Connect with us

Trending Spy News

IRL’s ex-CEO says he’s ‘shocked’ about the company’s fake users


IRL’s company logo.
IRL’s company logo. | Photo illustration by William Joel/The Verge

I have a packed issue this week, starting with the implosion of a once-hot social media startup called IRL. It raised $200 million at over a $1 billion valuation, and now its board is dissolving the company after finding that most of its users were fake. The SEC is investigating. It’s a whole mess. I have ex-CEO Abraham Shafi’s first public comments on the situation since he was ousted by the board.

Also in this issue: a convo with the earliest investor in MosaicML, the AI startup that just sold to Databricks for more than $1 billion; some notes on Google’s AR strategy; a scoop about Meta getting into the app distribution game; drug use in Silicon Valley; a Musk-versus-Zuckerberg cage match update; and more…

But first: Thanks to Roblox CEO David Baszucki for taking the time to chat with me for the latest episode of Decoder. You can listen to our full conversation, read the transcript, or watch a shorter version on YouTube.

A schedule note: Command Line will be going dark next week. I’ll be back the week of July 10th.

A reminder that you can access the full archive of Command Line issues here online.


Abraham Shafi.

IRL’s ex-CEO speaks out

I first heard about a social media startup called IRL in April of 2021. A source who had invested in the company offered to connect me with its CEO, Abraham Shafi. Shafi is the real deal, I was told. He could be the next Evan Spiegel or Mark Zuckerberg.

These kinds of introductions happen all the time in my line of work, and you never want to miss out on building an early relationship with the next world-changing entrepreneur. I accepted the intro offer and was soon on the phone with Shafi.

He immediately impressed me with his charisma and vision. He eloquently explained how he wanted to build “Facebook groups and events for the generation that doesn’t use Facebook.” He said IRL (“in real life”) had found its early users by focusing on middle America, with most of its usage coming from young people in states like Texas. “We wanted to build for the average user, not the cool tech user,” he told me.

Shafi was a repeat entrepreneur who had successfully sold his prior startup, a recruiting software platform. And even though IRL was quite small relative to the companies I normally cover, he’d already attracted a who’s who of investors and advisors, including Scott Banister, a member of the PayPal mafia. “Some consumer products are an overnight success, but a lot of the ones that have worked for us can take years,” Mike Maples, an IRL board member and early backer of Twitter and Twitch, told me at the time. “Now we’ve got enough critical mass that things are starting to get interesting.”

Things did get interesting, just not in the way Maples was suggesting to me then. After raising $200 million at a unicorn valuation and claiming to have 20 million users, IRL announced last week that it’s shutting down after a special committee of its board found that 95 percent of those users are fake. The SEC has been conducting an investigation for months, and Shafi was ousted as CEO at the end of April after the board said it discovered a pattern of misconduct.

These kinds of implosions have grown increasingly common as more money has funneled into tech. Even still, I feel a personal connection to this story, given the early platform I gave IRL both at The Verge and my former employer, The Information. I realized Shafi, who goes by Abe, had yet to speak out since he was fired, so I decided to reach out after the news of IRL’s dissolving came out. I knew it was a long shot, especially given that the SEC is still investigating the situation.

But to my surprise, he responded.

“Hey Alex,” he texted me the other day. “The 95% number shocked me and many others at the company. We have not been given a basis to justify this number and believe it is wrong. Although all social media companies contend with bots, my team and I routinely took steps to detect and expel them, and believed in our metrics.”

After defending his approach to fighting spam and fraud while CEO, he went on to tell me that “my suspension had nothing to do with bots. I was told that the action was taken due to a personal expenses issue. I had put both personal and business charges on the same credit card. Although the account was opened on my personal credit and I was responsible for all charges, the company paid the monthly statements and I reimbursed the company for the amounts owed to cover my personal charges. Although I am not proud of this mess, these expenses do not impact our product, our metrics, or our team.”

