The Pavlok: A Shockingly Bad Idea

By Duff McDuffee on October 23rd, 2014 1

Wearable health technology is all the rage these days. Silicon Valley just can’t get enough little gadgets to track our movements, weight and bodyfat percentage, blood pressure, and other vital statistics.

Granted, some of these gadgets are pretty cool, and maybe even useful. And yet most current wearables like the Jawbone and Fitbit lines of products are little more than glorified pedometers.

Enter Pavlok.

Pavlok is advertised as the first wearable gadget that shocks you if you don’t do what you committed to doing.

This latest wearable is the product of energetic, young Maneesh Sethi whose website advertises “cheat codes for life.” Sethi is brother of Ramit Sethi, Ramit being the creator of the popular financial advice blog “I Will Teach You To Be Rich.”

Maneesh and I had a vigorous debate on Twitter about his product, which ended when Ramit encouraged me to write a longer article. So here goes!

You, dear reader, can be the judge of whether my arguments are convincing or not. I doubt I will convince Maneesh, for as Upton Sinclair wrote, “It is difficult to get a man to understand something, when his salary depends on his not understanding it.”

But more important than this particular product, I think the Pavlok is an iconic example of an entire paradigm of personal development that thinks self-aggression is the best way to go about personal change. Read to the end to find out why!

Shocking the Press

Pavlok has been making the rounds through tech news and personal development blogs alike, no doubt as a result of the the PR wizardry of Maneesh’s brother, Ramit. Whether it deserves such attention on the merits of the product itself and not just on the shock appeal has not been questioned at length by most of these sources. (Maneesh is, after all, a fan of Kim Kardashian’s strategy for gaining fame.)

So, how did this electrifying idea come about? As the founder states on the IndieGoGo crowdfunding page

“Hi! I’m Maneesh Sethi. A few years ago, I hired a girl [sic] to slap me in the face whenever I used Facebook. My productivity increased 4x — so I knew I was on to something. A few weeks later, I asked my friend Dan Kaminsky if we could build a shock collar that shocked me when I went on Facebook. Dan said: Let’s go to RadioShack.”

Despite calling it “shock therapy,” this device has little to do with the barbaric psychiatric practice of electroconvulsive therapy aka electroshock therapy, where a patient is shocked into a partial seizure to create a similar effect to a lobotomy. It has more to do with torturing non-human mammals in the name of science. Or in this case, torturing human mammals in the name of personal development.

Shocking You into Happiness?

While other interventions are paired with the electric shock including beeping, vibrating, and tracking steps (of course), the electric shock is clearly the emphasis in the marketing of this product. The item itself has a yellow lightning bolt logo on it!

So any criticism of this product should focus on the efficacy of electric shock specifically in behavior change, not on any of the other secondary features it may have. The electric shock feature is central to the unique value proposition and marketing positioning of this particular gadget. So either Pavlok is primarily an electric shock bracelet, or it’s a bait-and-switch (aka a con) with the shock aspect just a marketing gimmick.

But for now let’s assume the developers in good faith believe this product to actually be what they say it is: a “habit changing device that shocks you.”

So, a bracelet that shocks you when you go on Facebook, or don’t go to the gym, or whatever. What could possibly go wrong?

Lots of things.

Pavlok is not Pavlovian

First, despite being named after Ivan Pavlov, the founder of Classical Conditioning, the Pavlok wearable device is better understood as a product of the later school of behavioral psychology known as Operant Conditioning.

Classical, Pavlonian conditioning is where you pair a new stimulus with an already existing stimulus, to trigger a response, as in the iconic example of Pavlov ringing a bell when presenting a dog with food.

After enough trials, the bell itself can then trigger salivation…basically because the dog learns that it’s the dinner bell. But Pavlov and other behaviorists didn’t want to go there when it came to mental events, preferring only to study that which was observable from the outside. (The cognitive turn in the 70’s started to reverse that trend, luckily, and neo-Pavlonians also aren’t so picky anymore.)

So really Pavlok is misnamed…which already worries me about whether the founders took Psych 101 before naming their product after the wrong school of psychology.

