Upstart stock dropped as much as 23.5% in Wednesday trading despite beating earnings expectations.
The lending software company offered lower guidance for the fourth quarter.
Upstart is coming off a strong recent rally of 129% over the past three months.
Upstart stock plummeted as much as 23.5% Wednesday as the lending software company sees EBITDA slipping in the current quarter.
The sell-off came despite strong third-quarter results. Net income tripled from a year ago to $29.1 million, topping Wall Street forecasts for $20 million. Revenue soared 249% to $228.5 million, beating analyst expectations of $214.9 million.
But management sees fourth-quarter EBITDA falling to $51 million-$53 million from $59.1 million in the third quarter.
Analysts at Bank of America cut their price target on Upstart stock to $255 and maintained an underperform rating.
Shares were down 18% at $256.11 at 12:01 p.m. ET after diving as low as $239.98 earlier. Wednesday’s plunge follows a recent rally for Upstart stock, which before this week was up 129% over the past three months, and 670% so far on the year.
The company, which uses artificial intelligence to determine loan eligibility, first went public in December 2020.
Upstart connects banks with customers interested in taking out loans, then it collects referral fees from partner banks but doesn’t do the lending itself.
According to Upstart, its AI technology makes lending decisions based on considerations from over 1,600 data points, which allow it to reduce risk while facilitating lower rates for loans.
“Since Upstart’s IPO a year ago, we’ve more than tripled our revenue, tripled our profits, tripled the number of banks and credit unions on our platform, and tripled the number of auto dealerships we serve,” said Upstart CEO Dave Girouard.
Meredith Whittaker (right), a former Google AI researcher, is joining the Federal Trade Commission.
AP Photo/Bebeto Matthews
Former Google AI researcher Meredith Whittaker has joined the Federal Trade Commission.
While at Google Whittaker helped organize a massive employee walkout.
Her role at the FTC will be senior adviser on AI.
Meredith Whittaker, a former Google AI researcher and activist employee who left the company in 2019, announced she is taking a job with the Federal Trade Commission (FTC).
Whittaker said in a tweet Wednesday she will be joining the FTC as a senior adviser on AI.
“I’m humbled and beyond excited to get to work with + learn from @linakhanFTC [FTC chair Lina Khan] and her brilliant team,” Whittaker said.
Whittaker was one of the lead organizers of a massive employee walkout at Google in 2018, where more than 20,000 workers left work to protest the company’s handling of harassment allegations.
She sent an internal company memo in April 2019 saying she was facing retaliation as a result of her involvement in the walkout, and left the company in July 2019.
Whittaker is currently faculty director of the AI Now Institute at New York University.
FTC Chair Lina Khan was appointed to her position in June and is viewed as a major critic of Big Tech companies.
Former Google CEO Eric Schmidt said the metaverse is “not necessarily the best thing for human society.”
Schmidt spoke with the New York Times about his concerns about the future of artificial intelligence technology.
The former executive said he believes AI technology like the metaverse will eventually replace human relationships.
Former Google CEO Eric Schmidt is joining the sea of voices weighing in on Facebook’s metaverse and expressing concern about the future of artificial intelligence technology.
Schmidt, who served as Google’s top executive from 2001 to 2011 and as executive chairman until his departure in May 2020, told the New York Times that while he believes the technology will soon “be everywhere,” he warns it is “not necessarily the best thing for human society.”
“All of the people who talk about metaverses are talking about worlds that are more satisfying than the current world – you’re richer, more handsome, more beautiful, more powerful, faster,” Schmidt told the Times. “So, in some years, people will choose to spend more time with their goggles on in the metaverse. And who gets to set the rules? The world will become more digital than physical. And that’s not necessarily the best thing for human society.”
“It will be everywhere,” he told New York Times opinion columnist Maureen Dowd. “What does an A.I.-enabled best friend look like, especially to a child? What does A.I.-enabled war look like? Does A.I. perceive aspects of reality that we don’t? Is it possible that A.I. will see things that humans cannot comprehend?”
The former Google executive isn’t alone in his concerns about AI. The technology has been increasingly criticized by business leaders in recent months, including Tesla CEO Elon Musk, who said his confidence is “not high”in the transparency and safety of AI within his own company. Meanwhile, some analysts say augmented reality poses even more risks of abuse than social media.