In an internal email explaining the company shutdown last week, IRL employees were told that an “SEC investigation into potential violations of federal securities laws” kicked off the process. “Based on company counsel’s investigation into the allegations raised by the SEC, as well as concerns about management’s response to the SEC investigation, in March 2023 the Special Committee engaged outside counsel to conduct an independent review of the allegations connected to the SEC investigation,” according to the email, which concludes that “the company’s going forward prospects are unsustainable.”

Based on my recent conversations with current and former IRL employees, including those who are sympathetic to Abe and those who think he is a scheming liar, there’s clearly more to this story than I can fit in one column. A spokesperson for the remaining IRL management team declined to comment for this story on the record. (Needless to say, if you know more about what went down, get in touch.)

For now, what I can say is that things went very wrong at IRL. Its blue chip investors are now scrambling to save face, and its employees are suddenly out of a job with little to no severance and worthless equity. “I’m saddened, stunned and frustrated,” one early employee who was there till the bitter end told me.

While I don’t know what will happen next, I won’t be surprised if the SEC files charges against Shafi. (When I asked if he was expecting to be charged, he said “I can’t comment on that.”) I ended our text conversation by asking him what his message was to IRL employees.

Here’s what he wrote back:

“What we built had immense value and ended too soon. The worst part of seeing the headlines is knowing the unfair damage they will cause to the employees who worked their hearts out for years. They deserve nothing but respect for building a product and set of tools to bring people closer together. I just want them to know that I was honored to work alongside them.”


The Databricks logo.

A big AI acquisition and what it means

One of the biggest stories this week was the $1.3 billion purchase of MosaicML by Databricks. It’s a watershed acquisition that, as I alluded to in an earlier issue of this newsletter, signals how important open-source AI models are quickly becoming. It’s also an eye-popping number for a 60-person startup that is less than three years old, though it’s worth noting that the deal is mostly in Databricks stock.

“We were not expecting to sell the company anytime soon,” Lux Capital general partner Shahin Farshchi, who was the earliest investor in Mosaic, told me over the phone the other day. “If you would have asked me a year ago if a company like Snowflake or Databricks would have acquired Mosaic, I would have said, ‘huh?’”

What changed? The “zeitgeist around AI” and the “concerns created” by the rise of OpenAI and other centralized companies in the space, according to Farshchi. “That brought about all these questions around privacy [and] compliance, now that companies have to hand over their data to someone else to benefit from this technology.”

Mosaic’s insight is unique algorithms that take open-source AI models and pretrain them on proprietary company data for a fraction of the cost of other solutions. Unlike OpenAI, Anthropic, and the other closed-source AI firms, the models that Mosaic produces are owned by the companies that create them. By combining Mosaic’s tech with the data management tools and massive sales force at Databricks, the result is a vertical stack for companies to manage and deploy their own models.

Farshchi, who recently testified before Congress about the risks and opportunities of AI, compared it to the advent of Amazon Web Services and the impact that had on startups that suddenly had easy access to cloud hosting: “Companies that couldn’t leverage AI because they couldn’t let their data go somewhere else now have an option to access, train, and deploy.” He expects an “explosion” of new startups now that the cost of building generative AI experiences with open-source technology is rapidly coming down.

Another side effect is that large funding rounds for early-stage, pre-revenue AI startups — like the $1.3 billion round Inflection AI announced just yesterday — may become a thing of the past. “If you go out to build a foundational model company, you need $100 million plus to train models,” explained Farshchi. “As a result, you have to raise at a big valuation to avoid too much dilution. This will break that cycle. You no longer need to raise at a crazy price to do interesting things in AI.”


Google logo

Notes on Google’s AR pivot

Insider published an interesting story this week about Google’s changing AR strategy. It builds off a report I published last year about the company’s “Project Iris” headset, which has now been canned after the departure of its leader, Clay Bavor.

The big shift is that, according to Insider and my own sources, Google has given up on building its own in-house pair of AR glasses. It’s now shifting to the same playbook for Android on mobile and hoping to power devices made by hardware companies like Samsung. All of the work that went into building custom silicon and a fully custom OS for Iris has been thrown out in favor of an “Android for XR” approach that can be licensed out.

