Welcome to the Techmeme Ride Home for Wednesday, February 28th, 2024. I'm Brian McCalla. Today, the Apple Car Project is dead. Really this time, we think. Google is working to fix Gemini. If you're using Tumblr, your posts might soon be training open AI's models. Clarnas AI is doing the work of hundreds of humans and another fever dream of the most recent boom years might be over. Here's what you missed today on the world of tech.
Bombshell Apple News for you right now. The Apple Car Project is dead. So says Apple rumor expert Mark Gurman in Bloomberg. He says Apple is winding down Project Titan. It's decade-long effort to make an electric car. Employees on some car teams will move to Apple's AI division. This is shocking news. Maybe a decade of effort here. Untold number of dollars spent. But maybe the AI moment has claimed one of Apple's wilder products.
quoting Gurman. Apple made the disclosure internally Tuesday, surprising the nearly 2000 employees working on the project said the people who asked not to be identified because the announcement wasn't public. The decision was shared by chief operating officer Jeff Williams and Kevin Lynch, a vice president in charge of the effort according to
the people. The two executives told staffers that the project will begin winding down and that many employees on the car team known as the special projects group or SPG will be shifted to the artificial intelligence division under executive John Giannandrea. Those employees will focus on generative AI projects and increasingly key priority for the company. The Apple Car team also has several hundred hardware engineers and car designers. It's possible they will be able to
apply for jobs on other Apple teams. There will be layoffs, but it's unclear how many. The decision to ultimately wind down the project is a bombshell for the company ending a multi-billion dollar effort that would have vaulted Apple into a whole new industry. The tech giant started working on a car around 2014, setting its sights on a fully autonomous electric vehicle with a limousine like interior and voice guided navigation. Apple's most senior executives finalized the
decision in recent weeks according to the people. It comes just a month after Bloomberg news reported that the project reached a make or break point. The most recent approach discussed internally was delaying a car release until 2028 and reducing self-driving specifications from level 4 to level 2 plus technology. Most recently Apple had imagined the car being priced at around $100,000, but executives were concerned about the vehicle being able to provide the profit margins that Apple
typically enjoys on its products. The company's board was also concerned about continuing to spend hundreds of millions of dollars a year on a project that may never see the light of day. Somewhere, Gene Munster is thinking, wait, maybe we can move those dollars into my long dreamed of Apple TV. I wanted to follow that up with a segment pointing out why the Apple car, if it is dead, is a big deal. Longtime listeners know this project has gone back and forth several times.
I've never really known if it was real or not. I mean, it was real in terms of the work hours and dollars Apple invested, but was Apple really ever going to do this? I think they were for one simple reason, the law of large numbers. If you're Apple and you want to grow revenue, what consumer products are available to you that you can sell enough of to meaningfully contribute revenue in a way that it will move the needle? Fones are this great moneymaker because
they're basically $1,000 products that every adult human upgrades regularly. Cars are a much larger ticket item that most households in North America have and upgrade regularly too. Now, Apple might have had to walk away from this project because it just wasn't going to work out for technical reasons, the tech or the product or whatever, but by walking away, that signals something about Apple's larger strategy over the next decade or more. This is something that
Mark German himself went into in a follow-up post. In short, was the AI moment the tipping point for the car project, quote, in abandoning plans for a self-driving car, Apple is giving up on billions in potential revenue and the dream of selling what one executive called the ultimate mobile device. The hope is that other big bets, including generative AI and mixed reality headsets, can make up the difference. The upshot, Apple's future isn't going to hinge on selling $100,000
cars with self-driving features. Instead, it will focus on catching up with rivals in the generative AI industry where chatbots from OpenAI and Google have captured the imagination of consumers and investors. The shift also lets Apple concentrate on turning the Vision Pro headset, still a fledging product into a mainstream hit. But the decision also eliminates a future revenue source at a time when Apple has been struggling to maintain growth. Though the company managed to pull out of a
sales slump last quarter, it warned that the current period will be sluggish again. The Vision Pro just launched this month and isn't expected to be a major contributor to growth for years, if ever. With a car, profit margins would have been slim, but the revenue potential was massive. The idea was long-tatted as one of Apple's famous next big things and could have more firmly
locked consumers into the company's ecosystem. Tesla, which led the EV revolution in the US, generated nearly $100 billion in revenue last year, furthermore, tech giants like Alphabet and Chinese Rivals remained focused on cars. With the car project off the table, Apple can concentrate on applying AI to its current products, including the iPhone and iPad, and avoid falling further behind tech peers like Google, Amazon, and Microsoft.
