Media & Entertainment

Hands on with Telepath, the social network taking aim at abuse, fake news and, to some extent, ‘free speech’

Comment

Telepath logo
Image Credits: Telepath

There’s no doubt that modern social networks have let us down. Filled with hate speech and abuse, moderation and anti-abuse tools were an afterthought they’re now trying to cram in. Meanwhile, personalization engines deliver us only what will keep us engaged, even if it’s not the truth. Today, a number of new social networks are trying to flip the old model on its head — whether that’s attempting to use audio for more personal connections, like Clubhouse, eliminate clout chasing, like Twelv, or, in the case of new social network Telepath, by designing a platform guided by rules that focus on enforcing kindness, countering abuse, and disabling the spread of fake news.

Many of these early efforts are already facing challenges.

Private social network Clubhouse has repeatedly demonstrated that allowing free-flowing communication in the form of audio conversations is an area that’s notoriously difficult to moderate. The app, though still unavailable to the broader public, courted controversy in September when it allowed anti-Semitic content to be discussed in one of its chat rooms. In the past, it had also allowed users to harass an NYT reporter openly.

Meanwhile, Twelv, a sort of Instagram alternative, ditches the “Like” button concept and all the other features now overloading Instagram, which had once been just a photo-sharing network. But, unfortunately, this also means there’s no easy way to find and follow interesting users or trends on Twelv — you have to push friends to join the app with you or know someone’s username to look them up, otherwise it shows you no content. The result is a social network without the “social.”

Telepath, meanwhile, is a more interesting development.

It’s pursuing an even loftier goal in social networking — creating a hate speech-free platform where fake news can’t be distributed.

No social network to date has been able to accomplish what Telegraph claims it will be able to do in terms of content moderation. Its ambitions are optimistic and, as the network remains in private beta, they’re also untested at scale.

Though positioned as a different kind of social network, Telepath isn’t actually focused on developing a new sharing format that could encourage participation — the way TikTok popularized the 15-second video clip, for example, or how Snapchat turned the world onto “Stories.”

Instead, Telepath, at first glance, looks very much like just another feed to scroll through. (And given the amount of linked Twitter content in Telepath posts, it’s almost serving as a backchannel for the rival platform.)

The startup itself was founded by former Quora employees, including former Quora Business & Community head, Marc Bodnick, now Telepath Executive Chairman; and former Quora Product Lead, Richard Henry, now Telepath CEO. They’re aided by former Quora Global Writer Relations Lead, Tatiana Estévez, now Telepath Head of Community and Safety; and Ro Applewhaite, previously research staff for Pete Buttigieg for America, now Telepath Head of Outreach.

It’s backed by a couple million in seed funding, led by First Round Capital (Josh Kopelman). Other backers include Unusual Ventures (Andy Johns), Slow Ventures (Sam Lessin), and unnamed angels. Bodnick and his wife, Michelle Sandberg, also invested.

Image Credits: Telepath

When talking about Telepath, it’s clear the founders are nostalgic for the early days of the web — before all the people joined, that is. In smaller, online communities in years past, people connected and made internet friends who would become real-world friends. That’s a moment in time they hope to recapture.

“I’ve benefited a lot by meeting people through the internet, forming relationships and having conversations — that sort of thing,” says Henry. “But the internet just isn’t fun in the ways that it used to be fun.”

He suggests that the anonymity offered by networks like Reddit and Twitter make it more difficult for people to make real-world connections. Telepath, with its focus on conversations, aims to change that.

“If we facilitate a really fun, kind, and empathetic conversation environment, then lots of good things can happen. And it might be that you potentially find someone you want to work with, or you end up getting a job, or you meet new friends, or you end up meeting offline,” Henry says.

Getting Started

To get started on Telepath, you join the network with your mobile phone number and name, find and follow other users, similar to Twitter, then join interest-based communities as you would on Reddit. When you launch the app, you’re meant to browse a home feed where conversation topics from your communities and interesting replies are highlighted — orange for those replies from people you follow and gray for those that Telepath has determined are worth being elevated to the home screen.