Schmidt’s comments come after Facebook announced Thursday it was changing its corporate name to Meta, and creating the metaverse as a virtual space where people can interact digitally using avatars. The company has been at the center of significant criticism in recent weeks after leaked documents exposed the company’s controversial business practices and technology.
Since then, Facebook has increasingly emphasized its metaverse mission in an attempt to distance itself from the controversy. The company has since pushed back against the reports, calling them mischaracterizations. Facebook CEO Mark Zuckerberg told The Verge it was “ridiculous” for people to think that he changed Facebook’s name to Meta because of the backlash surrounding the leaked documents.
“From now on, we’ll be metaverse first, not Facebook first,” CEO Mark Zuckerberg said during the company’s Oculus Connect event. “Over time, you won’t need to use Facebook to use our other services.”
Facebook and Instagram usage among younger populations is already dwindling, as the platforms are increasingly being replaced by apps like TikTok and Snapchat. According to Piper Sandler’s “Taking Stock With Teens”, 81% of teens surveyed said they used Instagram, the highest percentage out of all the platforms. 77% said they use Snapchat and 73% said they use TikTok. Only 27% of respondents said they use Facebook, the least of all the platforms.
Visitors watch a digital twins simulation service system for sports venues by Intel at the Apsara Conference, a cloud computing and artificial intelligence (AI) conference, in Hangzhou, in China’s eastern Zhejiang province on October 19, 2021.
Photo by STR/AFP via Getty Images
Digital twins are used by businesses to run simulations and identify issues before they occur.
Digital twins use data on social media, consumers, and global factors to run simulations.
As supply chain disruptions pummel industries across the globe ahead of the holiday season, a simulation tool powered by artificial intelligence (AI) has the power to help businesses predict disruptions to their supply chains and minimize their effects, according to MIT Technology Review.
Digital twins are virtual representations of real-world objects or systems, such as global supply chains, that are used to run simulations and identify issues before they occur. MIT Data Science Lab Director David Simchi-Levi told Technology Review that an increasing number of companies are using this tool to stress-test their supply chains.
“What if there’s a drought in Taiwan and the water shortage shuts down microchip manufacturing? A digital twin could predict the risk of this happening, trace the impact it would have on your supply chain, and – using reinforcement learning – suggest what actions to take to minimize the harm,” Technology Review reported.
From consumer behavior to geopolitical implications and social media trends, digital twins use vast amounts of data on a variety of factors to run simulations, which trains their AIs to analyze the data and make predictions, according to Technology Review.
A pilot digital twin was created for a Unilever facility in Valinhos, Brazil, that saved the company $2.8 million dollars by cutting down on energy use and driving productivity, the company told WSJ.
While it’s mostly large companies currently utilizing the technology, Simchi-Levi told Technology Review that with a million dollars and 18 months, a company could enjoy “many of the benefits” of digital twins.
ARK chief executive Cathie Wood is firmly on the side of the debate that believes the current surge in inflation will likely be short-lived. In a series of Twitter exchanges on Tuesday with the likes of Jack Dorsey and Elon Musk, she explained the deflationary forces that will help quell the pickup in price pressures, including the role of technology.
Supply-chain bottlenecks at ports, warehouses and rail depots around the world that built up during the worst of the Covid-19 lockdowns, along with shortages of labor and raw material prices have sent the cost of consumer and producer goods and services to multi-year highs.
Central bankers face a delicate balancing act of curbing excessive price increases without dampening economic growth, which in many cases, is still below pre-pandemic levels.
On Monday, Wood took to Twitter to detail three deflationary forces that will “overcome the supply chain-induced inflation,” responding to Jack Dorsey less than 72 hours after he sounded an alarm on rising prices.
Tesla chief executive Musk weighed in. “I don’t know about long-term, but short-term we are seeing strong inflationary pressure,” he tweeted in reply to part of Wood’s thread.
“Inflation has flared in response to COVID-related supply chain bottlenecks and oil supply constraints but, IMHO (in my humble opinion), the powerful and converging deflationary forces associated with AI, energy storage (EVs!), robotics, genomic sequencing, and blockchain technology will bend the curve,” Wood said in response to Musk.
