An outdated Meta AI model was apparently at the center of the Department of Government Efficiency's initial ploy to purge parts of the federal government.
Wired reviewed materials showing that affiliates of Elon Musk's DOGE working in the Office of Personnel Management "tested and used Meta’s Llama 2 model to review and classify responses from federal workers to the infamous 'Fork in the Road' email that was sent across the government in late January."
The "Fork in the Road" memo seemed to copy a memo that Musk sent to Twitter employees, giving federal workers the choice to be "loyal"—and accept the government's return-to-office policy—or else resign. At the time, it was rumored that DOGE was feeding government employee data into AI, and Wired confirmed that records indicate Llama 2 was used to sort through responses and see how many employees had resigned.
Meta is beginning to test showing 3D photos on Instagram if you’re looking at your feed on a Meta Quest VR headset.
“Thanks to our AI view synthesis algorithms, we can transform the existing photos that show up in Instagram feed pixel by pixel – no fancy 3D cameras required,” Meta says in a blog post. “That means flat photos that weren’t originally captured in 3D will automatically be converted into an immersive format that gives 2D images a sense of depth when you view them on Quest.”
The test starts this week. “Not everyone will have access to 3D Instagram photos on Quest – and even if you’re in the test group, you may not see it right away,” Meta says. And it’s not the first time Meta has dabbled with 3D photos; the company rolled out a 3D photo feature for Facebook in 2018.
The test of 3D photos in Instagram is being introduced as part of Meta Quest’s v77 update, which includes a bunch of other features. One notable one, also in testing, is Navigator, which Meta describes as “a new home for your games and apps, friends, notifications, essential system settings, and more on Quest.”
You’ll get to the Navigator by pressing the Meta or Oculus button on the right controller. It offers access to recently-used apps, and you can pin up to 10 items to make them easier to get to. Meta has more information about Navigator in a video on its website.
The v77 update also adds experimental support for Bluetooth Low Energy audio devices, lets you select a window that can follow you as you move, and combines the Meta Quest Link, casting, and remote desktop into a single PC app, Horizon for PC.
In its antitrust case against Meta, the US Federal Trade Commission is asking a judge to consider an alternate reality. In that world, the company never bought Instagram and WhatsApp. The two apps remained competitive with Facebook, developing features that competed for users' attention. And that competition created a thriving ecosystem of social media apps where people can connect with their friends and family.
Meta has spent the past several days - during which it's begun lodging its case-in-chief in a Washington, DC, courthouse - building a counternarrative. In its telling of this alternate present, Instagram and WhatsApp are shadows of what they are in our world. They lacked the resources, expertise, and vision to become robust and valuable online platforms, let alone formidable competitors. And consumers are the ones who ultimately suffered.
One of Meta's key witnesses for this defense is WhatsApp cofounder Brian Acton, who was called on Tuesday to help make its case that WhatsApp users, just like Instagram ones, benefited from Meta's acquisition. Acton was the second app founder to testify in the case, after Instagram cofounder Kevin Systrom delivered mostly blistering te …
Meta is launching a new program to incentivize startups to adopt its Llama AI models. The program, Llama for Startups, provides companies “direct support” from Meta’s Llama team, as well as funding in certain cases. Any U.S.-based firm that is incorporated, has raised less than $10 million in funding, has at least one developer on […]
Mark Zuckerberg, Meta's CEO, at LlamaCon this year.
AP Photo/Jeff Chiu
Meta has instructed managers to rank more employees in its "below expectations" performance rating.
The midyear reviews could set the stage for more performance-based cuts.
The move follows CEO Mark Zuckerberg's internal announcement to "move out low-performers" faster.
Meta is expanding the ranks of its lowest-rated employees in their midyear performance reviews, months after it laid off nearly 4,000 employees whom itlabeled low performers.
It's telling managers to put more employees in its "below expectations" tier, the lowest performance bucket, during this year's midyear performance reviews, according to a memo shared on Meta's internal forum on May 14, which was viewed by Business Insider. For teams of 150 or more, Meta wants managers to put 15% to 20% of employees in the bottom bucket compared with 12% to 15% last year.
The expanded range includes employees who have already left the company as part of "nonregrettable attrition," Meta's term for staff considered noncritical to operations, including those who resigned or were dismissed for underperformance.
The midyear performance review process is "an opportunity to make exit decisions," the memo said. It added that "there will be no company-wide performance terminations, unlike earlier this year," and that leaders are expected to manage the performance of their reports.
Managers can select employees for performance cuts based on criteria including a "below expectations" rating in their midyear review, if they were formally disciplined within the past six months, or if they had an "employee relations" case in the first quarter. Those cases are when an employee was on a plan to manage their performance.
The review process is set to begin on June 16, and conversations between managers and employees on performance are scheduled for between July and August.
The change comes just months after Meta laid off nearly 4,000 employees — about 5% of its workforce — over their performance. Internal documents seen by BI earlier this year suggested such layoffs could become an annual fixture, with CEO Mark Zuckerberg telling staff he had "decided to raise the bar on performance management" and move faster to "move out low-performers."
Meta declined to comment.
The new midyear targets echo a move Meta made at the end of 2022, when it roughly doubled the share of employees placed in its lowest performance categories during annual reviews. At the time, managers were instructed to classify up to 16.5% of staff as underperformers, up from the previous range of 7% to 12%.
As with the current midyear cycle, that figure included employees already marked for nonregrettable attrition. The company also told managers to be more rigorous when evaluating employees on the borderline between performance tiers.
The repeated tightening of performance review criteria underscores Meta's effort to reshape its workforce following years of overhiring. Meta executives have increasingly used performance management as a mechanism to streamline teams and cut costs. Meta's human resources leaders have emphasized a need to "move faster" in managing out underperformers so that new, stronger talent could be brought in.
Meta's move mirrors a broader trend in the tech industry, where companies are sharpening their focus on performance, while doubling down on artificial intelligence investments. Earlier this month, Microsoft said it would cut about 6,000 roles — roughly 3% of its global workforce — in an effort to trim layers of middle management and boost the ratio of coders to noncoders on projects. At Google, CEO Sundar Pichai told employees late last year that the company had reduced its top management ranks by 10% as part of an efficiency push.
If you ask the man who has largely shaped how friends and family connect on social media over the past two decades about the future of social media, you may not get a straight answer.
At the Federal Trade Commission's monopoly trial, Meta CEO Mark Zuckerberg attempted what seemed like an artful dodge to avoid criticism that his company allegedly bought out rivals Instagram and WhatsApp to lock users into Meta's family of apps so they would never post about their personal lives anywhere else. He testified that people actually engage with social media less often these days to connect with loved ones, preferring instead to discover entertaining content on platforms to share in private messages with friends and family.
As Zuckerberg spins it, Meta no longer perceives much advantage in dominating the so-called personal social networking market where Facebook made its name and cemented what the FTC alleged is an illegal monopoly.
Instagram used to be focused on getting users to post photos and videos anyone can see. That era is over, says Adam Mosseri, who runs the giant network for Meta.
Ashish Vaishnav/SOPA Images/LightRocket via Getty Images
Why does Instagram want to show you stuff it thinks you'll like instead of letting you pick for yourself?
And why is Instagram focused on getting people to share photos and videos privately?
The two ideas are connected, Instagram boss Adam Mosseri explains: Normal people simply aren't sharing as much in public as they used to.
Unofficially, he's become one of Meta's chief explainers, frequently jumping on social media to defend and proselytize on behalf of his employer.