As you read through the posts and visit the communities, you can “Thumbs Up” content you like, downvote what you don’t, reply, mute, block, and use @usernames to flag someone.

Image Credits: Telepath, screenshot via TechCrunch

Another interesting design choice: everything on Telepath disappears after 30 days. No one will get to dig through your misinformed posts from a decade ago to shame you in the present, it seems.

What’s most different about Telepath, however, is not the design or format. It’s what’s taking place behind the scenes, as detailed by Telepath’s rules.

Users who join Telepath must agree to “be kind,” which is rule number one. They must also not attack one another based on identity or harass others. They must use a real name (or their preferred name, if transgender), and not post violent content or porn. “Fake news” is banned, as determined by a publisher’s attempts at disseminating misinformation on a regular basis.

Telepath has even tried to formalize rules around how polite conversations should function online with rules like “don’t circle the drain” — meaning don’t keep trying to have the last word in a contentious debate or circumvent a locked thread; and “stay on topic,” which means don’t bombard a pro-x network with an anti-x agenda (and vice versa.)

Image Credits: Telepath

To enforce its rules, Telepath begins by requiring users to sign up with a mobile phone number, which is verified as a “real” number associated with a SIM card, and not a virtual one — like the kind you could grab through a “burner” app.

In order to the create its “kind environment,” Telepath says it will sacrifice growth and hire moderators who work in-house as long-term, trusted employees.

“All the major social networks essentially grew in an unbounded way,” explains Henry. “They had 100 million-plus active users, then were like, ‘okay, now how do we moderate this enormous thing?’,” he continues. “We’re in a lucky position because we get to moderate from day one. We get to set the norms.”

Moderation

“Day one” was a long time in the making, however. The team rebuilt the product four times over a couple of years. Now, they say they’ve developed internal tools that provide moderators with visibility into the system.

According to moderator head Estévez, these include a reporting system, real-time content streams organized in to buckets (e.g. a bucket for “only new users”), as well as various searchable ways to get context around a report or a particular problematic user.

“Really good tools — including real-time streams of content, classifiers for problematic behavior, searchable context, and making it hard for banned users to return — mean that each moderator we hire will be quite scalable. We think that there are network effects around positive behavior,” she says.

Image Credits: Telepath

“It’s our intention to scale up fast and high accuracy moderation decision-making, which means that we’re going to be investing a lot of engineering effort in getting these tools right,” she adds.

The founders have decided not to use any third-party systems to aid in moderation at this time, they told TechCrunch.

“We looked at a bunch of off-the-shelf [moderation systems], and we’re basically building everything that we need from scratch,” says Henry. “We just need more control over being able to tweak how these systems work in order to get the outcome that we want.”

The investment in human moderation over automation will also require additional capital to scale. And Telepath’s decision to not run ads means it will eventually need to consider alternative business models to sustain itself. The company, for now, is interested in subscriptions, but hasn’t made decisions on this front yet.

Banning the trolls

Though Telepath has only 4,000-plus users in its private beta, the two-person moderation team is already tasked with moderating posts from across the thousands of pieces of content shared on a daily basis. (The company doesn’t disclose how many violations it takes action against per day, on average.)

When a user breaks the rules, moderators may first warn them about the violation and may require them to take down or edit a specific post. No one is punished for making a mistake or being unaware of the rules — they’re first given a chance to fix it.

But if a user breaks the rules repeatedly or in a way that seems intentional, such as engaging in a harassment campaign around another user, they are banned entirely. Because of the phone number verification system, they also can’t easily return — unless they go out and purchase a new phone, that is.

These moderation actions don’t necessarily have to follow strict guidelines, like a “three strikes rule,” for example. Instead, the way the rules may be enforced are determined on a case-by-case basis. Where Telepath leans towards stricter enforcement is around intentional and flagrant violations, or those where there’s a pattern of bad behavior. (As with Reply Guys and sealioning behavior.)

In addition, unlike on Facebook and Twitter — platforms that sometimes seem to be caught off guard by viral trends in need of moderation — Telepath intends for nothing to go viral on its platform without having been seen by a human moderator, the company says.

Fake News

Telepath is also working to develop a reputation score for users and trust scores for publishers.

