When for-profit companies fund research, how is science affected?

Case Study: The Release of AlphaFold 3 by Google DeepMind (May 2024)

Introduction to AlphaFold 3

1. AlphaFold 3 is a powerful AI tool released by Google DeepMind in May 2024, designed to predict the shapes of proteins.
2. This new version is based on previous versions of AlphaFold and AlphaFold 2, both of which were made open-source, meaning the code was available for free for scientists to use and improve.
3. AlphaFold 3, however, was different.
4. The full code wasn't shared with the public.
5. Some parts, like the protein-drug interactions simulator, were kept private.

Why Was Information Withheld?

1. Isomorphic Labs, a company spun off from DeepMind, is using AlphaFold 3 to help develop new drugs, which is why DeepMind chose to keep some parts of the code private.
2. Head of AI science at DeepMind, explained that they wanted to balance sharing the tool with making sure Isomorphic Labs could still make money from its drug research.

The Debate: Science vs. Intellectual Property (IP)

1. The main issue here is the conflict between making scientific work open and protecting it for profit.
2. For-profit companies, like those behind AlphaFold 3, often use patents and intellectual property (IP) laws to protect their ideas and inventions.
3. However, science traditionally works best when it is open, allowing other researchers to check, build upon, and improve the work.

Key Views on the Issue

1. A professor at the University of Toronto, argues that keeping scientific discoveries secret goes against the goal of advancing science.
2. He believes that making research available to everyone benefits society as a whole.
3. Some scientists, like Haibe-Kains, believe in sharing the basic software or algorithms but keeping a commercial version for sale or exclusive use.
4. This lets researchers contribute to public knowledge while also making money from a more advanced version.
5. Pressure to Commercialise: Many universities and research institutions rely on commercialisation to fund their research.
6. This creates pressure for scientists to patent their discoveries and potentially keep parts of their work private.

How Can Scientists Balance Openness and Commercialisation?

1. One possible approach is to publish the basic version of the code and algorithms openly, but keep the more advanced version, which is ready to use, for commercial use.
2. This allows scientists to contribute to the public but still make money from their work.
3. Thomas Hemmerling’s Approach: Hemmerling, a professor in anesthesiology, worked on a robot that administers anesthesia automatically.
4. He published the basic algorithms but patented specific parts of the technology.
5. This allowed others to use his methods while protecting his intellectual property.

The Role of Government Funding

1. Government Funding vs. Private Companies: Some argue that government funding could help solve the conflict between openness and commercialisation.
2. With government support, researchers wouldn't be tied to private companies and could publish their findings without worrying about profit.
3. Advantages of Public Funding:
   1. Researchers can focus on innovation without restrictions from companies.
   2. It reduces dependence on corporate funding, which often comes with strings attached, like limiting what can be published or researched.
4. Even with more public funding, universities and research institutions often still want to commercialise their discoveries to earn revenue.
5. This desire to profit remains strong, even if government funding is available.

Ethical Concerns and Long-Term Implications

1. Many scientists believe that withholding key details, like in the case of AlphaFold 3, undermines the integrity of science.
2. If other scientists can't verify or replicate the work, it can slow down progress.
3. Commercialisation and Ethics: While it's important for researchers to make money from their discoveries, some argue that keeping too much secret harms the wider scientific community.
4. Other scientists should be able to see the data, understand the methods, and build on the work.

Real-World Example: McSleepy

1. Thomas Hemmerling and his team developed a robot called McSleepy, which could give anesthesia on its own.
2. They published the algorithms used to create the robot, allowing other researchers to understand and build upon their work.
3. This example shows that commercialisation and openness can coexist.
4. While some parts of the technology were patented, the basic research was made available for others to use.

Conclusion: Is AlphaFold 3 a Good Example?

After facing criticism, the authors of the AlphaFold 3 paper agreed to release the full code within six months, and they made it available earlier in November 2024. Haibe-Kains believes that while releasing the code later is better than not releasing it at all, science should be open from the start. Waiting to release the full details undermines the research process.

Must Check: Best IAS Coaching In Delhi

UPSC Prelims Result 2024 Out: Expected Cut Off & Other Details, UPSC Prelims 2024 Answer with Explanation, Daily Prelims Quiz, Daily Current Affairs, MONTHLY CURRENT AFFAIRS TOTAL (CAT) MAGAZINE, Best IAS Coaching Institute in Karol Bagh, Best IAS Coaching Institute in Delhi, Daily Mains Question Answer Practice, ENSURE IAS UPSC Toppers, UPSC Toppers Marksheet, Previous Year Interview Questions, UPSC Syllabus