Pavlok is a Positive Punishment Device

But I digress…in Operant Conditioning, adding electric shock to cease a behavior like going on Facebook would be called “positive punishment.” For those of you reading who haven’t taken Psych 101 in a while (*cough* Maneesh *cough*), allow me to explain.

First off, we all have habits we want to increase in ourselves (e.g. going to the gym), and habits we want to reduce or get rid of (e.g. going on Facebook).

In Operant Conditioning, positive and negative have different meanings than what we normally mean. “Positive” means a stimulus is delivered following a response, and “negative” means a stimulus is withdrawn following a response. Positive doesn’t mean “good” but “added stimulus,” and similarly negative doesn’t mean “bad” but “subtracted stimulus.”

To reinforce a behavior means we get more of it, and to punish a behavior means we get less of it. Note that if you intend to punish a behavior but you get more of it, you actually reinforced it. These terms are determined by actual behavior changes, not intent.

Combining the two sets of terms we can talk of positive and negative reinforcement, and positive and negative punishment.

I won’t bore you with the details. The key thing here is that electric shock upon performing an undesired behavior is a classic example of positive punishment. The addition of a shock makes it “positive” (adding a stimulus), and reducing the behavior makes it “punishment.”

This graphic from the IndieGoGo Page indicates behaviors you can allegedly reduce, presumably through self-administered electric shocks:

To be fair, the crowdfunding site also includes a list of behaviors that allegedly can be increased through use of the device:

Positive Punishment is not Negative Reinforcement

In some places I’ve seen Sethi and his staff claim that Pavlok works by means of “negative reinforcement,” but that claim is simply false. However, it is a common mistake made by Psych 101 students when learning about Operant Conditioning.

Again, as Sethi wrote, “Hi! I’m Maneesh Sethi. A few years ago, I hired a girl [sic] to slap me in the face whenever I used Facebook. My productivity increased 4x — so I knew I was on to something.”

That’s practically the definition of positive punishment: reducing a behavior (going on Facebook) by adding an aversive stimulus (being slapped in the face by a woman). That’s literally the origin story for this product – positive punishment to reduce an unwanted behavior.

Because this confusion seems to go so deep over at Pavlok HQ, I’ll give another example from the Indiegogo page:

What habits can I break with Pavlok?

Pavlok is designed to help to break ANY habit. This ranges from visiting distracting websites, going to fast food restaurants, not standing up from your desk often enough, going to bed late, biting your nails, and TONS of others.

Breaking habit = punishment. That’s the definition of punishment in Operant Conditioning. Punishment is reducing a behavior, reinforcement is increasing a behavior. So the method of breaking a bad habit by adding electric shock is punishment. Always. 100% of the time.

Ok, so I hope that is finally clear.

Again, this confusion about basic psychology doesn’t boost my confidence about this project.

That said, since negative reinforcement is for increasing a behavior’s frequency by removing an aversive stimulus, you could use shock as negative reinforcement, but that would be if you wanted to increase a behavior.

For example, many Beeminder goals utilize negative reinforcement – if you don’t do what you say you will, say bike 50 miles a week, you get charged money automatically. (Sadly some other popular Beeminder goals are positive punishment, e.g. quit smoking, or aren’t even directly behaviorally in one’s control, e.g. lose 30 lbs).

At best, Pavlok could be a geofence for a negative reinforcement tool like Beeminder, e.g. tracking GPS data to know whether you went to the gym for an hour on Monday…but that’s already possible with a smartphone, which Pavlok relies on in any case for geofencing. In fact, there are already app integrations for Beeminder with various other tools – no doubt there will be integrations with Pavlok.

On Pavlok’s home page, again there is confusion between positive punishment and negative reinforcement:

Choose your commitment level, and when you succeed, you’ll earn money. But, if you fail, you’ll have a penalty to your Facebook friends. Either you’ll give them your money — or they’ll get to shock you.

Here they don’t state whether the commitment is to increase a behavior (“positive”) or to reduce a behavior (“negative”). That difference matters – a lot. Because while negative reinforcement can be quite effective, positive punishment has largely been abandoned by Operant Conditioning since at least the early 1970s for some very good reasons.