I’m told that the team building AR hardware was severely impacted by the recent layoffs, putting the future of Google’s acquisition of the MicroLED company Raxium into question. Shahram Izadi oversees the roughly 800-person AR software organization and reports directly to Hiroshi Lockheimer, the SVP of Android and Chrome.

Another layer here is that the dynamic between Samsung and Google seems to be growing increasingly complicated as Google continues investing in its own Pixel line of hardware. Google needs Samsung for Android distribution, and now it’s working with Samsung on some kind of mixed reality headset.

“Google big bets are only as strong as the leader advocating for the bet/funding to continue,” one person who worked on Iris recently told me. “So for AR at the moment, Samsung is the sugar daddy supporting the team building goggles. If there is success with Samsung on goggles, then maybe Iris will have another shot at life? That’s the current level of commitment.”


Image of Meta logo

Meta plans its own kind of app store

Speaking of Google, I have a feeling that the folks over in Mountain View are paying attention to this: Meta is planning to let people in the EU directly download apps through Facebook ads, setting the company up to eventually compete with both Google and Apple’s app stores.

The new type of ad is set to start as a pilot with a handful of Android app developers as soon as later this year, I’ve learned. Meta sees an opening to try this thanks to new regulation in the EU called the Digital Markets Act (DMA) that is expected to go into effect next spring. It deems Apple and Google as “gatekeepers” and requires that they open up their mobile platforms to alternative methods of downloading apps.

Android technically allows sideloading already, though Google makes it difficult by coupling its in-app billing and licensing with the Play Store, along with the scary warnings it shows when someone tries to download an Android app from another source. Even still, Meta clearly thinks it’s safer to try its test first on Android rather than Apple’s iOS.

Meta’s pitch to developers participating in the pilot is that, by hosting their Android apps and letting Facebook users download them directly without being kicked out to the Play Store, they’ll see higher conversion rates for their app install ads. At least initially, Meta doesn’t plan to take a cut of in-app revenue from participating apps, so developers in the pilot could still use whatever billing systems they want.

A spokesperson for Meta, Tom Channick, confirmed the plan to me in an emailed statement: “We’ve always been interested in helping developers distribute their apps, and new options would add more competition in this space. Developers deserve more ways to easily get their apps to the people that want them.” Google didn’t respond to a request for comment.

Meta isn’t alone in wanting to become a distributor of mobile apps when the EU’s DMA goes into effect. In March, Microsoft said it hoped to launch an alternative app store for games on iOS and Android in Europe next year.


Image of Elon Musk

Drug use in Silicon Valley

This Wall Street Journal story about the widespread drug use in tech circles is a fun read. I have personally been offered ketamine by an executive at a fancy tech conference, and I don’t even work in tech!

I’m jealous of the anecdotes that Kirsten Grind and Katherine Bindley from WSJ got into their piece. It all reminds me of a juicy tip I recently heard about a very prominent tech CEO who may or may not be mentioned in the story.

According to someone who said they witnessed the situation unfold firsthand, this CEO’s security team made everyone leave an upstairs room at a mansion party in Los Angeles so that he could drink molly water without any witnesses. As the saying goes, it’s lonely at the top.



A Musk versus Zuckerberg fight update

Lex Fridman has trained with both men and posted a video of his session with Zuckerberg. Musk also wants to train with UFC legend Georges St-Pierre, while heavyweight champion Jon Jones has offered to be Zuckerberg’s training partner.

Meanwhile, the Italian government has apparently offered to host the fight in the actual ​​Colosseum.

Musk’s dad on the situation: “The thing is, if this crazy fight goes ahead, if Elon beats this guy, Elon will be called a bully, being so much heavier and taller…While if he loses, the humiliation would be total…Elon loses if he wins and loses if he loses.”


Satya Nadella.

Quote of the week

“The dominant player there [Sony] has defined market competition using exclusives, so that’s the world we live in. I have no love for that world.” – Microsoft CEO Satya Nadella.