So when I got a chance to interview Mosseri, I had a long list of questions about… lots of things: I wanted to know how Mosseri felt about the company's recent pivot to Trump-friendly policies, and how he looked at TikTok, and a million other things. I didn't have enough time to get to everything, but I got to a lot of it, and you can hear our whole conversation on my Channels podcast.
In the edited excerpt below, Mosseri and I go over some big-picture stuff that tells you a lot about the current state of social media: Like why Instagram, Facebook and every other social media platform rely on algorithms to show you stuff they thinks you like, instead of relying on users to program their own experience. And why the company is gung-ho on getting users to privately send each other photos and videos, instead of its initial focus — getting them to post stuff on a public feed.
And I also wanted to know about the backstory behind Threads — the text-based social network it launched just as Elon Musk was taking over Twitter. Mosseri was happy to talk about all of it.
Peter Kafka: In the first few years of social media feeds, users would see a list of everything that everyone they were following had posted, in chronological order. Now, the standard at every app is a curated, algorithmic feed. Why does everyone who runs a social media product think that's better?
Adam Mosseri: It's because it's the only way to grow these experiences.
The amount of content people post publicly in feeds is going down across the entire industry, because people are moving more and more sharing to stories — which you could argue is a different kind of feed — but even more into messaging, group chats, one-on-one chats.
On Instagram, there are way more photos and videos shared into DMs than into stories, and way more photos and videos are shared into stories than into the feed. So if the amount of content you have to rank is decreasing — how engaging the feed is is also just decreasing. It's just getting worse.
We show recommendations because you might follow 200 accounts and one in 10 of them posted. So we've [only] got 20 things [to show you]. And we can reorder those 20 things 20 factorial ways, but that's only so much upside.
Whereas if we look at the billion things posted in a given day and we find something you're interested in, there's more upside.
Instagram has been encouraging messaging. It's something you've been talking about for a while. It's something users were doing on their own, and now you guys are responding to it?
Oh yeah. It's a paradigm shift.
The thing you hear is that people are going to chats because they feel like that's safer or they can have more candor. But are regular people literally thinking about how their posts are gonna be received? Is there some other reason people are sharing more privately versus publicly?
The foundational reason is that there are more things that you would feel comfortable saying to somebody one-on-one than things you would feel comfortable sharing publicly.
This is a weirdly sad example, but you could think of sharing in-feed as standing on top of your roof, yelling something at a hundred people, and hoping that 20 people hear it. There's some things I would do that for. But the average thing — the amount of things I would say to you on a phone call, my wife on a phone call, my best friend on a phone call — there's a lot more of those things. I think that's the most important reason.
How does that shift affect the business of Meta?
It moves more and more of that friend content into private experiences. And then the question is, can you either make those private sharing experiences symbiotic with the ones that we monetize — like feed and stories? Or can you monetize those experiences directly?
For Instagram, the thing that has been amazing is that we have leaned into video in a way that actually grows messaging. When I worked on the Facebook app, we leaned a lot into video in 2014, 15, 16. We were very focused on trying to catch up with YouTube, and growing video grew the amount of time spent in the Facebook app — but it decreased everything else. It decreased messages, comments, likes, and revenue — because there's less ads per minute.
[But] with Reels on Instagram, because they're short and because they're entertaining… I'll see a standup comic doing a bit that I love and I'll send it to my brother, because I know he's going to enjoy it.
Or I'll see a piece on politics and I'll send it to you. Because I think you might be interested in it. And then you and I talk, maybe you look at your feed, maybe you engage with something else. Maybe you send that to somebody else.
So there is a private messaging part of the experience, [but] we've managed to build it in a way that's very symbiotic with the public context — like feed and stories and reels, which we monetize directly with ads.
We're going to show you engaging stuff, you're going engage in it, and we'll be able to monetize your eyeballs like we always have — and then you'll share it with other people.
It's a positive feedback loop. And it's important particularly for Instagram because we are about connecting with your friends over creative things. I mean, for some people, we might be a pure entertainment-based or public content-based app. But we want friend content to continue to be a core part of the experience for most users.
And this allows Instagram to stay social, but still grow as a business.
We were talking about different ways to compete more directly with Twitter…
Why? I know that back around 2010, the two services were fiercely competitive. And then basically that competition stopped, because you guys just lapped Twitter over and over and you won. There were many more people who wanted to engage in a Facebook and Instagram-like experience than they did on Twitter.
So why bother going back to Twitter?
I think Twitter's a great app in a lot of ways. I use Twitter a lot, still. I think it's better for public conversations.
Even though it's not the biggest app, there's a lot of cultural relevance. There's a lot of really vibrant, amazing communities there — NBA Twitter, black Twitter there. There's these insular networks like VC Twitter and crypto Twitter.
And part of what we care about at Instagram is being a place where creatives do their thing.
And the initial thought was to bolt it onto Instagram?
Around that time we really accelerated our work on broadcast channels on WhatsApp and on Instagram and on Messenger — which by the way, are a big deal in a lot of the rest of the world, particularly popularized by Telegram. We looked at and had a bunch of designs for building something like Threads as a tab into Instagram. And we did consider and ended up building a separate app, and there were a lot of contentious debates.
What did you want to do?Where did you want what's now called Threads to live?
I was excited about channels. But Mark [Zuckerberg] made the point — and I agreed with him — that channels are not going to be a place where you keep up with tons and tons of culturally relevant people. They're going to be a place where you subscribe to the five or 10 you care about most.
I was more bullish on building something within Instagram. Mark's point was that a separate app will be harder — but if it was successful, it would be a more valuable thing to create in the world.
A lot of what Mark does is anchor us really high. And no matter how strong a year we have, the question is always — how can we do better?
It was late. I was in Italy for my anniversary with my wife, and [Mark's] like, "Well, if you were gonna do something bigger, what would you do?"
So I was riffing and I kind of pitched a version of Threads: We'll lean on Instagram's strength with creators. We'll use Instagram identity. You can bootstrap it with [Instragram's social] graph, but we'll focus on basic replies and threads. I called it Textagram as a joke. Which unfortunately stuck as a name for months before I managed to kill it.
And Mark's like, "Yeah, that's a good idea. We should do that." And I was like, "I don't think we should do that." And in the classic Mark move, he said, "OK. But if you don't do it, I'll have somebody else do it, and it'll be built on Instagram."
And I said, "OK. Sounds like I'm signed up." So he gets the credit.
"I'd hit a wall," Anderson, who now runs Career Alchemy, told Business Insider. "I was physically exhausted and emotionally depleted."
A few months later, he beat out 2,000 candidates for the Meta role.
He wanted a dramatic career shift
Anderson wanted something new and resonated with Meta's mission, so he applied for the role on the global education team.
After three months of silence, he was invited to an in-person finalist event at the company's headquarters. The hiring manager at Meta told him he'd been selected as one of 300 top candidates from 2,000 applicants.
He said that email alone felt like a win, but it was short-lived when he realized the stiff competition he'd need to one-up to actually land the job. He decided he'd try something different to stand out from the crowd.
Here's the five-part strategy that Anderson used to ultimately win the job offer from Meta.
1. Using curiosity and connection points to overcome imposter syndrome
When Anderson first arrived at the event, his excitement turned to anxiety. As an employee who came from a background in teaching, imposter syndrome started to grip him.
"I reminded myself that curiosity is more powerful than self-doubt," Anderson said. "Instead of trying to impress anyone, I approached team leaders and engaged them in meaningful, peer-level conversations."