Positive Punishment Doesn’t Work, and Has Bad Side-Effects

As B.F. Skinner, one of the most prominent founders of Operant Conditioning put it in 1972…

“A person who has been punished is not thereby simply less inclined to behave in a given way; at best, he learns how to avoid punishment” (B.F. Skinner, Beyond Freedom and Dignity).

B.F. Skinner practically invented Operant Conditioning, and he was totally opposed to punishment – not just because he felt it was inhumane, but because he observed that it actually doesn’t even work very well.

Honestly I feel like I should just stop there. Skinner abandoned it, end of story.

And yet, the story continues. In parenting literature and research, punishment has been increasingly recommended against. For instance, this author says punishment “results in immediate compliance in the moment, but not in the long run.” That sort of thing tends to be the standard advice now, because of extensive research into punishments like spanking.

Spanking research clearly indicates that not only does spanking not work, it causes behavioral problems including increased aggression. Mind = blown. You mean that kids who are hit by their parents go on to hit others? I wonder where they learned to do that….

Here’s what the APA president said:

“You cannot punish out these behaviors that you do not want,” says Kazdin, who served as APA president in 2008. “There is no need for corporal punishment based on the research. We are not giving up an effective technique. We are saying this is a horrible thing that does not work.”

A HORRIBLE THING THAT DOES NOT WORK. But electric shocks for going on Facebook, yea, brilliant idea guys. Almost as brilliant as a meal-replacement shake for geeks who hate eating because it interrupts their World of Warcraft. I know, we’ll call it Soylent! Food is so human body 1.0!

Seriously. WTF happened, Silicon Valley? Is it something in the water out there? Can’t you make a cure for malaria or something?

A Shock Collar for Humans

Remember when Sethi wrote “I asked my friend Dan Kaminsky if we could build a shock collar that shocked me when I went on Facebook. Dan said: Let’s go to RadioShack.”

Let’s talk about shock collars, shall we?

Shock collars are basically punishment devices for dogs, or depending on who you ask, torture devices for dogs. They can be part of an “invisible fence” or part of a training regimen. Unsuprisingly, shock collars are highly controversial.

Shock collars sometimes work to keep dogs on the property and out of harm’s way, but they are also known to increase aggression and anxiety in dogs.

“Dogs may associate the pain from the shock with the environment or with objects in the environment…this may lead to anxiety or negative associations with those objects, which can ultimately result in aggression”

But that totally couldn’t happen in humans, right?!? Right?

I mean being shocked randomly throughout the day totally wouldn’t change my state to that of being slightly on edge…especially if the device malfunctioned in some way, which never happens with technology, right guys?

With shock collars, the shock must be strong enough that it causes pain and does not habituate, but not too strong as too be torture. So that means shock collars necessarily must hurt. As the author of the previously linked blog says, “This electric current will cause pain and physical discomfort to the dog, otherwise it would not be effective in conditioning him.” So if a dog shock collar doesn’t cause harm, it doesn’t work, and if it works, it’s abuse. What a great invention, guys! How do we adapt this for humans?

If the Pavlok fails to be painful enough, the mild shock will be easy to ignore. However, according to TechCrunch, “you can modify the intensity of the electric shock from a very mild 17 volts up to 340 volts.” Indeed, what often happens with positive punishment is habituation, where more intensity is required over time to get the same aversive effect.

In fact, electric shock in classical conditioning is kinda unpredictable. In a study on rats in 1966, electric shock didn’t work for preventing rats from drinking a sweet liquid, but a substance that made them ill did (stupid researchers torturing mammals for science – grrrr). Associating electric shock with light and sound did lead to avoidance. (Source: Relation of cue to consequence in avoidance learning.)

So perhaps Pavlok’s shocking feature could help you avoid Facebook (light and sound) but not soda (sweet liquid). In fact, foul taste may be a naturally occurring “aversion therapy” specific to foods that can even associate to the wrong thing easily with one-trial learning, as when a person gets sick from the flu but thereafter cannot eat a certain food they ate that day. So even if Pavlok came with a nausea generator app, it might still lead to the wrong kind of learning.

