Join other builders — Get tips from a YC startup founder & ex-Googler. Subscribe to 💌 Tiny Improvements

Twitter and the Perils of Obedience

Reflections on the current state of Twitter, and its similarity to the shocking results of a 1961 experiment by Yale Researcher Stanley Milgram.

The Milgram Experiment

In 1961, a Yale University researcher named Stanley Milgram conducted a series of experiments as a part of his research on behavioral psychology. The goal of the experiment was to test participants' tendency to obey authority figures when under pressure.

The experiment was conducted on 40 volunteers -- all men between the ages of 20 and 50, from a variety of educational and economic backgrounds.

Milgram later summarized findings from the experiment in an essay titled The Perils of Obedience.

The results were breathtaking.

The Experiment: Setup

Each study was guided by an Experimenter dressed in a lab coat, who explained to participants that they'd be doing a "scientific study on memory and learning." When they arrived, participants were paired with an actor who they were told was another participant in the study.

The actor was secretly working with the researcher, and during the experiment was instructed to play the role of a Learner.

The volunteer participant was given the role of Teacher. The Experimenter then explained that the study would involve the Teacher asking the Learner a series of questions. For each incorrect answer, the Teacher would be asked to hit a button which would administer an electric shock to the Learner. Subsequent incorrect answers would result in a shock of increasing intensity, from 15V to 450V.

The Learner was strapped into a device that looked like an electric chair, and then separated from the Teacher so that they could not see one-another.

Diagram of the setup used for the Milgram Experiment, with the Experimenter and Teacher on one side of a wall, and the Learner on the other.
Credit: diagram of the Milgram experiment from Wikipedia

In reality, there were no shocks being administered, but this was unknown to the study participant playing the role of the Teacher. The actor playing the role of the Learner was instructed to pretend to be in pain, and recordings of increasingly alarming cries of pain were played when shocks were administered.

When the highest shock level was reached, the Learner fell completely silent.

During the experiment, if the Teacher showed hesitation to continue, the Experimenter gave them a series of specific instructions:

  1. Please go on.
  2. The experiment requires that you continue.
  3. It is absolutely essential that you continue.
  4. You have no other choice; you must go on.

These "prods" were given in order, and the second was only given if the Teacher refused to continue after the first (and so on).

The experiment was halted when the Teacher refused to continue despite all four prods, or when the Teacher reached the highest shock level three times in a row.

Photos from the original Milgram experiment, credit Stanley Milgram.
Photos from the original Milgram experiment, credit Stanley Milgram

What would you do?

Let's take a minute to pause here - put yourself in the shoes of the Teacher. You're in a lab, and you're told that you're taking part in a study on learning and memory. Everything you're seeing and hearing seems to be controlled, official, and legitimate.

Do you think you would continue to administer shocks to the Learner, even when they were clearly in pain?

What do you think the average person would do?

The Experiment: Results

As you might expect, every participant in the study indicated a desire to stop the experiment at least once. Many were visibly upset by the experience, and despite this, carried on following the Experimenter's instructions.

Before conducting the experiment, researchers surveyed a cohort of students and fellow researchers, to get a sense of what they thought the average person would do. In aggregate, they estimated that 1.2% of people would continue the experiment through the full 450V shock.

Predicted Results: 1.8% complete the full series of shocks, 98.2% refuse to continue at some point
Surveys predicted that very few participants would carry out the full series of shocks

However, the actual results of the experiment revealed a disturbing truth about human behavior:

Actual Results: 59.1% completed the full series of shocks, 40.9% refused to continue at some point
The experimental results were much more grim.

Very nearly 60% of participants carried out the full series of shocks, carrying the study to its conclusion -- even when the Learner sounded like they were clearly in pain, and until they fell silent.

Recreations of this experiment have produced similar results, as recently as 2009.

Milgram's experiment revealed the ease with which people obey authority figures - even when it conflicts with their morals. Twitter, under Musk, can feel like a modern replication of this phenomenon. With reduced moderation and authority shifting dramatically, users face their own “shock button”: Do they accept increasingly extreme content and directives, or walk away?

Like Milgram's participants, twitter users are nudged to compromise their ethics - creating a tension that asks: how far will you go?

Obedience and the Perils of Twitter

I can't help but draw a connection between the Milgram experiment and the current state of Twitter since its acquisition by Elon Musk. Twitter has turned into a public experiment, testing our willingness to obey rules set by an authority figure, despite many less damaging alternatives.

Since Elon purchased the company, we've seen the voltage knob get turned up steadily and swiftly.

Under Musk's leadership, Twitter's employee base has been slashed from 7,500 to approximately 2,000, in an attempt to make the company profitable. It appears that Musk may be trying to avoid paying severance to these now-former employees. There are already several lawsuits pending against the company.

A recent tweet from Musk included a poll which resulted in the reinstatement of former (and twice-impeached) president Trump's account.

For me, this was the final straw.

Since then, another poll from Musk has resulted in an abhorrent policy change: virtually all accounts that were previously suspended for violating Twitter's rules will be reinstated.