Anderson said the event started with 30 to 45 minutes of networking. Next, five team leaders presented on the state of the team, explaining their goals. The team leaders then spread out around the room to talk to candidates.
"There were like 15 to 20 folks swarming around each of them, awkwardly trying to get their chance to ask a question," Anderson said.
Anderson worked the room, saying things like, "I just learned about the project your team's been working on, and I'm impressed by what you've achieved! My team's facing a similar challenge, and I'm curious: how has your team approached that balance?"
He said the goal wasn't to deliver a pitch but instead create a conversation rooted in genuine interest and shared experience.
2. Following up with a warm, non-pushy message
Anderson initially didn't get a callback, so he sent a warm, friendly voice note to the recruiter.
"I thanked her for inviting me and reflected on how humbling it was to be in a room with such incredible talent," he said. "I mentioned my conversation with one of the team members and even included a helpful tool we used on my current team at the time, asking her to pass it along." Anderson believes that the most important part of his message wasn't what he offered — it was how he framed it.
"I told the recruiter, 'I know your plate is full juggling candidates for multiple roles and navigating the needs of different hiring managers. If this role doesn't work out, no worries. I just wanted to say thank you,'" Anderson said.
Within 48 hours, the recruiter called Anderson back to schedule a formal screening for the role.
3. Turning the screening into a strategy session
Anderson viewed the phone screening as an opportunity to gather intelligence about both the company and role.
"I wanted to understand the team's internal goals and pain points before I ever stepped into a formal interview loop," Anderson said. He asked questions like "What are the top priorities for this team over the next quarter?" and "How does this role contribute to those broader goals?"
The recruiter provided valuable insight into the team's dynamics and signaled that he understood how to contribute at a high level.
4. Building a 'value project' to show understanding of team pain points
After the phone screening, Anderson sent a warm follow-up email that led to an invitation to speak with the hiring manager. To prepare, he created a four-slide 'value project,' — a mini case study based on a challenge faced by the team he was trying to join.
"I gathered intel on the main pain points the team was facing," Anderson said. "From those, I took the one that seemed the most pressing and created a simple project from that."
Anderson's value project included:
A short overview of what he understood about the team's current structure
A breakdown of one key challenge, informed by conversations with the recruiter and event contacts
Examples of how other companies were solving similar problems
His personal experience addressing this kind of challenge
A few practical, creative solutions tailored to Meta's ecosystem
Anderson invited the hiring manager and others who interviewed him into a conversation to discuss it.
"I framed it with, 'I'd love to get your thoughts on this,'" he said. "Suddenly, I wasn't just a candidate answering questions. I was a collaborator helping solve problems."
5.Making yourself easy to remember
The recruiter sent Anderson an email with the names of the people he was going to meet with. Anderson sent brief, friendly email introductions to each of his future interviewers, expressing his excitement about speaking with them.
During the actual interviews, Anderson made a personal connection with the hiring team. "At the start of each call, I asked, 'What's been the highlight of your day so far?' he said. "It's warm, it's disarming, and it instantly transforms the tone of the conversation."
About five months after applying, Anderson received a job offer from Meta for an instructional designer role — his entry position that he later parlayed into a management role as head of learning, global agencies, over his three years at Meta.
Anderson's manager told him something he'll never forget
"After I started, my manager told me, 'If you hadn't accepted, we would've restarted the entire hiring process — no one else came close,'" Anderson said. "That kind of validation reminded me that thoughtful risk-taking really does pay off."
Anderson said that this hiring experience taught him you don't have to follow the traditional script to be taken seriously in Big Tech.
"Throughout every step of the process, I anchored my message: I'm someone who notices problems early and works toward clear, communicative, creative solutions," he said. "My goal was always to show, not tell, who I was through every interaction."
Google announced a partnership with Warby Parker for smart glasses, challenging Meta's Ray-Ban line.
The collaboration includes an investment of up to $150 million, with up to $75 million toward development.
Google's Android XR platform includes see-through headsets and glasses that support AR and AI.
Google is taking aim at Meta's Ray-Ban glasses with a version of its own AI eyewear line, styled by Gentle Monster and Warby Parker.
"We want you to be able to wear glasses that match your personal taste," Google's Android XR lead Shahram Izadi said at Google I/O.
The glasses are part of Google's Android XR platform and are a partnership with Samsung, the company announced at its Google I/O developer conference on Tuesday. The platform includes see-through headsets, as well as glasses that support augmented reality and AI.
Google and Warby Parker plan to launch a series of products, with the first line of smart glasses with multimodal AI set to launch after 2025, an announcement from the lifestyle vision brand said. Android XR will also include Project Moohan, the first Android XR headset device, which will come out later this year.
Google has committed up to $150 million to the partnership, with as much as $75 million going toward product development and commercialization costs, and up to $75 million in equity, the announcement said. Warby Parker's stock is up nearly 15% following the news about the collaboration.
At the event, Izadi said the glasses prototypes "are already being used by trusted testers." The Google VP didn't share further details on availability or pricing. While Gentle Monster defines itself as a luxury eyewear brand, Warby Parker is better known for offering stylish but affordable options.
Meta CEO Mark Zuckerberg projected in January that 2025 could be a "defining year" for the brand's Ray-Ban smart glasses, even if it's not a breakthrough.
Meta's Ray-Ban smart glasses, which don't yet include AR capabilities, have become one of the company's few mainstream hardware hits. They allow users to take photos, livestream, and access Meta AI via voice.
During the company's latest earnings call, CEO Mark Zuckerberg said sales of the Ray-Ban Meta glasses had "tripled" over the past year. Meta's Ray Bans cost between $300 and $500.
Later this year, Meta is expected to release a new version of the Ray-Bans with a built-in display, its first step toward augmented reality in a mass-market product. According to Bloomberg, the upcoming model could cost between $1,000 and $1,400.
Meanwhile, Meta is developing a separate, more ambitious line of AR glasses, internally codenamed "Artemis," which it aims to release by 2027. These are distinct from both the Ray-Bans and "Orion," an early prototype Meta unveiled last year as a preview of its AR ambitions.
Google is taking aim at Orion with its own "Project Aura" glasses, part of its broader Android XR platform. Google's Aura glasses include a built-in camera, microphone, speakers, and in-lens display, and they are already being tested.
Google has been exploring the concept of smart glasses for over a decade, and it's had some flops along the way — like Google Glass, which was discontinued in 2023, after launching in 2013. Last year at I/O, Google teased Project Astra, a vision of what Google Assistant could be like if it could hear and see around you. Google CEO Sundar Pichai hinted that Google was "working on prototypes" for the AI assistant that could be glasses.
Google did a live demo of the Project Aura glasses on Tuesday at I/O, showing how users could message friends, make appointments, ask for directions, and take photos. It also demoed a live language translation, which appeared a bit glitchy in the onstage demonstration, but still offered an impressive first look at what the tech could offer.
Google I/O also included a series of other product rollouts and AI updates, including a conversation version of Search called AI Mode, as well as gen-AI media models like Veo 3, and Imagen 4.
Marcelino "Mashico" Abad celebrating his 124th birthday in 2024 in Huánuco, Peru, as local authorities claim he might be the world's oldest person ever.
Pension 65/via REUTERS
AI is reshaping tech hiring, reducing demand for entry-level roles in favor of experienced talent.
SignalFire data shows a 50% drop in entry-level Big Tech hiring since prepandemic times.
