directly recruiting high school graduates and establishing "corporate universities" has transformed from heresy into standard talent strategies for tech giants.

Data from polling agencies shows that in the United States, public trust in traditional institutions and media has plummeted to a historic low of 28%. People have finally realized that an expensive university degree is no longer a guarantee of upward social mobility, but rather more like a "lifetime education rent" paid annually.

Everywhere you hear pessimistic pronouncements that "traditional schools are doomed." But in my view, the truth is quite the opposite.

It's not education itself that's dying; it's the bureaucratic system that hijacked education and profited from selling diplomas.

The true schools of learning—the intellectual exchange of the ancient Athenian academies, the craftsmanship passed down in Renaissance master workshops, the practical guilds of the Middle Ages—are being reborn in Silicon Valley offices, carried by tech giants.

This new type of "corporate school" has no exorbitant tuition debt, no rigid dogmatism, only the most rigorous real-world elimination and the freest release of talent.

Universities have held the ultimate power of talent selection for centuries, but entrepreneurs have shattered that facade in less than a decade.

It's time for universities to wake up and sell "degree futures" based on their century-old reputations. When the brightest young people realize that it's better to work for top tech companies and gain real experience than to spend hundreds of thousands listening to professors read PowerPoint presentations, the walls of academia crumble.

This great reconstruction has only just begun. The next fifty years belong to those who dare to tear up diplomas and create value in the real world.

 企業培養速成人才

Palantir老闆Alex Karp推出了一個名為「精英主義獎學金」(Meritocracy Fellowship)的計畫。

他們直接給還沒入大學校門的少年每月約5400美元的津貼,加上福利和資源支持,高薪訓人!

Palantir老闆Alex Karp

錄取率很低

上一期計畫吸引了700多位學霸報名,最後只錄取了不到20名,錄取率比有些名校還低!

其中有一位,直接撕掉布朗大學的錄取通知書,毅然決然跑去Palantir報到。

這群18歲的年輕人不上高昂的大學課,白天直接在真實的軍工與企業級AI項目中敲代碼、做實習工程師;到了晚上,在兩次算法衝刺的間隙,他們圍坐在一起研讀柏拉圖、古希臘哲學和政治學著作。

Alex Karp管這套培養體系稱為「Palantir文憑」。

Palantir的高管Jordan Hirsch面對媒體極其坦白:“在過去,只要看到一張常春藤文憑,你就知道這個人肯定OK,但現在,那張紙除了代表他背了幾十萬美元的學費債務之外,什麼也保證不了。”

這好像是在針對美國的貸款上學制度麼?

這是一場正在發生的文明重構。

為何科技巨頭們寧可自己從零養,也不願意要名校畢業生?

原因很簡單:大學這套“學歷工業複合體”,已經徹底跟時代脫節了。

過去幾十年,企業把最神聖的使命託付給大學:幫社會挑選並培養最聰明、最有創意的頭腦。然而今天,這套系統已經被背叛了。

現在的頂尖大學裡發生了什麼事?

名校充斥著嚴重的「成績通膨」──只要按時交作業,高分滿天飛;課程設定跟不上AI時代的飛速演進,還在教著五年前的陳舊框架;招生標準被各種不透明的機制擠佔,真正的硬核能力反而被邊緣化。

保羅‧葛理翰看透這點:本該培養自由思考者的學術聖地,如今卻在批量生產帶著高傲姿態、卻連實際工程問題都解決不了的「標準件」。

對科技公司來說,招募名校畢業生就像在抽盲盒:

外包裝極為華麗,印著百年名校的校徽,打開一看,裡面裝著一個花了30萬美元學費、寫了上百頁學術論文,卻連基本的API介面都配置不好的「高學歷巨嬰」。

面對這樣一個背叛了使命的舊系統,企業家們做出了最符合他們基因的選擇。重讀安·蘭德的《源泉》,你就會明白當下的矽谷在發生什麼。

當建構者面對一個腐朽僵化的舊體制時,他們絕不浪費時間去說服它、改革它。他們在舊建築的旁邊,直接蓋一棟新樓。

就像馬斯克當年沒有坐在那裡等NASA醒悟,而是創立了SpaceX;