“If they expect lower prices, most consumers/businesses will defer purchases, exacerbating a decline in the velocity of money. Despite the burst in cyclical inflation during the last year, velocity is hovering at low levels. If ARKInvest is correct, the next leg will be down,” she continued.
She pointed to the decline in velocity – the rate at which money turns over per year – and how that will also help offset inflation.
“I am struck by the behavior of millennials who, at the margin, are sacrificing short term consumption to pay down student loans or invest in crypto and other assets. Bank loan growth also is tepid which would not be the case if velocity were increasing,” she said.
Mark Cuban praised cryptocurrencies and Robinhood for encouraging young people to invest.
The “Shark Tank” star compared bitcoin to gold, and called for stablecoins to be regulated.
The billionaire advised parents to join TikTok and YouTube to understand how children learn today.
Mark Cuban praised cryptocurrencies for enticing young people to invest, defended Robinhood, and compared bitcoin to gold on “The Prof G Pod with Scott Galloway” this week.
The “Shark Tank” investor and billionaire owner of the Dallas Mavericks also likened the crypto boom to the dot-com bubble, underscored the need to regulate stablecoins, and recommended parents join TikTok to better understand their children.
Here are Cuban’s 10 best quotes from the interview, lightly edited and condensed for clarity:
1. “Crypto reminds me of 1995. It’s not so much about all the tokens, and all the trading, and all the frenzy that’s going around – that reminds me more of the dot-com stock market. What’s interesting to me about crypto is the actual networking platforms themselves, and the ability to connect with smart contracts.”
2. “You’re seeing millions and millions of people investing in crypto, because it’s far cheaper and simpler than investing in stocks. I don’t think we fully understand the impact that crypto has had.”
3. “Robinhood’s gamification has done a lot of good, contrary to what a lot of people think. We’re not going to close the income gap by taking a hammer and smashing everybody’s income down, because there’s no efficient way to get that money to the people who need it most. Reducing income inequality is really about enabling people to access appreciable assets.”
4. “We should require digital accounts for everybody. It was hard to give millions of people their stimulus checks! That’s just crazy in every way, shape, and form. You can say a lot of bad things about Robinhood and its peers, but they enable anybody of a minimum age to own an appreciable asset at a very, very low cost.”
5. “AI, AI, AI. Crypto, crypto, crypto. That’s it.” – Cuban’s advice to people in their early 20s asking what they should learn about.”
6. “Bitcoin is a store of value, just like gold. The bitcoin maxis – the people who eat, sleep, and breathe bitcoin and think it’s gonna save the world – believe that it’s a reserve currency, and that when fiat crumbles for whatever reason, bitcoin will slide right in as the white knight. I don’t see that at all.”
7. “Gold is sold as a hedge against the apocalypse, a hedge against inflation. When there are things that look really bad, like a pandemic, it doesn’t really respond. Bitcoin is the same way, it’s purely a supply-and-demand play.”
8. “They’re gonna have to regulate stablecoins. There’s just so much money there. If one of these was a Ponzi scheme or a fraud, that would create significant instability.”
9. “We’re in this death war between blockchains. Not all of them are going to survive. It’s really going to be the applications and the innovation that drive their success, just like in the early days of the internet.”
10. “Parents need to understand how kids learn today because it’s different. You’ve gotta be on TikTok because you’ve gotta understand how they’re consuming information. Same with YouTube. – Cuban gave the example of his son learning about gross margin and other business concepts by watching “Shark Tank” clips on TikTok.
Grammy-winning artist Miguel spoke to Insider about the making of a “generative” AI music project he released with the tech company Endel this week.
In the interview, Miguel discussed the project’s aim to promote wellness and gave his perspective on future implementations of AI in music.
Grammy-winning artist Miguel spoke to Insider last week about his work with the tech company Endel on a “generative” AI music project that dropped in Endel’s app on Monday.
Insider beta tested the project, called “Clarity Trip,” ahead of a phone interview with Miguel. Set to ambient instrumentation and spare, wordless vocals from Miguel, the dynamic project uses your phone’s accelerometer and step counter to adapt Miguel’s production to your movement.