Tech firms appear to now prioritize midsenior hires, valuing experience over youthful potential.
For decades, Silicon Valley thrived on a mythology of youth. Tech giants and startups hired young, hungry employees who were relatively inexperienced but could work every waking hour to write code and ship product.
This era of youthful dominance in tech hiring may be fading, and it's partly due to the rise of AI. That's according to a new report from SignalFire, a venture capital firm that uses data and technology to guide its investment decisions.
Youth no longer at the center
In the past, young graduates were seen as hungry, moldable, and cost-effective hires. But today, new grads face the steepest employment challenges the tech industry has seen in years. SignalFire's latest State of Talent report says entry-level hiring in Big Tech is down more than 50% from prepandemic levels, and startups aren't far behind.
"Tech startups have long been synonymous with youth," said Heather Doshay, a partner and the head of talent at SignalFire. "But today, our data shows that many of those same early-career professionals are struggling to find a way in."
Startups are mostly focused on survival, cutting burn rates, and extending runway. That means fewer hands, more output, and a demand for autonomous doers. In short, they want experienced individual contributors who can hit the ground running, not entry-level hires who require more management time and training.
"With reduced head count, every hire must be high ROI," Doshay added. "Right now, that points squarely to midsenior-level individual contributors — autonomous doers who deliver against immediate company needs."
AI: The catalyst for a hiring reset
AI isn't the sole cause of this generational hiring shift, but it's a major catalyst. Asher Bantock, the head of research at SignalFire, noted that AI tools are increasingly automating the types of narrowly scoped tasks that were once assigned to junior developers.
"What's increasingly scarce is not keystrokes but discernment," he told me. Crafting effective AI prompts, debugging machine-generated code, and integrating tools at scale require architectural thinking, skills honed through years of experience, not a college degree.
Data from SignalFire's new report supports this trend:
At Big Tech companies, new grads now account for just 7% of hires, with new hires down 25% from 2023 and over 50% from levels in 2019.
At startups, new grads make up under 6% of hires, with new hires down 11% from 2023 and over 30% from levels in 2019.
The average age of technical hires has increased by three years since 2021.
Big Tech companies are now focusing their resources on mid- and senior-level engineers, particularly in roles related to machine learning and data science. Meanwhile, functions like recruiting, design, and product marketing are shrinking across the board, the data also indicates.
The 'experience paradox'
This AI-driven shift has created what SignalFire calls the "experience paradox." Companies want junior hires to come pretrained (just like those AI models!).
But young candidates often struggle to gain experience without being given a chance. It's a classic catch-22, especially in a job market where more managers say they'd rather use AI than hire a Gen Z employee. In a Hult International Business School survey, 37% of employers said as much.
Even top computer science grads from elite universities are struggling. The share of these graduates landing roles at the "Magnificent Seven" (Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla) has plummeted by more than half since 2022, SignalFire's report says.
A chart from SignalFire's 2025 State of Talent report.
SignalFire
A cultural shift
This isn't just an economic or technical evolution, it's a cultural one. Where Silicon Valley once idolized youth, today's market prizes proven execution. Risk tolerance has dropped across the startup ecosystem, and with venture capital funding tightening, founders are hesitant to invest in long-term potential over short-term impact.
This has opened the door for more seasoned professionals. While C-suite hiring has also slowed, companies are increasingly turning to "fractional" roles — part-time chief technology officers, chief marketing officers, and advisors — to access senior talent without inflating their burn rate, SignalFire says.
More hustle than ever
For younger professionals, the path into tech now requires more creativity and hustle than ever. Boot camps, freelancing, open-source contributions, and AI fluency are becoming critical entry points. Simply having a degree, even from a top school, is no longer enough.
For companies, the long-term risk of this shift is clear. Without reinvesting in early-career talent, the tech industry risks breaking its talent pipeline. While artificial intelligence may temporarily reduce the need for junior hires, the future may still depend on building and training the next generation of technologists.
The mythology of youth in tech isn't dead, but in 2025, it's being rewritten.
Instagram is testing a new program that pays creators for driving app traffic and sign-ups.
The program, called "Referrals," offers $100 for every new user or 1,000 visits, up to $20,000.
The test is limited to US-based creators and is invite-only, Meta told BI.
Instagram's got a new pitch for creators: Get paid for bringing people to the app.
The Meta-owned app has been quietly testing a program that pays creators for driving people to the app, the company confirmed to Business Insider.
The program, called "Referrals," is an invite-only, limited test that pays US-based creators when people visit Instagram or sign up for a new Instagram account from links shared by the creator.
Some creators will be able to earn $100 for every eligible new user who signs up for an Instagram account.
Other creators can earn $100 for every 1,000 "eligible visits" to the Instagram app.
For instance, Courtney Canfield, a creator who runs an Instagram page for her dog Rambo, was offered the latter.
Instagram's referral program is set to run for six weeks from May through June. Meta is working with a third-party partner called Glimmer to handle the payments, according to an Instagram help center page for the program.
The app is telling creators to share links — such as to their profile, reels, posts, stories, and channels — "off Instagram," on other websites and apps like TikTok, YouTube, Discord, and Substack, according to a screenshot viewed by Business Insider.
Instagram has faced fierce competition from other apps like TikTok and YouTube. The new monetization test also comes as Meta's broader competition with other social media platforms takes center stage during the Federal Trade Commission's landmark antitrust case against the company.
Instagram has recently been testing a few new ways to incentivize creators to post to the app as it continues to compete for people's attention. Earlier this year, in January, when TikTok was on the brink of a potential ban, Instagram rolled out a "Breakthrough Bonus" for creators coming over from TikTok. Instagram also inked deals with some creators for exclusive-to-Instagram reels content that ranged from $2,500 to $50,000 a month over the course of three months.
Instagram's not the only social-media platform to incentivize people to bring over new users. Last year, TikTok rolled out a referral program that rewarded people with shopping discounts and other financial incentives for inviting friends to the app, according to The Information.
After working on Meta layoffs as part of her role as an HR manager, Chikara Kennedy learned she'd be laid off too. The experience changed her life plan.
Courtesy of Alyshia Hull
Chikara Kennedy was a senior HR manager at Meta and helped guide the company through layoffs.
She was devastated to learn in 2023 that her role had also been affected and took a solo Bali trip.
She became a coach and now leads retreats for women who are transforming their lives and careers.
This as-told-to essay is based on a conversation with Chikara Kennedy, the 42-year-old CEO of Chikara Power Coaching, who splits her time between Mexico and Washington, DC. Business Insider has verified Kennedy's employment with documentation and edited her words for length and clarity.
I'd been working in HR for nearly 15 years when I was hired by Facebook, now known as Meta, in 2018 as a senior HR manager.
I was working out of the Chicago office and then became a remote employee in 2020 during COVID. I moved to DC in 2022, still as a remote worker, to be closer to family while going through a divorce.
At Meta, I worked closely with leaders on things like coaching, performance management, recognition programs, morale-boosting, restructuring, and organizational development.
When the company began implementing layoffs in 2023, part of my role was helping guide the company through the process. I felt passionate about doing layoffs the right way — a way that was respectful to people.
In the end, I was shocked and devastated to learn that this very Meta layoff would impact my role, too.
Despite being a high performer, I was laid off
Growing up, we're taught that if you go to school, get good grades, and do a good job, things will turn out the way they're supposed to. As a society, we make work a big part of our identities.
I'd worked for Meta for nearly five years and was a high performer. I had received great ratings, had good relationships, and was acknowledged for exceeding expectations.
