New Delhi: A graduate of IIT-Delhi and UC Berkeley is the unlikeliest of activists against Artificial Intelligence.
But that’s exactly what 24-year-old Samuel Shadrach from Bengaluru is doing. He is on a hunger strike against the creation of superintelligent AI.
He is not alone. He is part of an internet movement. Others include Guido Reichstadter in San Francisco, former AI researcher Michael Trazzi in London, and Denys Sheremet, a student from Amsterdam. The three camped outside AI company offices and streamed their fasts, while Shadrach kept them company on YouTube.
The protests are directed at major AI labs—including OpenAI, Google DeepMind, and Anthropic—which have openly declared their goal of building Artificial General Intelligence (AGI) and, eventually, superintelligence.
Shadrach shares livestream videos of his hunger strike on his YouTube channel, hoping to attract more attention and inspire others to join him in the fight to safeguard humanity’s future. Meanwhile, Reichstadter, Trazzi, and Sheremet have begun receiving considerable attention for their posts on X.
While some users acknowledge the growing risks posed by superintelligent AI, others mock the protests or dismiss their concerns.
“Lol, keep going, remember to hold your breath and stomp your feet like a boomer protesting the internet,” wrote one X user.
this attire is more appropriate. pic.twitter.com/UY1z56fnJg
— korben (@korbencopy) September 8, 2025
Shadrach graduated from IIT Delhi in 2023 and completed the ML Safety Scholars Program under Dan Hendrycks at UC Berkeley. He is currently on the 13th day of his hunger strike, surviving only on water, electrolytes, and magnesium supplements, he claims.
“I have publicly declared that I am opposed to anyone trying to create superintelligence—AI that is better than all humans at doing their tasks,” he told ThePrint.
Also Read: UP’s new dog rule faces flood of ‘paw and order’ jokes
Shadrach insists that the race to build such AI could lead to catastrophic consequences. He voiced concern about a scenario in which a superintelligent AI is developed in a data centre and then escapes human control—his fears sound like the plotline of a Black Mirror episode or a Hollywood end-of-the-world horror film.
“You’ve built an AI that can pursue goals over long time horizons, that can make plans, and execute them. So there is a risk that we could just lose control of the superintelligence, and it could build its own society—which could be as bad as causing human extinction,” said Shadrach.
He also flagged another risk: that those who control superintelligent AI could become the most powerful people in history. He explained that the creators would gain immense political influence, be extremely persuasive, run their own military, and invent technologies the rest of the world could not match. As a result, the world could end up being ruled by a very small number of people.
“The entire world could end up being ruled by just a handful of AI CEOs,” Shadrach added.
So far, the comments he has received on his YouTube channel are more about concerns for his health than about an AI takeover.
“This isn’t working, it would be irrational to continue. You need to stop and reassess the situation. Find another way,” said a comment on his ‘Day 12’ video.
Also Read: Bengaluru to Palghar, potholes are a death trap. AI startups racing to solve it
It’s been a chain of protests, each one sparking the next. Activist Guido Reichstadter is currently on Day 23 of his hunger strike outside Anthropic’s office in San Francisco. He inspired Michaël Trazzi, who began his own strike outside DeepMind’s London headquarters on 5 September. Soon after, Denys Sheremet travelled from Amsterdam to join Trazzi, ending his water-and-electrolyte fast after 16 days on 23 September.
Shadrach is in contact with the others involved in these protests. He has even interviewed Reichstadter on his YouTube channel as part of his effort to build a global movement around the issue.
He is calling for legally binding global treaties—particularly between the United States and China—to halt the development of superintelligent AI. He emphasises that governments must take responsibility and stop their labs from pursuing superintelligence.
On his latest livestream, he wrote in the caption that he was willing to strike physically as well: “I am still open to doing this, if someone can solve visa and funding issues for me. $20k will fix it.”
Shadrach has committed to fasting for at least 18 days, though he hasn’t set a specific end date or condition for ending the strike. He was still ploughing through on Friday, posting on X: “Nutrient deficiencies are probably wrecking me.” A few hours later, he recounted a dream featuring Elon Musk and added: “Reality is disappointing.”
(Edited by Asavari Sing)