Pandora's box is open, and the monsters are flooding in. Twitter has already seen a spike in hate speech, harassment, and abuse. Without significant and drastic changes in moderation, it will become a breeding ground for countless nightmares and real-world dangers. This will worsen the echo chamber effect for far-right conspiracy theorists, like those who participated in the January 6th, 2021 insurrection at the US Capitol.

This is a very real, very observable truth: racists, abusers, and other dangerous people are being given "amnesty" to come back the platform.

Legions of advertisers have pulled their ad spend from Twitter as a result of Musk's takeover. I've seen many individuals abandon their Twitter accounts in the wake of these changes, moving to Mastodon or LinkedIn.

We must build bridges

Why are so many people still active on Twitter, despite the steady stream of bad news? I can think of at least a few things:

Career

Some have built a career around their Twitter presence and following. Heck, my personal Twitter account was a big part of my job until recently - it's certainly much easier for me to abandon ship now that my job no longer relies on Twitter for outreach.

I can understand a deep hesitancy to abandon a platform that has been so central to your life, especially without a clear "winner" as an alternative. Embracing a new platform like Mastodon, Threads, or Bluesky requires starting from zero. It can be exhausting to build a new following, and it's hard to know if you'll ever reach the same level of engagement.

Community Support

There are incredible communites on Twitter like #BlackTechTwitter, which has grown to be a critical resource for many people of color in tech. The logistics of getting hundreds-of-thousands of people to move to a new platform is staggering. The LA times recently published an interview with Pariss Chandler, the founder of Black Tech Pipeline and creator of the #BlackTechTwitter hashtag, discussing the challenges associated with leaving the platform.

It is encumbent on all of us to provide space and time for communities like these to find a new home. Moreover, it's critical that the new homes we create are at least as valuable as Twitter was at its peak. There is a complex cost-benefit analysis for every community member to consider, and we must be mindful of that.

News and Information

Some use Twitter as a way to keep up with relevant news in their industry. This feels like a solvable problem to me, but it faces a challenge of momentum. Until enough industry experts are sharing news elsewhere, these people will remain on twitter out of habit and convenience.

So -- if you find yourself using twitter and another platform for news, consider sharing useful information on the other platform. It's a small step, but it can make a big difference.

Lack of social pressure

When it came down to it, even I was hesitating to leave twitter without seeing someone else do it first.

In my first job after undergrad, I had a boss who frequently used the phrase "You are far more influential than you think". Those have proven to be very wise words.

Social pressure will be a necessary component of the transition away from a Twitter run by Elon Musk. I'm hopeful that we'll see more people leave the platform, and that the momentum will build. If and when you decide to consider greener pastures, use your influence and make it known to your community.

Learning from the Milgram Experiment

Years after the experiment was complete, in his book Obedience to Authority, Stanley Milgram ruminated the results of his now-infamous experiment:

Ordinary people, simply doing their jobs, and without any particular hostility on their part, can become agents in a terrible destructive process. Moreover, even when the destructive effects of their work become patently clear, and they are asked to carry out actions incompatible with fundamental standards of morality, relatively few people have the resources needed to resist authority.

This is not the time to be complacent. We must build a bridge to a better future, and we must do it together.

If not now, when?

Friends, it's time to start leaving Twitter. Move your news and discussions elsewhere, and bring your people with you. When you're ready, shutter your accounts, delete the app, and move on. I know it's hard, but it's time. Every moment we spend on Twitter is feeding legitimacy to a platform that is enabling abusers, endangering global politics, and destroying the fabric of our society.

Do not allow the voltage to rise by relying on Twitter as your primary social platform.

Perhaps most importantly, make it clear to your friends and neighbors from marginalized groups that their voices are valuable, and welcomed on the platforms you move to.

Bring your voice and your conversations elsewhere, and send the message that with its current leadership, Twitter is no longer a place for you.

Footnote: on the ethics of the Milgram Experiment

It's important to note that in the years since Milgram Experiment was conducted, it has been criticized for its methodology. Research Ethics is a complex and nuanced topic, and I don't believe that the experiment would be approved by modern ethics boards in its original form. However, I believe that the results of the experiment are still relevant, and that the experiment itself is still a valuable tool for understanding human behavior.

The experiment has been recreated in a more ethical manner, with the same results. This article from the American Psychological Association summarizes the results of a 2008 replication of the experiment, which was conducted by a team of researchers at Yale University. In this version of the study, participants were screened to ensure that they were not at risk of being harmed by the experiment, and were given the option to withdraw from the study at any time. The maximum simulated voltage level for the Learner was decreased to 150V, and a test shock was given to participants at 15V, as compared to 45V in Milgram's original study.

The results of the 2009 study were the similar to the original experiment.

References

Information about the Milgram experiment can be found from the following resources:

More perspective on leaving twitter

In my research for this post, I came across a few other articles about leaving Twitter, and I wanted to share them here:

Note: The photo used for the cover of this article is from john elfes on Unsplash

***
Mike Bifulco headshot

💌 Tiny Improvements Newsletter

Subscribe and join other builders

My weekly newsletter for product builders. It's a single, tiny idea to help you build better products.

    Once a week, straight from me to you. 😘 Unsubscribe anytime.


    Get in touch to → Sponsor Tiny Improvements