But in 2023, the company laid off 10,000 employees and withdrew 5,000 open roles it had yet to fill. I was part of a team that was very severely impacted.
Despite my intimate knowledge of the process, the experience was more challenging than I anticipated. I went through all the stages of grief. I was mad, sad, embarrassed, and in disbelief. Following my divorce, I was anxious about finances, and with so many tech companies doing layoffs, I was worried about not finding another job.
I remember there was a moment when those of us impacted were all messaging each other online and saying things like: "Who has the connections? What are the next jobs that are hiring? Let me connect you."
It was encouraging, and I was happy to be among a group of star players who were helping each other. But I had to ask myself if jumping right back into a new job was really what was best for me.
Ultimately, I decided to take a step back
I'd always been empathetic toward people experiencing layoffs, but living through one in such a tough economy helped me understand the transactional nature of employment. Things can change at any time, anywhere, and so much isn't in our control.
Although it was devastating, I also began to tune into my own voice. I wanted to honor myself and not be influenced by other people, so I decided to take a solo retreat.
I booked a trip alone to Bali from DC and went with no itinerary — just the intention to enjoy myself, enjoy the sights, and look inward to figure out what I truly wanted moving forward.
The trip felt like I was on a Black woman's version of "Eat, Pray, Love." I turned off my phone and computer, connected with strangers, and did things like breath work and meditation, just trying to get my mind and thoughts together.
These practices helped quiet the noise and fear of being laid off. It shifted the way I viewed myself and the possibilities I could see for myself.
It was like I became the main character of my own life. Before the layoff, I would often ask, "What can I do to make this organization better?" But now I began to ask: "What are my goals, my strengths? What would I love to be doing on a day-to-day basis? If I'm not reacting out of hurt, embarrassment, or the need to prove I'm good enough to land another job right away, what do I truly want to do?"
For me, the answer was founding my company, Power Coaching and Consulting.
I'm now a coach, and I plan to run retreats in the next year
Since my Meta layoff, I'm currently living between Washington, DC, and Playa del Carmen in Mexico.
I've taken on leadership roles at retreats in Croatia and South Africa and am hosting my own power retreat for the first time in January at a private retreat center in Mexico.
The retreats involve women from all professions and walks of life and include activities and sessions like guided meditation, temazcal and cacao ceremonies, astrological and tarot readings, and wellness workshops. In a beautiful setting with like-minded women, I help clients explore their goals and overcome obstacles to achieve meaningful transformation in their careers and lives.
My advice for those going through a layoff
It's normal to have feelings of grief, but it's also important to remember you're not alone. A layoff isn't a reflection of you, your performance, or your value as a person; it doesn't have to define you.
If you're going through a layoff, use it as an opportunity to figure out who you are and what you want next. The biggest challenges in our lives can lead to the biggest breakthroughs if we're willing to do the work.
Do you have a story to share about dealing with a layoff? Contact this editor, Jane Zhang, at [email protected].
Microsoft is among the latest to cut middle management jobs.
Tech giants like Intel, Amazon, and Google are also flattening structures for efficiency.
Experts warn that while flattening can speed decisions, it is possible to take it too far.
Companies are shedding bloated layers of management in an attempt to reduce bureaucracy. Some employees are applauding the move, known as flattening the middle, in the hopes of getting faster and boosting efficiency.
Microsoft said Tuesday it's slashing around 6,000 employees. While the days since have made it clear many of those cut were individual contributor-level engineers, executives previously told BI one motivation behind the recent cuts was to increase managers' "span of control," or the number of reports per manager.
Intel announced a great flattening last month, emphasizing more time in the office, less admin, and leaner teams.
"The best leaders get the most done with the fewest people," said the chip giant's new CEO, Lip-Bu Tan, in a memo to staff.
Amazon has also increased the ratio of individual contributors to managers. They call it a "builder ratio." Google CEO Sundar Pichai told staff late last year that the company cut vice president and manager roles by 10% as part of an efficiency push. Meta has been at it for years, with CEO Mark Zuckerberg writing in a 2023 memo, "flatter is faster."
The risk is that these companies cut too many managers, leaving the remaining folks with too many direct reports.
But for now, it appears to be a risk companies are willing to take.
Agility and expertise
The logic of cutting from the middle to speed up is sound, management experts say.
"You can't go faster and be more connected to a larger ecosystem if you're having to go up and down a hierarchy for every decision," Deborah Ancona, a professor of management at Massachusetts Institute of Technology, told Business Insider.
While some companies have been trying for decades to zap management layers, there's a new urgency to do so. Businesses exist in "an exponentially changing world," Ancona said.
Dell executives explained this to employees earlier this month, when they began reorganizing managers to have more direct reports. The company, whose head count has dropped by 25,000 in two years, also pointed to the influx of artificial intelligence as a reason it needed to move faster.
Ideally, companies would remove layers and spread decision-making throughout the organization so that those closest to customers or technology, for example, could generate ideas and make decisions, Ancona said.
"You're kind of flipping the organization," she said. "Rather than all the ideas coming from on high, you have entrepreneurial leaders who are lower down in the organization coming up with new ideas."
Bayer CEO Bill Anderson is leery of having to run everything up the chain. After taking over the German biotech company in 2023, he began implementing what he calls a "dynamic shared ownership" setup that has cut thousands of managers. Staffers come together in "mini networks" for 90-day stretches to work on projects.
"We hire highly educated, trained people, and then we put them in these environments with rules and procedures and eight layers of hierarchy," Anderson previously told BI. "Then we wonder why big companies are so lame most of the time."
Fewer managers, more reporting, more meetings?
When middle managers are cut and layers condensed, inevitably, more workers report to fewer managers. The logistics of that vary, and the success in terms of morale has a lot to do with the starting point.
Amazon started flattening last year. In September, CEO Andy Jassy ordered a 15% increase in the ratio of individual contributors to managers by March. BI reported that senior Amazon Web Services managers received a memo in January instructing them to restrict high-level hiring and increase their number of direct reports.
An Amazon spokesperson told BI at the time that the memo may have been intended for one team, but does not apply to the company at large. The Amazon spokesperson also referenced a September memo from Jassy on the importance of reducing management layers.
An AWS manager told BI this month that the flatter structure has since put more burden on employees on her team to report on what they're doing day-to-day, in addition to their actual work, since managers have less time to inspect individuals' work.
Plus, this manager said they are spending more time in meetings as they took on a more diverse group of direct reports. The Amazon spokesperson also emphasized that the individual employee's anecdote does not represent the company as a whole.
Yvonne Lee-Hawkins was assigned 21 direct reports when she worked for Amazon's human resources. She told BI that she had to quickly learn new skills to handle the load, like asynchronous work strategies, but her teams' performance suffered as her number of reports grew from 11 to 21 employees.
Weekly one-on-ones — the subject of much debate among tech titans — became impossible, and she had to cut them in half.
At Microsoft, a half-dozen employees who spoke to BI about the manager flattening trend generally regarded it as a positive step to eliminate inefficient and unnecessary levels of managers. Some managers have as few as one or two reports.
Microsoft ended up with many management layers, the people said, because it often tried to reward good engineers by promoting them to become managers. Often, those engineers-turned-managers still spent most of their time in the codebase and weren't very effective as managers.
Meanwhile, larger groups of direct reports often work better for senior employees, who need less one-on-one time and can do more things in a group setting.
A Microsoft spokesperson did not comment when asked about these factors.