In the interview, Miguel discussed the making of the project and its aim to promote movement and wellness. He also touched on how his meditation practices intersect with music, and gave his perspective on future implementations of AI in music, including in his own work.
This interview has been lightly edited and condensed for clarity.
Miguel, this is a remarkable piece of generative music here.
Man.. when they were describing how it would work, I kinda had to go, “Okay, wait a second. Rewind. Okay, say that again. How’s this gonna work?”
[Laughs].
You know, just kinda trusting the guys and getting them what they needed, then letting the algorithm do what it’s programmed to do. It’s been such a dope experience. So yeah, man, I’m glad you’re enjoying it. And what’s awesome is that it’s different every time. So, hopefully it’s good enough to have you come back time and time again. And yeah, really happy with it.
What drew you to the project initially?
I mean, to be honest, just the “what the fuck” factor was enough. Just the idea of it, taking parts that I created, then having an algorithm tailor the listening experience to the listener’s movement. It sounded interesting and cutting edge, and also the team made me feel real comfortable. They seemed like they had a real sense of what they were going for, and they trusted me, and it felt it was going to be really collaborative. And it proved to be, absolutely collaborative. And the guys are excellent. So yeah, that was the drawing factor, the team and just the possibilities, and it being cutting edge.
With the concept behind it in mind, how do you approach creating the body of music that you put into it?
I mean, again, I have to give the team some credit, a lot of credit actually, ’cause they were so patient. I feel like at every turn I was like, “Okay, wait. So, how are we breaking this down?” Because I’m usually imagining a piece of music in.. I’m always starting with maybe a finished something, something finished in my mind is sort of the guiding light, and every choice is always sort of to get to that point of whatever that finished portion was in my mind. With this, it’s sort of more about understanding your choices and then letting the finished product be sort of a question mark, you know? So it was just more about trusting that each piece that you’re adding is reflective of your perspective, your choices, your personality, and so on and so forth.
And I think that’s the most exciting is that I didn’t have to be the visionary here at all. Not only was it collaborative in terms of piecing it all together, but every time any person listens to it, it’s a collaboration with the algorithm, with the program. So, it’s sort of a nice, I think, new way of imagining the possibilities of music, and how we’ll integrate human and technology in expression. I feel like it’s just a really dope.. if nothing else, a really amazing new step in a direction of the future in music and creativity.
I gotta tell you, it took a minute for me to get it right. I started on a walk through Downtown Brooklyn, getting stopped at a light with an ambient chord hanging, someone yelling some shit at me..
[Laughs].
I’d imagine you worked this out with scenic California landscapes in mind, a different type of environment.
[Laughs]. Yeah. You know, it was absolutely meant just to help add in that atmosphere, and then also like for those of us who aren’t in that atmosphere, to hopefully kind of take you to that same place and that same feeling. So yeah, hopefully it added to your experience and maybe transported you a little bit. I think that’s what great music does. I think that was the other interesting possibility, was that in some way, shape, or form, like allowing technology to emote, giving the machine the pieces to emote, and seeing what happens? I think just the sheer.. like just the idea of it is wild, to think that these are possibilities. But I think it’s also really fun. And as an experiment, just another step in new ways of creating, which I’m all about. I’m all for that.
Miguel.
Ethan Gulley
For you personally.. let’s say.. if you’re on an ideal walk, what do you get out of it from like a mental health standpoint, or what do you hope to get out of like a hike, for example?
Man. I mean, I typically hike for clarity. That was actually not intentional, ’cause I know that the piece is called “Clarity Trip.” But yeah, that’s the goal, and that’s what I normally get from just any exercise, but specifically like hiking, getting out of my space, getting out of my workspace, and just sort of recalibrating. That’s what normally happens for me when I’m on a hike, or when I go on a walk or get out of the house and get some exercise, man.
Is that the best form of meditation for you? I can’t like.. I don’t know how people sit down and meditate..
[Laughs].
.. maybe I’ve got like undiagnosed ADHD or something. But do you prefer to like get moving?