100% Automated Law Enforcement

It is extremely likely that some users will end up getting shocked for not breaking their commitments at all, so-called false positives. The reason for this is that many real-life contexts are fuzzy. For example, take the case of red light tickets generated automatically by cameras hooked up to computers. Someone who is almost half-way through the light will get a ticket 100% of the time even though no cop would ever be this harsh. This 100% automated law enforcement creates a very different environment where practical wisdom never enters the equation at all. (Interestingly, red light cameras actually increase accidents at intersections and thus in effect positively reinforce risky driving behavior.)

As an example, imagine someone wants to spend less time on Facebook so they set their Pavlok to shock them when they go on Facebook, mirroring Sethi’s experiment to have “hired a girl [sic] to slap me in the face whenever I used Facebook.”

What about using Facebook to log into an app? What about Facebook comments on a blog? What about sharing an article to Facebook through Buffer? What about sharing a YouTube video to Facebook through YouTube’s social media sharing? How about checking the TimeHop app? Some of these may trigger a shock, or maybe none of them, or maybe all of them.

There will always be more fuzzy use cases like this than can be anticipated through any set of programmed rules. To work at all, punishment must be 100% consistent and never arbitrary. Even then, punishment only trains us to know what NOT to do, not what to do, as B.F. Skinner pointed out in 1972.

This is not theoretical. In this video Maneesh says, starting at 20 seconds, “It’s been going off a lot today because I have it set to shock me every time I go on Reddit, and I’m doing a Reddit AMA.”

So already it’s giving false positives.

Aversion Therapy

As I mentioned before, Sethi has briefly corresponded with me on Twitter and in that conversation referenced some papers from an obscure rehab facility. This facility engages in “aversion therapy” for reducing addictions to drugs and alcohol.

The history of aversion therapy is extremely controversial, mostly due to the infamous case of Aubrey Levin. Levin was a psychiatrist in South Africa and in the late 60’s/early 70’s did a range of terrifying psychological experiments on gays and lesbians involving electric shock and other painful torture to try and “cure” individuals of homosexuality. This horrific abuse is part of the legacy of electric shock for human behavioral change.

Of course Pavlok’s marketing materials do not suggest in any way that it could be used for such a horrifying purpose, but again it is telling that the founder mentions “aversion therapy” without any seeming knowledge of the sordid history of this approach. Levin’s Aversion Project is one major reason why most researchers won’t touch aversion therapy with a 10-foot electrified pole.

There’s also the 1971 Stanley Kubrick film A Clockwork Orange. In the film, violent sex offenders in a dystopian future London are subject to aversion therapy as part of their rehabilitation. The main character is forced to watch videos of violence and sexual imagery while receiving a drug that causes nausea. The therapy works so well he becomes incapable of fighting back when abused, becomes ill when seeing a topless woman, and is no longer able to listen to his favorite symphony from Beethoven because of the new negative associations.

But that’s just fiction, representing our fears of the potential dangers of aversion therapy. In reality, it doesn’t work that well. For example, the most popular drug for aversion therapy still in use is Disulfiram, better known as Antabuse. If you take Antabuse, nothing happens…that is until you take a drink of alcohol. Then you start to feel incredibly nauseous and start vomiting uncontrollably, a classic positive punishment. Hailed as a miracle drug for alcoholism for a while, Antabuse remains highly controversial.

Antabuse seems to work well in supervised studies, but not at all in unsupervised ones. In other words, when the doc is watching you swallow your pill, you won’t drink. But if the doctor is away, it’s party time!

Or as Dr. Jonathan Tallman, a GP in Minneapolis person put it,

“[Antabuse] is not proven to work,” he says. “In a 52 week trial on veterans with a placebo, it did not reduce the time to first drink or increase abstinence.” Those that wanted to drink simply stopped taking the pill, waited two weeks for the drug to clear, and drank.

Indeed, that’s the general consensus with positive punishment. It works as long as the punishment occurs every time, aka the boss man is watching. But as soon as the parent or teacher or psychiatrist isn’t looking, the punishment’s effect on behavior change goes away. We’d hope that what it would do is make the alcoholic feel nauseous just thinking about drinking, but it doesn’t seem to work like that.