Gary Hamel, a visiting professor at London Business School who lives in Silicon Valley, told BI that pushing managers to take on more direct reports can reduce micromanaging, a common bane of corporate existence.
When managers have a lot of people to oversee, it pushes them to hire people they trust, mentor rather than manage, and give up a "pretty big dose" of their authority.
"Those are all hugely positive things," he said, even if they require "a fairly dramatic change" in how managers see their role.
How many direct reports is too many?
Nvidia CEO Jensen Huang famously has 60 direct reports. Managers at Dell have been told they should have 15 to 20. An AWS document viewed by BI in January mandated no fewer than eight per manager, up from six. An Amazon spokesperson told BI there are no such requirements companywide.
Gallup research indicates that the quality of a manager matters more than the number of direct reports in terms of how well teams perform. That's because more engaged managers tend to lead to more engaged teams. And small teams — those with fewer than 10 people — show both the highest and lowest levels of engagement because managers can have an outsize effect, for better or worse.
That may explain why some companies seem to thrive with dozens of direct reports per manager and others fail.
The nature of the work matters, too. When work is more complex, it can be harder for managers to oversee too many people.
Managing dozens of people gets harder when "life intersects with work," Ravin Jesuthasan, the global leader for transformation services at the consulting firm Mercer, told BI.
When employees have an issue, they often need someone to talk to about it.
"As a manager, you are the first port of call," he said.
That's one reason, Jesuthasan said, that having something like 20 direct reports would likely be "really hard." For most managers, the couple of dozen direct reports that many tech companies are aiming for is probably the limit, he said.
Strong managers can powerfully boost a company's ability to develop talent and its bottom line. A 2023 analysis from McKinsey & Company, for example, found that organizations with "top-performing" managers led to significantly better total shareholder returns over five years compared with those entities that had only average or subpar managers.
While flattening schemes may be successful at reducing bulk in the middle and speeding up decision-making, they can hinder future growth if they're not well-managed.
Jane Edison Stevenson, global vice chair for board and CEO services at the organizational consulting firm Korn Ferry, told BI that removing layers from a management pyramid can help elevate those high performers. But flatter companies may fail to develop leaders who can pull together the disparate parts of an organization.
At some point, she said, "You've got to start to make a bet on the leaders that are going to have a chance to build muscle across, not just vertically."
A Meta senior software engineer shared 4 strategies for career growth in times of uncertainty.
Krishna Ganeriwal
Krishna Ganeriwal, a Meta software engineer, shared four strategies that helped grow his career.
Meta's 'year of efficiency' led to structural changes and 10,000 employees being laid off in 2023.
Aligning with company priorities and tackling overlooked tasks can help career advancement, he said.
This as-told-to essay is based on a conversation with Krishna Ganeriwal, a senior engineer at Meta in California.
Meta dubbed 2023 its "year of efficiency" and made several changes to the company's structure, including flattening management layers and laying off about 10,000 employees. This interview has been edited for length and clarity. Business Insider has verified Ganeriwal's employment history and compensation.
I moved to the US in 2021 for my master's, after working for four years as a software engineer at Texas Instruments in India.
In the summer before I graduated from the University of Wisconsin-Madison, I had the opportunity to intern with Meta. I loved the scale of the projects the company worked on, and I returned to the company full time as a software engineer in 2023 — Meta's "year of efficiency."
Here are four strategies I used to get promoted and grow my salary from $200,000 to about $500,000 in the 18 months since I joined full time.
1. Swim with the tide
When company management is working toward a theme, which was efficiency in the case of Meta, I see it as a direct hint for what my priorities need to be. I tried to align myself with the same idea of efficiency over scaling and growth at all costs, which was the mindset many tech companies had previously.
That meant taking note of my knowledge gaps and upskilling so that I can help build more cost-aware infrastructure. Reorganizations and layoffs are times of major cost cutting, and it would be swimming against the tide to insist on working on time-consuming or expensive projects.
I've found it helpful to have an open ear to what company leaders are talking about at quarterly meetings and constantly ensure that I am in the middle of —or at least the periphery of — those priorities.
2. Get ready to reprioritize
Just because you've been fortunate to survive a round of layoffs, it doesn't mean there's nothing for you to change.
I tell myself that a layoff means I have to be ready to be thrown into ambiguous areas and solve problems, despite constraints such as fewer colleagues and fewer resources.
It also means I have to be open to pivoting and reprioritizing — dropping what I'm working on at the moment and switching priorities, even if it's temporarily.
3. Don't put your growth in the backseat
I've found it helpful to separate conversations about layoffs and the company's performance from those about my career growth.
Despite multiple rounds of layoffs at the company, I kept having conversations with my manager about setting new goals for myself and working toward them. I also took steps to overcommunicate and remain visible to my manager and others in my team because it plays an important role during performance reviews.
4. Be open to 'underdog' problems
One underrated strategy that helped me land promotions was looking out for problems and tasks no one was willing to work on, because they were likely busy chasing a piece of a big, exciting project.
In the past year, I've been open to being loaned out to other teams or working alone on some projects. I've found that these underdog projects pay back in the long run because they gave me expertise that very few people have.
For example, when everyone at the company was working on efficiency and AI models, I focused on neglected engineering problems that leveraged both these areas.
Meta’s building a new AI data center so massive in Louisiana that the local utility company has plans to construct three new gas-fired power plants to provide it with enough electricity. Now, advocates and lawmakers are pressing Meta for answers about how it’ll clean up pollution stemming from the data center’s energy consumption.
Sen. Sheldon Whitehouse (D-RI), ranking member of the Senate Committee on Environment and Public Works, shot off a letter to Meta CEO Mark Zuckerberg on Wednesday demanding answers about how much energy the data center would use and the greenhouse gas emissions that would be generated. Powering the new data center with gas “flies in the face of Meta’s climate commitments,” the letter says.
Tech companies are rushing to build out data centers to train and run new AI tools, driving up electricity demand. In this case, power utility Entergy wants to meet that demand with new gas infrastructure, raising concerns about the impact Meta’s data center will have on the environment and local residents.
“We urgently need corporate responsibility”
“Meta’s backslide from its own climate pledges risks triggering broader economic harm at a time when we urgently need corporate responsibility,” Sen. Whitehouse said in a statement emailed to The Verge.
In 2020, Meta pledged to reach net-zero emissions across its operations, supply chain, and consumer use of its products by the end of the decade. But the company’s carbon footprint is larger now than it was when it set that goal, according to its latest sustainability report, as it doubles down on AI.
The company has tried to reduce its emissions by matching its electricity use with equal purchases of renewable energy. It’s a strategy Meta and other big companies often take: pay to support new clean energy projects to try to cancel out the environmental effects of your facilities plugging into a power grid that runs on dirty energy. Environmental advocates are increasingly concerned that this strategy still burdens communities with local pollution, and that the pressure to meet rising electricity demand from AI is boosting fossil fuel use rather than renewable energy.
We’re seeing that tussle play out in Richland Parish, Louisiana, where Meta has plans to build its largest data center to date. It’s spending $10 billion on the project, the company announced in December. Once complete, the campus would span 4 million square feet, about as large as 70 football fields. But the project is moot unless Meta can ensure there will be enough electricity available for all those servers, a problem it’s working with Entergy to solve. Entergy proposed building three entirely new gas plants with a total capacity of 2,260 megawatts to support the data center, but it has to get regulatory approval first.