You know, there’s a couple of things that are meditative for me. Music is meditative for me. I would say that, yeah, hiking for me is absolutely.. it brings me that very similar kind of mental peace, recalibration. But no, I love to meditate as well, and often I do it with music that helps get me kind of in a good space. So, that’s what I think the goal of each one of the states that we built out for the algorithm to use.. the idea was to create music that sort of made each one of those experiences, whether you’re in a static state, like if you were sitting down and meditating, having a piece of music that would help you in that state or help add to that experience. And as you progress in movement, we tried to build something that really helped add to that experience. And I think, with the help of the team, we were able to create something that’s gonna add value to people, and that’s pretty much all you can hope for when you make things. Is it gonna add value? Is it gonna help people? Is it going to do something for someone? So, yeah.
What type of music would you say is meditative for you? Are you the type of artist who would prefer instrumental music in that type of space?
Yeah. Absolutely, more instrumental music, for me, is meditative. Because as a songwriter, I pay attention to lyrics, so it’s hard to shut that off. So yeah, if I’m looking to meditate, it will absolutely be with music that is just instrumental. Often like not even melodic. Sometimes it’s just tones, with no arrangement. I think that’s.. I mean, scientifically it’s proven that certain tones help put you in different states of mental activity.
Miguel.
Ethan Gulley
Do you go with like a Brian Eno-type vibe? What artists would you go to?
Man, I love.. I mean, jazz can be that way from me. I’m a big Miles head. And even though I know the melodies, it still for whatever reason can be meditative for me. I think because it’s not lyrical, and more expressionist, especially the stuff that I like to listen to from Miles. It’s not arranged, so it feels likes it’s less of a.. I don’t even know what the word would be right now.. You gotta excuse me, it’s Friday. I’m like, “Huh? What?”
[Laughs]. I’m in the same boat.
But yeah, I mean, in terms of stuff I listen to, honestly, for meditative music, it’s normally like downtempo jazz, or it’s literally like frequency tones. So not anything arranged. It’s just more tonal.
I was looking at this Brian Eno speech on generative music. This guy said that, after experimenting with generative music, he couldn’t listen to a normal record the same, it was “very difficult” for him. Has this mode of composition shifted your perception at all? Has the project changed the way you think about things musically at all?
Well, it absolutely opened up the possibilities, because understanding now that you can give.. you can actually program something to.. like AI can analyze the way a musician generally creates, and then create based off of that, and create pieces of music that would mimic those sorts of choices. I think the better and better we get at programming.. I mean, the possibility of programming, according to my past choices and then giving the machine a set of sounds or tones, or specific choices, and then letting it do things that are based on what it’s analyzed about my past choices. I think that’s an interesting way of integrating AI. I think there’s just so many different ways, and that’s just like.. that’s just one run. You can assign it to do specific tasks.
I think the possibilities are endless. I think it’s an interesting space in general. As technology becomes more and more integrated, it’s only going to become.. it’s just the way things are going, you know? Humans are gluttons for convenience, and technology obviously makes things a lot more convenient. So, as we progress, I feel like it’s just the natural way things are gonna go. So, finding ways that we work with technology, I think is just sort of a natural progression. And yeah, it’s just interesting to sort of start playing with and imagining new ways that we may use algorithm and AI and computer learning to help us create new, interesting things that we haven’t done before.
Do you see that factoring into your solo work at all? As such an organically talented musician, do you see that type of technology implementing into what you do, into the music itself?
Just collaborating with the Endel team, I think the hope is that we continue to collaborate and find new ways to implement the technology. So I would love to. I think right now.. not everybody has an Endel team and AI that can do what they program it to do. So, I think it’s early, but that’s sort of why I jumped at the chance because it’s these new ideas that will absolutely be normal in time. So, being able to be one of the first to jump in and learn about it and learn about the process, and be able to witness firsthand the possibilities was what made it a no brainer.
Facebook’s artificial intelligence removes less than 5% of hate speech viewed on the social media platform.
A new report from the Wall Street Journal details flaws in the platform’s strategy to remove harmful content.
Facebook whistleblower Frances Haugen said that the company dangerously relies on AI and algorithms.
Facebook claims it uses artificial intelligence to identify and remove posts containing hate speech and violence, but the technology doesn’t really work, according to internal documents reviewed by the Wall Street Journal.