Apple's artificial intelligence team for the car, which had been reporting to Stewart Bowers, will be shifted to work on generative AI in John Jean and Dre's machine learning division. Some of the car talent could also apply their skills to the Vision Pro. That product has many features, including the ability to create a virtual representation of its wearer that rely on artificial intelligence. The big question is how soon AI might make
serious money for Apple. It's unlikely that the company will have a full-scale AI line-up of applications and features for a few years. And Apple's penchant for user privacy could make it challenging to compete aggressively in the market. So in short, maybe they need to marshal their resources for the AI moment. But given the pure dollars and cents of it,
will the faint idea of an Apple car ever really go away? You could always do something like Byrvian and shortcut your way to a minimum viable player in the market. They've seen where Google CEO Sundar Pachai called Gemini's recent problematic responses around race, quote, completely unacceptable and vowed to make structural changes to fix the problem. Quote, Google suspended its Gemini Image Creation tool last week after it generated
embarrassing and offensive results. In some cases, declining to depict white people or inserting photos of women or people of color when prompted to create images of Vikings, Nazis and the Pope. I know that some of its responses have offended our users and shown bias to be clear that's completely unacceptable and we got it wrong, Pachai said. Pachai said the company has already made progress in fixing Gemini's guardrails.
Our teams have been working around the clock to address these issues. We're already seeing a substantial improvement on a wide range of props, he said. Our mission to organize the world's information and make it universally accessible and useful is stank-resynced. We've always sought to give users helpful, accurate and unbiased information in our products. That's why people trust them. This has to be our approach for all our products, including our emerging AI products.
We'll be driving a clear set of actions, including structural changes, updated product guidelines, improved launch processes, robust eva-hals and red teaming and technical recommendations. We are looking across all of this and will make the necessary changes end quote. Meanwhile, 404 media has seen internal documents suggesting Tumblr staff are compiling users data in preparation for deals where Tumblr owner automatic
would sell that data to open AI and mid-journey to train AI tools. Quote, the exact types of data from each platform going to each company are not spelled out in documentation we've reviewed. But internal communications reviewed by 404 media make clear that deals between automatic the platform's parent company and open AI and mid-journey are imminent. The internal
documentation details a messy and controversial process within Tumblr itself. One internal post made by Kyle Gage, a product manager at Tumblr states that a query made to prepare data for open AI and mid-journey compiled a huge number of user posts that it wasn't supposed to. It is not clear from Gage's post whether this data has already been sent to open AI and mid-journey or whether Gage was detailing a process for scrubbing the data before it was to be sent.
Gage's post makes clear that engineers are working on compiling a list of post IDs that should not have been included and that password protected posts, DMs and media flagged as CSAAM and other community guidelines violations were not included. Automatic plans to launch a new setting on Wednesday, Brian cutting in here to say that would be today, that will allow users to opt out of data sharing with third parties including AI companies according to the source who spoke on the condition of
anonymity and internal documents. A new FAQ section we reviewed is titled What Happens When You Opt Out, stating that quote, if you opt out from the start we will block crawlers from accessing your content by adding your site on a disallowed list. If you change your mind later we also plan to update any partners about people who newly opt out and ask that their content be removed from past sources and future training. A statement published by Automatic after this article was published
specifically mentions WordPress.com which are blogs that Automatic hosts as a service. There is separately an open source WordPress CMS that people and businesses use on self-hosted websites. What remains unclear is whether self-hosted WordPress blogs that use popular automatic plugins like Jetpack to connect those blogs with automatic infrastructure are subject to the company's AI scraping deals. Automatic did not immediately respond to a question about whether
sites using Jetpack are subject to its data sharing agreements. Another internal document shows that on February 23rd an employee asked in a staff only thread quote, do we have assurances that if a user opts out of their data being shared with third parties that are existing data partners will be notified of such a change and remove their data. Separately 404 media viewed a public now deleted post by Gage the product manager where he said that he was deleting all of his
images off of Tumblr and would be putting them on his personal website. A still live post says quote, I've deleted my photography from Tumblr and will be moving it slowly but surely over to KyleGage.com which I'm building into a photography portfolio that I can control and to end. At one point last week his personal website had a specific note stating that he did not consent to AI scraping of his images. Gage's original post has been deleted and his website is now a
blank page that just reads Kyle. Gage did not respond to a request for comment from 404 media. Meanwhile, Clorna claims its chat GPT-powered AI chatbot handles two-thirds of all of its customer service chats and does the equivalent work of 700 full-time human agents. The app-based AI chatbot handles two-thirds of all of its customer service chats.