Another big story this week is the FTC v. Microsoft trial, which will determine whether Microsoft’s $68.7 billion proposed acquisition of Activision Blizzard goes through. My colleague Tom Warren has been providing unparalleled, blow-by-blow coverage of the whole saga that I recommend you check out if you’re interested in the details.

For me, these kinds of trials are a great window into how big companies actually operate on the inside, thanks to the discovery process that produces internal strategy documents (sometimes by mistake).

I have no idea if Microsoft will win the case or not, but if you care about the business of gaming, this trial has been very illuminating.


People moves

  • Bret Taylor joined Shopify’s board of directors.
  • Kelsey Hightower, Google Cloud’s principal developer advocate, is retiring.
  • Tal Broda, formerly head of engineering for the crypto VC firm Paradigm, has joined OpenAI as “scaling team lead.”
  • Ren Ito, the former COO of Stability AI, was “let go,” according to the CEO, Emad Mostaque.
  • Danny Lange, Unity’s former SVP of AI, has joined Google as VP of “business intelligence” for AI.
  • Оlga Belogolova, who led Meta’s policy team for countering influence operations, is leaving.
  • Brian Janous, Microsoft’s VP of cloud energy strategy, is leaving.
  • Jason Mok has joined Brex from Andreessen Horowitz to be its new head of startups.
  • Brooke Oberwetter, one of TikTok’s most vocal spokespeople, is leaving.

Interesting links


That’s it for this week.

A reminder that I’m off next week. The next issue will be sent the week of July 10th.

In the meantime, I’d love to hear your feedback. Thanks for subscribing, and have a great holiday weekend.

Continue Reading

Trending Spy News

FTC investigating OpenAI on ChatGPT data collection and publication of false information


OpenAI CEO Samuel Altman Testifies To Senate Committee On Rules For Artificial Intelligence
Photo by Win McNamee / Getty Images

The Federal Trade Commission (FTC) is investigating ChatGPT creator OpenAI over possible consumer harm through its data collection and the publication of false information.

First reported by The Washington Post, the FTC sent a 20-page letter to the company this week. The letter requests documents related to developing and training its large language models, as well as data security.

The FTC wants to get detailed information on how OpenAI vets information used in training for its models and how it prevents false claims from being shown to ChatGPT users. It also wants to learn more about how APIs connect to its systems and how data is protected when accessed by third parties.

The FTC declined to comment. OpenAI did not immediately respond to requests for comment.

This is the first major US investigation into OpenAI, which burst into the public consciousness over the past year with the release of ChatGPT. The popularity of ChatGPT and the large language models that power it kicked off an AI arms race prompting competitors like Google and Meta to release their own models.

The FTC has signaled increased regulatory oversight of AI before. In 2021, the agency warned companies against using biased algorithms. Industry watchdog Center for AI and Digital Policy also called on the FTC to stop OpenAI from launching new GPT models in March.

Large language models can put out factually inaccurate information. OpenAI warns ChatGPT users that it can occasionally generate incorrect facts, and Google’s chatbot Bard’s first public demo did not inspire confidence in its accuracy. And based on personal experience, both have spit out incredibly flattering, though completely invented, facts about myself. Other people have gotten in trouble for using ChatGPT. A lawyer was sanctioned for submitting fake cases created by ChatGPT, and a Georgia radio host sued the company for results that claimed he was accused of embezzlement.

US lawmakers showed great interest in AI, both in understanding the technology and possibly looking into enacting regulations around it. The Biden administration released a plan to provide a responsible framework for AI development, including a $140 million investment to launch research centers. Supreme Court Justice Neil Gorsuch also discussed chatbots’ potential legal liability earlier this year.

It is in this environment that AI leaders like OpenAI CEO Sam Altman have made the rounds in Washington. Altman lobbied Congress to create regulations around AI.