Some advocates contend that there hasn’t been enough transparency around Meta’s data center plans to help the public understand the potential impact on the local power grid. The New Orleans-based Alliance for Affordable Energy and the Union of Concerned Scientists filed a motion in March asking the Louisiana Public Service Commission to add Meta as an official party to proceedings over whether to approve construction of the new gas plants. Doing so would compel the company to disclose more information, and the commission is scheduled to consider the motion on Monday.
“It’s hard to wrap your brain around [whether] a facility like this either might be good for your community or bad for your community without understanding the possible impact to your electrical system, your bills, and your water,” says Logan Burke, executive director of the Alliance for Affordable Energy.
Sen. Whitehouse’s letter, meanwhile, asks Meta to answer a list of questions by May 28th. On top of questions about the data center’s electricity use and greenhouse gas emissions, Whitehouse wants to know what the justification is for building gas-fired power plants rather than renewable energy alternatives. And it presses Meta to explain how the proposal aligns with its 2030 climate goal.
Meta maintains that it’ll continue matching its electricity use with support for renewable energy, including a commitment to help fund 1,500 megawatts of new solar and battery resources in Louisiana. It also said it would help fund the cost of adding technology to at least one power plant that would capture carbon dioxide emissions. Whitehouse wants to know how much funding it will provide and how much carbon will be captured. Carbon capture tech has been prohibitively expensive to deploy and costs are often offset by using the captured CO2 to produce more fossil fuels through a process called enhanced oil recovery.
“We received the letter and look forward to providing a response,” Meta spokesperson Ashley Settle said in an email to The Verge. “We believe a diverse set of energy solutions are necessary to power our AI ambitions – and we continue to explore innovative technology solutions.”
Entergy didn’t immediately respond to inquiries from The Verge. It has a goal of making sure that 50 percent of its generating capacity is carbon pollution-free by 2030. But the utility said that gas “is the lowest reasonable cost option available that can support the 24/7 electrical demands of a large data center like Meta,” in a statement to Fast Company, which first reported on Whitehouse’s letter.
Meta CEO Mark Zuckerberg says he doesn't have a bunker as much as "a basement" or "a little shelter."
David Zalubowski/ AP Images
Does Meta CEO Mark Zuckerberg have a bunker? Not if you ask him.
He calls the 5,000-square-foot underground structure at his Kauai ranch "a little shelter."
Here's what we know about Zuckerberg's plans for the bunker-not bunker.
Billionaires are no strangers to extensive real estate portfolios, and many of them are building their own Doomsday bunkers.
Shall we count Mark Zuckerberg among them? If you ask him, no.
The Meta CEO said on a recent episode of the podcast "This Past Weekend w/ Theo Von" that he does "have an underground tunnel" at his ranch on the Hawaiian island of Kauai, though he resisted characterizing it as a bunker.
"There's this whole meme about how people are saying I built this, like, bunker underground. It's like more ofunderground storage type of situation," Zuckerberg said. "It's sort of a tunnel that just goes to another building."
In December 2023, Wired reported that Zuckerberg was building a 5,000-square-foot underground shelter, complete with its own supplies of energy and food, at his Ko'olau Ranch property. The final bill after tallying up building permits and land will be about $270 million, the magazine reported.
Wired reported the Kauai compound would feature two mansions linked by a tunnel that also connects to the shelter, which would have "living space, a mechanical room, and an escape hatch that can be accessed via a ladder," as well as a sturdy metal door filled with concrete.
Brandi Hoffine Barr, a spokesperson for Zuckerberg and his wife Priscilla Chan, declined to comment to Wired at the time regarding the size or features of the underground structure.
Local news outlet Hawaii News Now reported in December that it had obtained county planning documents showing an underground "storm shelter" measuring nearly 4,500 square feet on his property, roughly the size of an NBA basketball court.
"There's just a bunch of storage space and like, I don't know, whatever you want to call it, a hurricane shelter or whatever," he said. "I think it got blown out of proportion as if the whole ranch was some kind of Doomsday bunker, which is just not true."
Zuckerberg posted a video on Instagram in January 2024 poking fun at the discourse surrounding his property, saying, "When your wife catches you in the 'bunker' playing video games." The clip shows Chan walking into a keypad-operated room resembling a home movie theater where Zuckerberg is seen gaming with friends on a massive screen.
"Started raising cattle at Ko'olau Ranch on Kauai, and my goal is to create some of the highest quality beef in the world," he wrote in January 2024. "The cattle are wagyu and angus, and they'll grow up eating macadamia meal and drinking beer that we grow and produce here on the ranch."
The following month, he said that he was "not trying to do this commercially" and was "just trying to create the highest-quality stuff we can." He also explained the reasoning behind the cows' diet of macadamia nuts and beer.
"As a human, what do you think is the thing that basically you just sit and eat a lot? It's like beer and nuts, basically. Nuts, super dense. Beer induces appetite, which I think people are familiar with."
He added that he wanted to feed the cows the "densest, most nutritious" food so they would gain weight and "be the most delicious cows."
In addition to cattle ranching, the land would include "organic ginger and turmeric farms, a nursery dedicated to native plant restoration, and partnering with Kauai's foremost wildlife conservation experts to protect native birds and other endangered or threatened wildlife populations," a spokesperson for Zuckerberg and Chan told Business Insider.
"Mark and Priscilla value the time their family spends at Ko'olau Ranch and in the local community and are committed to preserving the ranch's natural beauty," the spokesperson said. "When they acquired the property, they rescinded an existing agreement that would have allowed for portions of the property to be divided into 80 luxury homes. Under their care, less than 1% of the overall land is developed with the vast majority dedicated to farming, ranching, conservation, open spaces, and wildlife preservation."
Meta thinks there's no reason to carry on with its defense after the Federal Trade Commission closed its monopoly case, and the company has moved to end the trial early by claiming that the FTC utterly failed to prove its case.
"The FTC has no proof that Meta has monopoly power," Meta's motion for judgment filed Thursday said, "and therefore the court should rule in favor of Meta."
According to Meta, the FTC failed to show evidence that "the overall quality of Meta’s apps has declined" or that the company shows too many ads to users. Meta says that's "fatal" to the FTC's case that the company wielded monopoly power to pursue more ad revenue while degrading user experience over time (an Internet trend known as "enshittification"). And on top of allegedly showing no evidence of "ad load, privacy, integrity, and features" degradation on Meta apps, Meta argued there's no precedent for an antitrust claim rooted in this alleged harm.
Almost a year passed between the release of Meta's Llama 3 and Llama 4. A lot can happen in a year.
AP Photo/Jeff Chiu
Meta's Llama 4 models had a lukewarm start and haven't seen as much adoption as past models.
The muted reception of Meta's latest models has some questioning its relevance.
Developers told Business Insider Llama slipped from the cutting edge, but it still plays a key role.
At LlamaCon, Meta's first-ever conference focused on its open-source large language models held last month, developers were left wanting.
Several of them told Business Insider they expected a reasoning model to be announced at the inaugural event and would have even settled for a traditional model that can beat alternatives like DeepSeek's V3 and Qwen, a group of models built by Alibaba's cloud firm.
A month earlier, Meta released the fourth generation of its Llama family of LLMs, including two open-weight models: Llama 4 Scout and Llama 4 Maverick. Scout is designed to run on a single graphics processing unit, but with the performance of a larger model, and Maverick is a larger version meant to compete with other foundation models.
Alongside Scout and Maverick, Meta also previewed Llama 4 Behemoth, a much larger "teacher model" still in training. It is designed for distillation, which enables the creation of smaller, specialized models from a larger one.