Facebook senior engineers say that the company’s automated system only removed posts that generated just 2% of the hate speech viewed on the platform that violated its rules, the Journal reported on Sunday. Another group of Facebook employees came to a similar conclusion, saying that Facebook’s AI only removed posts that generated 3% to 5% of hate speech on the platform and 0.6% of content that violated Facebook’s rules on violence.
The Journal’s Sunday report was the latest chapter in its “Facebook Files” that found the company turns a blind eye to its impact on everything from the mental health of young girls using Instagram to misinformation, human trafficking, and gang violence on the site. The company has called the reports “mischaracterizations.”
Facebook CEO Mark Zuckerberg said he believed Facebook’s AI would be able to take down “the vast majority of problematic content” before 2020, according to the Journal. Facebook stands by its claim that most of the hate speech and violent content on the platform gets taken down by its “super-efficient” AI before users even see it. Facebook’s report from February of this year claimed that this detection rate was above 97%.
Some groups, including civil rights organizations and academics, remain skeptical of Facebook’s statistics because the social platform’s numbers don’t match external studies, the Journal reported.
“They won’t ever show their work,” Rashad Robinson, president of the civil rights group Color of Change, told the Journal. “We ask, what’s the numerator? What’s the denominator? How did you get that number?”
Facebook’s head of integrity, Guy Rosen, told the Journal that while the documents it reviewed were not up to date, the intel influenced Facebook’s decisions about AI-driven content moderation. Rosen said it is more important to look at how hate speech is shrinking on Facebook overall.
Facebook did not immediately respond to Insider’s request to comment.
The latest findings in the Journal also come after former Facebook employee and whistleblower Frances Haugen met with Congress last week to discuss how the social media platform relied too heavily on AI and algorithms. Because Facebook uses algorithms to decide what content to show its users, the content that is most engaged with and that Facebook subsequently tries to push to its users is usually angry, divisive, sensationalistic posts that contain misinformation, Haugen said.
“We should have software that is human-scaled, where humans have conversations together, not computers facilitating who we get to hear from,” Haugen said during the hearing.
Facebook’s algorithms can sometimes have trouble determining what is hate speech and what is violence, leading to harmful videos and posts being left on the platform for too long. Facebook removed nearly 6.7 million pieces of organized hate content off of its platforms from October through December of 2020. Some posts removed involved organ selling, pornography, and gun violence, according to a report by the Journal.
However, some content that can be missed by its systems includes violent videos and recruitment posts shared by individuals involved in gang violence, human trafficking, and drug cartels.
Facebook is working on tech that will monitor human life, the company said in a new blog post.
The idea is to build AI that sees the world as humans do, from a first-person perspective.
This AI could be used for what Facebook envisions as the future of smartglasses.
Facebook envisions a future where smartglasses “become as useful in everyday life as smartphones,” the company said in a new blog post.
In order to achieve that future, such devices will require powerful AI software that can read and respond to the world around the headset’s user. And the only way to train AI to see and hear the world like humans do is for it to experience the world like we do: from a first-person perspective.
“Next-generation AI will need to learn from videos that show the world from the center of action,” the blog post said.
Facebook’s solution to this problem is a new project, titled, “Ego4D,” which will collate data from “13 universities and labs across nine countries, who collected more than 2,200 hours of first-person video in the wild, featuring over 700 participants going about their daily lives.”
The data will be open to the research community, the blog post said, but the goal of the project is clear: To create the type of AI that can power a slew of Facebook devices currently in the works.
There’s even a Facebook division, known as Reality Labs, that’s focused on research and development for the future of VR and AR tech.
That division is headed by longtime Facebook exec Andrew “Boz” Bosworth, who shared images of himself in various prototypes this past week:
The company already makes a very popular virtual reality headset in the Oculus Quest 2, and it has plans to transition from VR to augment reality (AR) in the coming years.
Got a tip? Contact Insider senior correspondent Ben Gilbert via email (bgilbert@insider.com), or Twitter DM (@realbengilbert). We can keep sources anonymous. Use a non-work device to reach out. PR pitches by email only, please.