Continue Reading

Trending Spy News

OpenAI will use Associated Press news stories to train its models


An illustration of a cartoon brain with a computer chip imposed on top.
Illustration by Alex Castro / The Verge

OpenAI will train its AI models on The Associated Press’ news stories for the next two years, thanks to an agreement first reported by Axios. The deal between the two companies will give OpenAI access to some of the content in AP’s archive as far back as 1985.

As part of the agreement, AP will gain access to OpenAI’s “technology and product expertise,” although it’s not clear exactly what that entails. AP has long been exploring AI features and began generating reports about company earnings in 2014. It later leveraged the technology to automate stories about Minor League Baseball and college sports.

AP joins OpenAI’s growing list of partners. On Tuesday, the AI company announced a six-year deal with Shutterstock that will let OpenAI license images, videos, music, and metadata to train its text-to-image model, DALL-E. BuzzFeed also says it will use AI tools provided by OpenAI to “enhance” and “personalize” its content. OpenAI is also working with Microsoft on a number of AI-powered products as part of Microsoft’s partnership and “‘multibillion dollar investment” into the company.

“The AP continues to be an industry leader in the use of AI; their feedback — along with access to their high-quality, factual text archive — will help to improve the capabilities and usefulness of OpenAI’s systems,” Brad Lightcap, OpenAI’s chief operating officer, says in a statement.

Earlier this year, AP announced AI-powered projects that will publish Spanish-language news alerts and document public safety incidents in a Minnesota newspaper. The outlet also launched an AI search tool that’s supposed to make it easier for news partners to find photos and videos in its library based on “descriptive language.”

AP’s partnership with OpenAI seems like a natural next step, but there are still a lot of crucial details missing about how the outlet will use the technology. AP makes it clear it “does not use it in its news stories.”

Did you miss our previous article…
https://eyespypro.com/congressistrying-to-stop-discriminatory-algorithms-again/

Continue Reading

Trending Spy News

Congress is trying to stop discriminatory algorithms again


A person with their hand hovering over the Like button on Facebook.
Photo by Amelia Holowaty Krales / The Verge

US policymakers hope to require online platforms to disclose information about their algorithms and allow the government to intervene if these are found to discriminate based on criteria like race or gender.

Sen. Edward Markey (D-MA) and Rep. Doris Matsui (D-CA) reintroduced the Algorithmic Justice and Online Platform Transparency Act, which aims to ban the use of discriminatory or “harmful” automated decision-making. It would also establish safety standards, require platforms to provide a plain language explanation of algorithms used by websites, publish annual reports on content moderation practices, and create a governmental task force to investigate discriminatory algorithmic processes.

The bill applies to “online platforms” or any commercial, public-facing website or app that “provides a community forum for user-generated content.” This can include social media sites, content aggregation services, or media and file-sharing sites.

Markey and Matsui introduced a previous version of the bill in 2021. It moved to the Subcommittee on Consumer Protection and Commerce but died in committee.

Data-based decision-making, including social media recommendation algorithms or machine learning systems, often lives in proverbial black boxes. This opacity sometimes exists because of intellectual property concerns or a system’s complexity.

But lawmakers and regulators worry this could obscure biased decision-making with a huge impact on people’s lives, well beyond the reach of the online platforms the bill covers. Insurance companies, including those working with Medicaid patients, already use algorithms to grant or deny patient coverage. Agencies such as the FTC signaled in 2021 that they may pursue legal action against biased algorithms.

Calls to make more transparent algorithms have grown over the years. After several scandals in 2018 — which included the Cambridge Analytica debacle — AI research group AI Now found governments and companies don’t have a way to punish organizations that produce discriminatory systems. In a rare move, Facebook and Instagram announced the formation of a group to study potential racial bias in its algorithms.

“Congress must hold Big Tech accountable for its black-box algorithms that perpetuate discrimination, inequality, and racism in our society – all to make a quick buck,” Markey said in a statement.

Most proposed regulations around AI and algorithms include a push to create more transparency. The European Union’s proposed AI Act, in its final stages of negotiation, also noted the importance of transparency and accountability.

Continue Reading

Trending