Why Open Source Is Shaping the Future of Software: From AI to Blockchain

01

The split occurs because of differing beliefs. Some see it as knowledge, others as product, leading to the division into closed‑source software and open‑source software.

02

Commercial software tends to chase eye‑catching new features and concepts. Open‑source software values code simplicity and quality, reflecting deep internal kernel development.

03

Support for commercial software is driven by product defects and cost; support for open‑source software aims to give users better service and generates commercial profit.

04

Originality and legal risk of commercial software are borne by the vendor, whereas open‑source software’s copyright and patent risks are passed directly to users.

05

Early open source, exemplified by Linux, mainly imitated commercial software. Today, open‑source projects like Hyperledger blockchain and TensorFlow AI are not just copycats but can lead technological trends.

06

It is not only software that defines the world, but open‑source software that defines the world.

07

The Industrial Revolution eliminated most handicraft groups but gave rise to programmers, the largest remaining manual‑skill community.

08

Programmers are using past craftsmanship to hand‑craft a promising new future.

09

Science is essentially an open‑source endeavor. The scientific method requires reproducibility, which demands sharing hypotheses, test environments, and results. Shared scientific knowledge fuels invention, and open‑source code drives innovation. Open‑source software extends the scientific method to IT; in computer science, reproducibility and proof of program correctness are achieved by “sharing code”.

10

Microsoft is the giant of the closed‑source world. Previously people said “open source is a cancer”; now they say “open source is the future”.

11

There are hundreds of open‑source licenses (GPL, BSD, MIT, Mozilla, Apache, LGPL, etc.), yet no widely adopted “open data license”. The lack of such a license seriously hampers the data‑open movement.

12

Previously development, testing, and operations were separate; now, with open source, social networks, and mobile internet, DevOps integrates development and operations.

13

Commercial software’s advantage is abundant funding; open‑source software’s advantage is knowledge sharing.

14

China’s open‑source force is rising quickly. Three years ago we localized international projects; three years later we internationalize local projects.

15

Software production evolved from the 1950‑60s individual‑hero era, through the 1970‑2000 corporate division of labor, to a post‑2010 open‑source‑based social collaboration model.

16

Open source underpins software’s social division of labor. Software defines industry, yet the software industry itself must be produced industrially, with open source as the foundation.

17

Enterprises earning revenue by servicing open‑source software resemble lawyers: both offer professional consulting and training, the former built on open software, the latter on legal transparency.

18

In the open‑source era, innovation should start with re‑packaging, but not be limited to it.

19

In 20 years, languages will still be split between human mother tongues and machine foreign languages; the latter is computer machine code, while Chinese and English are dialects of the human mother tongue.

20

Programming language standards fall into natural‑language and machine‑language categories. Traditional standards are based on human language; open source is based on machine language.

21

In cloud computing, we must rethink the open‑source spirit. Initially it was for practitioners to experiment and improve while giving back, before any company offered infrastructure as a service or rebranded open‑source projects for profit with little contribution back.

22

Traditional machine learning required code changes; modern deep learning uses “parameter programming”, adjusting data instead of code.

Open source is a bright strategy; open is always better than closed.