他沒有試圖說服傳統媒體,而是直接買下了Twitter。

今天,科技巨頭們不再試圖向大學喊話,他們直接在大學放棄使命的地方,取代大學行事。

為了讓大家更容易理解Palantir這套「公司大學」是怎麼運作的,我們可以舉個形象的例子。

傳統大學的計算機系,教的是怎麼在大黑板上推導數據結構,讓你花兩週時間寫一個簡單的冒泡排序;而Palantir的“高中文憑班”,直接把18歲的學生推到了真實戰場。

例如,公司接到一個跨國供應鏈或防務AI的緊急項目,需要把分散在幾十個老舊資料庫裡的幾百億條亂七八糟的資料抓取出來,用AI模型進行即時分析和預測。

18歲的高中生進公司第一天,不是去聽緒論課,而是直接擔任「前線部署工程師」(Forward Deployed Engineer)。他們要面對的是真實客戶的怒火、混亂的底層程式碼和隨時可能崩潰的系統。他們在實踐中學會如何在48小時內把混亂的資料治理得井井有條,如何在AI模型出現幻覺時快速定位Bug。

更妙的是他們的哲學課。很多人不理解,程式設計就編程,讀柏拉圖幹嘛?

柏拉圖:讀我幹嘛?

因為在AI時代,寫程式的技術門檻正被大模型迅速抹平。 AI可以幾秒鐘產生幾百行Python程式碼,但AI不知道「為什麼要寫這段程式碼」、「這段程式碼在倫理和政治上會帶來什麼後果」。

Alex Karp本身就是法蘭克福大學的哲學博士,他太清楚了:未來的頂級人才,絕對不是只會敲代碼的“碼農”,而是能敲代碼、能深度思考、能解決真實問題、還能對技術倫理提出質疑的“現代通才”。

其實,繞過大學的絕不僅僅是Palantir一家。

看看過去幾年的科技圈:Google、蘋果、IBM、埃森哲等巨頭,早已悄悄取消了大量科技職缺的「四年制大學學位」硬性要求。十年前,彼得·蒂爾推出“蒂爾獎學金”,給退學的年輕人發10萬美元創業,當時全網都笑他是狂妄的異端;而到了今天,直接招高中生、自己開辦“企業大學”,已經從異端邪說變成了科技巨頭們的標準人才戰略。

民調機構的數據顯示,在美國,民眾對傳統機構和媒體的信任度已經跌至28%的歷史冰點。人們終於看清了:昂貴的大學文憑,不再是通往階級躍遷的保票,而更像是一種按年繳納的「​​終身學歷租金」。

到處都能聽到「傳統學校要完蛋了」的悲觀論調。但在我看來,事情的真相恰恰相反。

死的絕不是教育本身,死的只是那些劫持了教育、靠賣文憑割韭菜的官僚體系。

真正的學校——古希臘雅典學園裡的思想碰撞、文藝復興時期大師工坊裡的手藝傳承、中世紀師傅帶徒弟的實戰行會——正在以科技巨頭為載體,在矽谷的辦公室裡浴火重生。

這種新型的“企業學校”,沒有高昂的學費債務,沒有僵化的教條主義,只有最嚴苛的實戰淘汰和最自由的天賦釋放。

大學擁有幾個世紀的時間來壟斷人才的終審權,但創業家只花了不到十年,就把這塊牌坊拆得稀碎。

那些還在靠著百年名校光環賣「學歷期貨」的高校,是時候清醒了。當最聰明的一群年輕人發現,與其花幾十萬去大學聽老教授念PPT,不如直接去頂尖科技公司拿高薪打真軍,大學的城牆就已經轟然倒塌。

這場偉大的重建才剛開始。未來的五十年,屬於那些敢於撕掉文憑、在真實世界裡製造價值的人。

GCIOSGF

Strategy is important

GCIOSGF Newsletter: Reprinted Minted grew rapidly after founder Mariam Naficy changed her strategy. She Raised $2.5 Million From Friends and...