‘National effort required to prevent the knowledge inside Indian libraries from disintegrating’

Carl Malamud spoke to The Hindu about Gyan Satyagraha and his thoughts on AI’s potential to democratise knowledge

After his 20-minute long talk at the Underline Center in Bengaluru, Carl Malamud, technologist and public domain advocate who took on the Indian government to make the building codes accessible, gets addressed as the ‘original internet pirate.’ 

“I’m not a pirate,” he corrects, and explains why. 

In the years 2012 and 2013, Malamud bought over 18,000 Indian Standards from the Bureau of Indian Standards spending more than $10,000 and published them online. And then, he wrote a letter to the Bureau. 

“You will be delighted to know that more people have access to your standards,” the letter read.  “They were not delighted,” Malamud laughs. 

A legal battle ensued. A few years later, BIS published most standards on its website for free. The petition in court was withdrawn. Nowhere during this process, Malamud chose to hide behind anonymity, unlike a pirate.  

“I’ve learned a lot from not only Gandhi but also from Martin Luther King and Nelson Mandela. There is a way to question authority and to make change,” says the self-proclaimed Gandhian.

He points out how Gandhi himself wrote a letter in 1930 to Lord Irwin, then Viceroy of India, informing him of his plans to begin the Dandi March and defy the Salt Laws. Malamud, a staunch believer in nonviolence, believes it’s important to have a story and explain why you are doing a certain thing. He’d like to consider himself a civil servant. 

“Knowledge needs to be available to everybody,” he says. This pretty much forms the basis of all works of Malamud who is the founder of Public.Resource.Org and co-founder of Servants of Knowledge Association.  

A champion of open knowledge access and proponent of what he calls ‘the Gandhian Internet’, Malamud believes what we have today is an ‘East India Company Internet’ where entities such as Meta or Elsevier gatekeep flow of information. 

“We have a lot of gateways out there, people that are intermediating. You’ve become the product. They’re monetizing you. And it’s very much the case in the access to scientific knowledge. How can you stand on the shoulders of giants if you can’t read?” he wonders. 

Pointing out that access to knowledge is key for the functioning of any democracy Malamud has been calling for a Gyan Satyagraha. “Because I believe we need to fight for open knowledge and open access!” he says. 

The Hindu caught up with Malamud to understand more about Gyan Satyagraha, his views on AI and about his latest works. 

What would a Gyan Satyagraha entail?

Satyagrahas are both offensive and defensive.  

It is about protesting and making an impact, but it is also about going to the villages and working with people on a daily basis.  

In our Gyan Satyagraha, on a daily basis, we scan books and make them available, we talk to people about how to use the internet properly... We spend an awful lot of time on such things. When we adopt a library, we not only work with them to scan their books but help fix their library, add some software, fix their internet...   

Bread labour is the idea that you need to do something every day. I think open-source programmers are a really good example of that. They may be in business or may work for companies such as Google, but they come home, and do something, either as part of their job or on the side, for the broader community. There are a whole bunch of ways in which you can give back. If you have the tech skills, you could adopt your local NGO and help fix their computers.  

In simple words, you have to keep working to do Gyan Satyagraha.  

Now what I don’t like is the idea of hackathons. You go in for two hours, meet up, have some pizza, put in a few hours and then you have a hack. That’s not a sustained effort.  

Hackathons are good for bringing new people in and getting them to understand that there’s a whole world out there. But from there you can only hope that they go on to pick something that’s important.  

You have to put in the work every day. I think that’s what Gandhi meant by bread labour.

AI seems to be the buzzword lately. What are your thoughts on AI’s potential to democratise knowledge? 

I am not a fan of the kind of AI which is about ‘ask me anything, I’ll give you an essay.’ I don’t think that’s right.  

I use AI for optical music recognition. We scan sheet music, transform it into “Music XML” and then create a number of downstream things, such as WAV files, MIDI files, and also Braille Music.  

We have an MoU with IIT Madras in the AI4Bharat programme. What we’re going to try and do is high-quality translations of the Indian Standards into Indian languages. That is fairly hard because Indian Standards are fairly technical. So, we’re going to work with the IIT Madras team as well as enlist the help of bilingual engineering students to help us figure out what works and what doesn’t, use that to re-transform the standards, and then keep iterating until we have something that everybody agrees is good. This has great potential.  

AI has great potential in the area of scientific information. There’s a big problem with people forging experiments in their scientific articles. Pattern recognition is a real good way to spot fakes. There are humans who do that now. But AI can assist in doing things.  

I think, in AI, focused applications are the future. I think this whole singularity thing of ‘ask me anything, I’ll write you an essay’ has huge bad effects. I think many AI companies are systematically violating copyright.  

So, I’m a big fan of being able to use materials for training the models, but not so much of this business of spinning it back out. I think there are tremendous issues that need to be resolved, and governments need to step up.  

I think AI as a buzzword is the wrong road. Pattern recognition, better translation, better character recognition...I think that’s the potential of using the technology. 

After your long battle, the building codes or most of them are in the public domain in India. What’s next? 

Getting the standards available for everyone is step one, really important. The next step would be building that awareness in civil society, among unions and so on and getting people to understand what these safety standards are.  

Another one is that in India there’s a huge amount of knowledge lying in libraries. Much of it is disintegrating. There needs to be a national effort to stop that from happening. It can’t be a top-down thing coming from a single point like the big government. It’s got to happen all over. That’s what we’re doing at Servants of Knowledge. We’re trying to teach everyone how to scan, build scanners and operate them.  

We have taken some open-source library management software. It’s very difficult to use. We’re turning it into something that any moderately skilled IT person can work with. We’re trying to educate librarians. It’s again raising awareness.  

Anything specific that you’re working on in Karnataka? 

We have an MoU with Kannada University to digitise their PhD dissertations and old and rare books. We’ve been working with the Lalbagh Library for digitisation and have been in talks with other botanical and horticultural organizations too. We hope to go on from Lalbagh to make accessible a broader amount of information useful for farmers and researchers. We’re working with Azim Premji University to scan their library. We have also been scanning materials of the Sochara Community Health Organization to make them more broadly available. In Karnataka, we’re deep inside. 

Published - December 17, 2024 09:00 am IST