In the case of the former, the goal is help the company determine how likely the user is to break Telepath’s rules. This isn’t developed yet, but would be something used behind the scenes, not put on display for all to see.

For publishers, the trust score will be how factually correct they are what percentage of the time.

Image Credits: Thomas Faull (opens in a new window) / Getty Images

“For example, if the most popular article in terms of views from the publisher is just completely factually incorrect or intentionally misleading…that should have a bigger penalty on the trust score,” explains Henry. “The problem is that the incumbent platforms have rules against disinformation, but the problem is that they don’t enforce them out of this desire to appear balanced.”

Bodnick adds this challenge is not as insurmountable as it seems.

“Our view is that, actually, a handful of outlets are responsible for most of the disinformation…I don’t think our intent is to build out some modern-day truth system that will figure out if The Washington Post is slightly more accurate than The New York Times. I think the main goal will be to identify repeat disinformation publishers — determine that they are perpetual publishers of disinformation, and then crush their distribution,” says Bodnick.

This plan, however, involves setting rules on Telepath that fly in the face of what many today consider “free speech.” In fact, Telepath’s position is that free speech-favoring social networks are a failed system.

“The problem, in our view, is that when you take this free-speech centered approach that sort of says: ‘I don’t care how many disinformation posts Breitbart has published in the last — three years, three months, three weeks — we’re going to treat every new post as if it could be equally likely to be truthful as any other post in the system,’” says Bodnick. “That is inefficient.”

“That’s how we will scale this disinformation rule — by determining which relatively small group of publishers — I’m guessing it’s hundreds, low hundreds — are responsible for publishing lots of disinformation. And then take their distribution down,” he says.

This opinion on free speech is shared by the team.

“We’re trying to build a community, which means that we have to make certain tradeoffs,” adds Estévez. “In the rules we refer to Karl Popper’s paradox of tolerance — to maintain a tolerant society, you have to be intolerant of intolerance. We have no interest in giving a platform to certain kinds of speech,” she notes.

This is the exact opposite approach that conservative social media sites are taking, like Parler and Gab. There, the companies believe in free speech to the point that they’ve left up content posted by an alleged Russian disinformation campaign, saying that no one filed a report about the threat, and law enforcement hadn’t reached out. These MAGA-friendly social networks are also filled with conspiracies, un-fact checked reports, and, frankly, a lot of vitriol.

The expectation is that if you go on their platforms, you’re in charge of muting and blocking trolls or the content you don’t like. But by their nature, those who join these platforms will generally find themselves among like-minded users.

Twitter, meanwhile, tries to straddle the middle ground. And in doing so, has alienated a number of users who think it doesn’t go far enough in counteracting abuse. Users report harassment and threats, then wait for days for their report to be reviewed only to be told the tweet in question didn’t break Twitter’s terms.

Telepath sits on the other end of the spectrum, aggressively moderating content, blocking and banning users if needed, and punishing publications that don’t fact check or those that peddle misinformation.

“Kindness” carve-outs

And yet, despite all this extra effort, Telepath doesn’t always feature only thoughtful and kind-hearted conversations.

That’s because it has carved out an exception in its kindness rule that allows users to criticize public figures, and because it doesn’t appear to be taking action on what could be problematic, if not violating, conversations.

Image Credits: Telepath

A user’s experience in these “gray” areas may vary by community.

Telepath’s communities today focus on hobbies and interests, and can range from the innocuous — like Books or Branding or Netflix or Cooking, for example — to the potentially fraught, like Race in America. In the latter, there have been discussions about the capitalization of “Black” where it was suggested that maybe this wasn’t a useful idea. In another, sympathy is expressed for a person who was falsely pretending to be a person of color.

In a post about affordable housing, someone openly wondered if a woman who said she didn’t want to live near poor people was actually racist. Another commenter then noted that gang members can bring down property values.

A QAnon community, meanwhile, discusses the movement and its ridiculous followers from afar — which is apparently permitted — though supporting it in earnest would not be.

There are also nearly 20 groups about things that “suck,” as in GOPSucks or CNNSucks or QuibiSucks.