Self-administered electric shocks would be easily avoidable by simply removing the device. I imagine thousands of people saying “F— this sh–!” and throwing their Pavloks across the room, freeing themselves from their shocking shackles.

Or is Pavlok a Con?

…unless…unless the whole thing is a con, and the shock is just for effect. And that’s what I believe is actually going on here.

Again, either this product is really about positive punishment through electric shock, as the origin story and logo and marketing message all claim, or it is all a bait-and-switch to get you to buy a product that most people will not end up using for that purpose but for other “secondary” features.

In the company’s own ebook, the word “shock” appears only 3 times, and none of those times is the shocking feature of Pavlok mentioned in any way indicating a method of behavioral change. In fact, the approach in the ebook focuses almost entirely on positive reinforcement:

…smaller habits can intentionally be formed in a three-step process: cue, routine, and reward.

It is very hard to destroy habits. Instead, they should replaced. And the best way to replace habits is to keep the same cue and reward, and replace the routine.”

This completely contradicts the whole marketing about ending bad habits through the positive punishment of electric shock. It contradicts the origin story of Maneesh hiring a woman to slap him, and building a shock collar with his friend. It contradicts everything.

The passage I just quoted from the ebook basically admits that the shocking aspect doesn’t even work. The ebook quotes BJ Fogg’s Tiny Habits (renaming it to microhabits for no reason at all), which Fogg says repeatedly is only for creating new good habits, not for breaking bad habits.

This short book goes on to talk about keystone habits, community support, accountability partners…but no mention at all throughout these sections of using THE MAIN FEATURE OF THE PRODUCT, the one thing that differentiates it from all other wearable health tech. The three case studies all involve creating new habits through positive and negative reinforcement, despite confusingly using the word “punishment” once or twice. Only when we get to Chapter Six we hear that one can use the shock aspect as negative reinforcement for getting out of bed on time, and only as a last resort. The central feature is extremely marginalized throughout this entire manual.

In other words, the shock aspect – which is absolutely central to this product – is just a gimmick.

It’s a confusing, stupid, and harmful gimmick, a gimmick that doesn’t work, a gimmick that has likely negative side-effects, and a gimmick that aligns with the dominant paradigm: self-aggression as personal change. Whether it’s Jillian Michaels yelling at obese people to push themselves brutally hard in the gym, or Steven Pressfield writing about part of you that you need to fight an endless inner war against (“Resistance”), or Tony Robbins beating his chest screaming “COOL MOSS!” while walking across hot coals, people love self-aggression. We love it precisely because it fits our map of the world, because it’s more of the same.

Worse still, Pavlok is self-aggression built into an artifact through intentional design, like how “anti-homeless spikes” turn classism into hostile architecture.

The fact that Maneesh is in this paradigm is unsurprising, given that his “friends” include Tim Ferriss, Dave Asprey (“the bulletproof coffee” guy), Chris Guillebeau, a “pick up artist” known as “Tynan,” and a host of other shady characters.

The thing is we already are in an inner war. We already punish ourselves when we do behaviors we don’t like by making ourselves feel bad through horrifically abusive self-talk.

Do you really think doing more of the same will work? More self-hatred, more causing ourselves pain, more self-shaming. And why stop there? Project that strategy outwards too (we always do) and motivate others with threats, verbal abuse, and even physical abuse.

The enemy only understands force.

We cannot negotiate with terrorists.

And so on and so forth…the cycle of violence continues in gross and subtle ways, internally and externally. But we can also break that cycle through methods like Core Transformation which involve practicing radical non-violence towards all parts of ourselves, or by taking courses in non-violent direct action and nonviolent communication for transforming the ways in which we communicate with others.

It’s not as shocking, but hey, it actually works so there’s that.

Other references:

Maneesh did an AMA (Ask Me Anything) on Many commenters were excited about the product, others shared similar concerns as me, especially here and here.

Hey, if you liked this post and want more stuff like this, you can subscribe to my new newsletter over at Scientific Goals as I don’t really post here much anymore. Thanks for reading!



Powered by Facebook Comments

Tags: , , , , , , , , , , , , , , , , , , , , ,

Comments are closed.