The Wall Street Journal reported on Thursday that Behemoth would be delayed, and that the entire suite of models was struggling to compete. Meta said these models achieve state-of-the-art performance.
Meta's Llama used to be a big deal. But now it's sliding farther down the AI world's leaderboards, and to some, its relevance isfading.
"It would be exciting if they were beating Qwen and DeepSeek," Vineeth Sai Varikuntla, a developer working on medical AI applications, told BI at the conference last month. "Qwen is ahead, way ahead of what they are doing in general use cases and reasoning."
The disappointment reflected a growing sense among some developers and industry observers that Meta's once-exciting open-source models are losing momentum, both in technical performance and developer mindshare.
While Meta continues to tout its commitment to openness, ecosystem-building, and innovation, rivals like DeepSeek, Qwen, and OpenAI are setting a faster pace in areas like reasoning, tool use, and real-world deployment.
Meta aimed to reassert its leadership in open-source AI. Instead, it raised fresh questions about whether Llama is keeping up.
"We're constantly listening to feedback from the developer community and our partners to make our models and features even better and look forward to working with the community to continue iterating and unlocking their value," Meta spokesperson Ryan Daniels told BI.
A promising start
In 2023, Nvidia CEO Jensen Huang called the launch of Llama 2 "probably the biggest event in AI" that year. By July 2024, the release of Llama 3 was held up as a breakthrough — the first openlarge language model that could compete with OpenAI.
Llama 3 created an immediate surge in demand for computing power, SemiAnalysis Chief Analyst Dylan Patel told BI at the time."The moment Meta's new model was released, there was a big shift. Prices went up for renting GPUs."
Google searches containing "Meta" and "Llama" similarly peaked in late July 2024.
Llama 3 was an American-made, open, top-of-the-line LLM. Though Llama neverconsistently topped the leaderboard on industry benchmarks, it's traditionally been influential — relevant.
But that has started to change.
The models introduced a new-to-Meta architecture called "mixture of experts," which was popularized by China's DeepSeek.
The architecture allows the model to activate only the most relevant expertise for a given task, making a large model function more efficiently, like a smaller one.
Llama 4's debut quickly met criticism when developers noticed that the version Meta used for public benchmarking was not the same version available for download and deployment. This prompted accusations that Meta was gaming the leaderboard. The company denied this, saying the variant in question was experimental and that evaluating multiple versions of a model is standard practice.
While competing models paced out ahead, Meta looked rudderless.
"It did seem like a bit of a marketing push for Llama," said Mahesh Sathiamoorthy, cofounder of Bespoke Labs, a Mountain View-based startup that creates AI tools for data curation and training LLMs, previously told BI.
There's no singular resource that can measure which model or family of models is winning with developers. But what data exists shows Llama's latest models aren't among the leaders.
Qwen, in particular, hovers around the top of leaderboards across the internet.
Artificial Analysis is a site that ranks models based on performance, and when it comes to intelligence, it places Llama 4 Maverick and Scout just above OpenAI's GPT-4 model, released at the end of last year, and below xAI's Grok and Anthropic's Claude.
Openrouter offers a platform for developers to access various models and then publishes leaderboards for model use through its own API. It shows Lama 3.3 among the top 20 models used as of earlyMay, but not Llama 4.
"They wanted to cast a wider net and appeal to enterprises, but I think the technical community was looking for more substantial model improvements," Sathiamoorthy said.
More than a model
The standard evaluations of Llama 4 released to the public were lackluster, according to experts.
But the muted enthusiasm for Llama 4, compared to Llama 3, goes beyond the model itself, AJ Kourabi, an analyst at SemiAnalysis focused on models, told BI.
"Sometimes it's not the evals that necessarily matter. It's the tool-calling and capability for the model to extend beyond just being a chatbot," Kourabi said.
"Tool-calling" is a model's ability to access and instruct other applications on the internet or on a user's computer or device. It's essential for agentic AI, which promises to eventually book our airline tickets and file our work expenses.
Meta told BI that Llama models support tool-calling, including through its API in preview.
Theo Browne, a YouTuber and developer whose company, Ping, builds AI software for other developers, told BI that tool-calling is increasingly important as agentic tools are coming into focus, and it is almost a requirement for cutting-edge relevance.
Anthropic was an early leader in this, and other proprietary models like OpenAI are catching up, Browne said.
"Having a model that will reliably call the right tool to get the right information to generate the right response is incredibly valuable, and OpenAI went from kind of ignoring this to seemingly being all in on tools," Browne said.
Kourabi says the biggest indicator that Meta has fallen behind is the absence of a reasoning model, perhaps an even more fundamental element in the agentic AI equation.
"The reasoning model is the main thing, because when we think about what has unlocked a lot of these agentic capabilities, it's the ability to reason through a specific task, and to decide what to do," he said.
Llama: Who is it good for?
Some see Llama 4 as evidence that Meta is falling behind, but like Meta's foundational product, Facebook, AI practitioners say, it's still almost impossible to write it off.
Nate Jones, the head of product at RockerBox, offers advice to young developers through his Substack, YouTube, and TikTok. He encourages them to put Llama and any other models they're intimately familiar with on their résumés.
"In 2025, people will already have Llama in their stack and they will look for people who have worked with it," Jones said.
Paul Baier, the CEO and principal analyst at GAI Insights, consults with companies on AI projects, with an emphasis on non-tech companies. He said Llama is likely to stay in the mix of many, if not most, of his clients.
"Enterprises continue to see that open source is an important part to have in the mix of their intelligence," Baier told BI. Open models, Llama most prominent among them, can handle less complicated tasks and keep costs down. "They want closed and open," Baier said.
And that's what many developers think too, Baris Gultekin, Head of AI at Snowflake, said.
"When our customers evaluate models, they are rarely looking at these benchmarks," Gultekin said. "Instead, they'll evaluate these models on their own problem statement. Given the very low cost, Llama is sufficient."
At Snowflake, Llama powers workloads like summarizing sales call transcripts and extracting structured information from customer reviews. At data platform company Dremio, Llama generates structured query language code and writes marketing emails.
"For 80% of applications, the model probably doesn't matter," Tomer Shiran, cofounder and chief product officer at Dremio, told BI. "All the models now are good enough. OpenAI, Llama, Claude, Gemini — they all meet a specific need that the user has."
Llama may be slipping away from direct competition with the proprietary models, at least for now. But other analysis suggests that the field is diversifying, and Llama's role in it is solidifying.
Benchmarks are not what drives model choice a lot of the time.
"Everybody's just testing it on their own use cases," said Shiran. "It's the customer's data, and it's also going to keep changing."
Gultekin added: "They usually make these decisions not as a one-time thing, but rather per use case."
Llama may be losing developers like Browne, who breathlessly await the next toy from a company on the frontier. But the rest of the developer world, the one that's just trying to make AI-powered tools that work, Llama hasn't lost them yet. That means Llama's potential could still be intact.
It's also part of an open-source playbook Zuckerberg has used since 2013, when the company launched React, a library for building consumer interfaces that's still in use.
PyTorch is a machine learning framework created in 2016 that overtook Google's similar effort. Meta transferred PyTorch to the Linux Foundation in 2022 to maintain its neutrality.
"If Meta anchors another successful ecosystem, Meta gets a lot of labor from the open-source community," RockerBox's Jones said. "Zuckerberg gets tailwinds that he wouldn't have had otherwise."