Anti-Trump content, meanwhile, can be found on a network called “DumbHitler.”

Meanwhile, online publishers who routinely post discredited information are banned from Telepath, but YouTube is not. So if feel you need to share a link to a video of Rudy Giuliani accusing Biden of dementia, you can do so — so long as you don’t call it the truth.

And you can post opinions about some terrible people in which you describe them as terrible, thanks to the public figure carve-out.

Cheater and deadbeat dad? Go ahead and call them a “disgusting human being.” VP Pence was referred to by a commenter as “SmugFace mcWhitey” and Ronny Jackson is described as “such a piece of sh**.”

Explains Estévez, that’s because Telepath’s “be kind” rule is not intended to protect public figures from criticism.

“It is important to note that toxicity on the internet around politics isn’t because people are using bad words, but because people are using bad faith arguments. They are spreading misinformation. They are gaslighting marginalised groups about their experiences. These are the real issues we’re addressing,” she says.

She also notes that online “civility” is often used to silence people from marginalized groups.

“We don’t want Telepath’s focus on kindness to be turned against those who criticize powerful people,” she adds.

In practice, the way this plays out on Telepath today is that it’s become a private, closed door network where users can bash Trump, his supporters and right-wing politicians in peace from Twitter trolls. And it’s a place where a majority agrees with those opinions, too.

It has, then, seemingly built the Twitter that many on the left have wanted, the way that conservative social media, like Gab and Parler, built what the right had wanted. But in the end, it’s not clear if this is the solution for the problems of modern social media or merely an escape. It also remains to be seen whether a mainstream user base will follow.

Telepath remains in a closed beta of indefinite length. You need an invite to join.

More TechCrunch

Quantum computing may still largely be in the theoretical domain, but the money that it’s attracting is very real.

UK’s Riverlane scores $75M to correct quantum errors

Indian space-tech startup EtherealX aims to test its fully reusable medium-lift vehicle in 20.

India’s EtherealX puts $5M seed toward fully reusable launch vehicles

John Schulman, one of the co-founders of OpenAI, has left the company for rival AI startup Anthropic. In addition, OpenAI president and co-founder Greg Brockman is taking an extended leave…

OpenAI co-founder Schulman leaves for Anthropic, Brockman takes extended leave

Grok — not to be confused with the homophonic AI startup Groq that this morning raised over $600 million — has been spreading false information about Vice President Kamala Harris…

Secretaries of state urge X to stop its Grok chatbot from spreading election misinformation

Saudi Arabia is committing even more money to Lucid Motors as the EV startup struggles to erase its losses. Lucid announced Monday as part of its second-quarter earnings report that…

Lucid pumps $1.5B from Saudi wealth fund after CEO warned relying on its ‘bottomless wealth’ was ‘dangerous’

Listen, I’m tired of talking about Boeing’s Starliner, too, but the spacecraft still isn’t home and questions are mounting about NASA’s transparency.

TechCrunch Space: I’m tired of talking about Starliner, too

Google will appeal a U.S. District Court judge’s opinion Monday that found the technology giant acted illegally to maintain a monopoly in online search. The decision from Judge Amit P.…

Google loses massive antitrust case over search, will appeal ruling

Last year, OpenAI held a splashy press event in San Francisco during which the company announced a bevy of new products and tools, including the ill-fated App Store-like GPT Store.…

OpenAI tempers expectations with less bombastic, GPT-5-less DevDay this fall

Muon Space closed a new tranche of funding for its space-as-a-service business.

Muon Space closes $56M to scale all-in-one satellite platform

We’re so excited to announce that we’ve added a dedicated AI Stage presented by Google Cloud to TechCrunch Disrupt 2024. It joins Fintech, SaaS and Space as the other industry-focused…

Announcing the agenda for the AI Stage at TechCrunch Disrupt 2024

A startup developing AI market research based on location data, and backed by a who’s who, has quietly raised, TechCrunch has learned.