Two-thirds of its customer service chats have been added to some 2.3 million conversations so far with the virtual assistant earning customer satisfaction ratings at the same level as human agents. Clorna is expected to go public this year and will need all the hype it can get at a time when investors have been generally frosty toward IPOs. Estimates that the chatbot could
help improve its profits by $40 million in 2024. Announcing a partnership with OpenAI earlier last year, Clorna said it was one of the first companies to integrate the firm's groundbreaking chat GPT technology into a plug-in for shopping. The natural language interface initially helped customers choose items and make other shopping-related decisions based on personalized queries. A feature Clorna described as smooth shopping. The company has continued to build out its AI
offering since then. Its app-based assistance are now available to customers worldwide and handle a variety of tasks including refunds, cancellations and even disputes. Clorna boasted in its announcement on Tuesday that the AI assistant is doing the equivalent work of 700 full-time agents. That statement may raise eyebrows for anyone who remembers the middle of 2022 when the company
laid off roughly the same number of employees than about 10% of its staff. At the time, CEO Sebastian Simitowski cited economic uncertainty and the likelihood of a recession as reasons for the cuts. He was criticized for his handling of the staff reduction after he shared a public spreadsheet on LinkedIn that contained the names of many of the layoff workers. Fast company asked Clorna how the company arrived at its calculation for its AI assistant's human equivalent productivity.
The company said the number of equivalent jobs the AI could perform wasn't related to the layoffs and a statement a spokesperson said the company's customer service is supported by four to five large third parties that collectively have over 650,000 employees and that it offers customers the option to speak with human agents if that's what they prefer. This is in no way connected to the
workforce reductions in May 2022 and making that conclusion would be incorrect. The statement read, we chose to share the figure of 700 to indicate the more long-term consequences of AI technology where we believe it is important to be transparent in order to create an understanding in society. We think it's important to proactively address these issues and encourage a thoughtful discussion around how society can meet and navigate this transformation. End quote.
Finally is another dream of the low interest rate investing environment. Dead Thrasio, which raised more than $3 billion in equity and debt to acquire third-party Amazon sellers, has filed for Chapter 11 bankruptcy and secured $90 million in emergency funding. The news should not come as a surprise. There have been multiple rumors of the companies impending bankruptcy since last year. Since 2022, the company has been laying off employees and taking other
steps to restructure its business, such as pulling out of certain markets. Thrasio overall has been a victim of a perfect storm of market conditions, plus its own business model. Amid the major downturn and fundraising that hit privately held tech companies starting at the end of 2021 and still ongoing, late-stage companies which needed the most to stay afloat yet were not in a position to IPO, were especially in a tight bind to stay afloat. Thrasio was a case study in
late-stage startups. Over several years, it raised well over $3 billion in funding across equity and debt rounds. Money it pulled together from investors like Silver Lake, Oak Tree, Ennova, and many more to itself buy up a wide range of smaller e-commerce businesses, built to run on Amazon's fulfillment infrastructure, but with little appetite to continue
and scale those enterprises on their own. Thrasio's pitch, the same one used by the many other roll-up plays that are still on the market today, was that by buying up the best of these companies, there are millions of them in existence globally. It could consolidate production distribution in marketing. It would have unprecedented access to data that it could use across the wider business to improve results overall, and it could build new technology to improve that larger operation.
Our business is getting better as it gets bigger and these investments will be invaluable as we continue on that path, said Carlos Cashman, one of the co-founders in 2021 when he was still the CEO. At the time, the company had just raised $1 billion at evaluation, it said of up to 10 billion. Josh Silverstein, another co-founder who is no longer with the company, told TechCrunch in 2021 that Thrasio made a profit of $100 million on revenues of $500 million in 2020. None of that
really played out as planned as you can probably guess. Consolidating disparate businesses is not as easy to do as it is to say. Consumer taste for goods shift all the time and more over, e-commerce has seen a lot of pressure due to the economy tightening, meaning sales targets were
likely hard to make on what might have been a wobbly cost base. Thrasio is the most notable of the rollups to collapse, but with companies like branded, Berlin Brands Group, seller X, Hayday, Heroes, Perch, and more, collectively raising more than $1 billion to jump into the aggregation race, it is unlikely to be the last. Quick ask of the hive mind. Anyone know Nat Friedman well enough to do an email intro? Would love to interview him on this show.
We do have loose connections as we're co-investors and a couple of AI companies, but I've not been able to make solid contact with Nat for the purposes of coming on the show, so if you thought you could get me some meaningful intro, to Nat, please get in touch, Brian at techmeam.com, talk to you tomorrow.