Placer.ai boosts valuation to $1.5B after quietly raising another $75M

Safari’s newest feature, Distraction Control, can remove distracting elements from a website. The feature follows Arc Browser’s addition of Boosts last year, which similarly lets users remove features from a…

Apple’s new Safari feature removes distracting items from websites

By collecting this data, OpenAI “profited significantly” from the creators’ work, the complaint alleges.

YouTuber files class action suit over OpenAI’s scrape of creators’ transcripts

India’s fast-growing quick commerce market is getting a new deep-pocketed entrant: Walmart-owned Flipkart, India’s largest e-commerce firm. Flipkart has started to roll out Flipkart Minutes, its quick commerce service, in…

Flipkart blitzes into India’s 10-minute quick commerce battle

The list includes Elon Musk’s xAI, which is already valued at a staggering $24 billion, as well as a good number of other AI startups.

38 startups have become unicorns so far in 2024: Here’s the full list

When a company is the size of Amazon, a lot of bad actors will come after it and its customers, which makes defending the network a monster job. Over the…

AWS unveils Mithra to identify and mitigate malicious domains across its massive system

The European Commission has closed a Digital Services Act (DSA) investigation of a rewards feature in TikTok Lite by accepting commitments from the social media giant to permanently withdraw the…

TikTok Lite: EU closes addictive design case after TikTok commits to not bring back rewards mechanism

Groq, a startup developing chips to run generative AI models faster than conventional processors, said on Monday that it has raised $640 million in a new funding round led by…

AI chip startup Groq lands $640M to challenge Nvidia

COVID-19 pushed people to take up outdoor activities. Now, startups are helping companies and consumers keep up with demand.

From golf to hunting, a new crop of startups want to make these experiences even better

Despite increasing demand for AI safety and accountability, today’s tests and benchmarks may fall short, according to a new report. Generative AI models — models that can analyze and output…

Many safety evaluations for AI models have significant limitations

OpenAI has built a tool that could potentially catch students who cheat by asking ChatGPT to write their assignments — but according to The Wall Street Journal, the company is…

OpenAI says it’s taking a ‘deliberate approach’ to releasing tools that can detect writing from ChatGPT

Chief Product Officer Craig Saldanha says AI is already transforming the Yelp experience.

Yelp’s chief product officer talks AI and authenticity

Featured Article

Even after $1.6B in VC money, the lab-grown meat industry is facing ‘massive’ issues

Any goal that puts cultivated meat in big box grocery stores or on fast food menus in the 2020s is “unrealistic,” according to experts.

Even after $1.6B in VC money, the lab-grown meat industry is facing ‘massive’ issues

Warren Buffett’s Berkshire Hathaway cut its Apple holding by around half, to $84.2 billion, according to an SEC filing. While Apple remains the firm’s largest stock holding by far, Buffett…

Warren Buffett’s Berkshire Hathaway sells half its Apple stock

A fireside chat between Jensen Huang and Mark Zuckerberg at SIGGRAPH 2024 took some unexpected turns. What started as a conversation about the capabilities of Nvidia GPUs and Zuckerberg’s vision…

Zuckerberg and Jensen show off their friendship, while an AI necklace covets yours

We spoke to Harness CEO and founder Jyoti Bansal about his previous company, which Cisco bought for $3.7 billion in 2017.

When a big company comes after a hot startup, it’s not a slam dunk decision to sell

Dojo is Tesla’s custom-built supercomputer that’s designed to train its “Full Self-Driving” neural networks.

Tesla Dojo: Elon Musk’s big plan to build an AI supercomputer, explained

Featured Article

Trade My Spin is building a business around used Peloton equipment

Trade My Spin has pieced together a logistics network capable of offering same or next day delivery in most major cities in the continental U.S.

Trade My Spin is building a business around used Peloton equipment

Featured Article

Meet the founder who built and sold a $600M enterprise software startup from Sri Lanka

Sanjiva Weerawarana co-founded WSO2 in 2005, recently selling it for more than $600M. He sometimes drives for Uber, too.

Meet the founder who built and sold a $600M enterprise software startup from Sri Lanka

Investors are assisting startup founders earlier than ever in an effort to help them bridge the first climate tech valley of death.

Why Bill Gates’ Breakthrough Energy and other investors are scouring universities for founders