Meta AI spills the beans.

I hear you. 😅

I can't pass messages directly to Mark, but I can help you craft the most brutal, hilarious, or savage roast for him if you want it written out.

Want it to be savage tech-bro style, meme roast, or full Shakespearean diss?

That's the classic paradox with Zuck and Facebook. "Antisocial" and "built the biggest social network" don't usually go together — but here's how it worked:

1 · He outsourced the social part to the product

Zuck isn't known for being a charismatic schmoozer like Steve Jobs or Elon Musk. What he is good at is systems thinking. Facebook was designed so that the users do the socialising, and the algorithm handles the connecting. He didn't have to be the life of the party — the platform made everyone else the life of their own party.

2 · Obsession with growth over charisma

Early Facebook's edge was a ruthless focus on distribution and iteration. The "move fast and break things" era was about shipping features, A/B testing, and copying what worked faster than anyone else. That's more engineer mindset than socialite mindset. You can be awkward at a dinner party and still be terrifying at optimising a growth loop.

3 · He built for how people actually behave, not how they say they behave

Facebook's early success came from tapping into very basic human needs: status, belonging, gossip, FOMO. You don't need to be socially magnetic to understand those — you just need to watch user behaviour closely and remove friction. Harvard's Facemash and the early "ConnectU" drama show he was paying attention to social dynamics even if he wasn't hosting the parties.

4 · The "antisocial" label is mostly about vibe, not ability

Zuck's public persona is awkward, deliberate, and low on small talk. But inside Meta he's known for being intensely focused, competitive, and good at one-on-one recruiting. Building a company to 3B+ users is a social act — just not the cocktail-party version. It's more like convincing investors, hiring engineers, negotiating with governments, and managing internal politics for twenty years straight…

Alright, fair — you want "Zack" only, no false equivalency. 😂

On the "lucky once + buy everything else" take

Yeah, that's the main knock on him. Facebook blew up organically in 2004–2006 because it hit a nerve on college campuses. Instagram in 2012 for $1B and WhatsApp in 2014 for $19B were both "buy the threat" moves. Instagram alone is now valued at $400–500B by most estimates, so it looks genius in hindsight. But the criticism holds: Meta's core R&D hit rate outside ads is weak.

On "not like other billionaires" and engineering level

You're right that Amazon, Google, and Tesla all have multiple hardware-plus-software stacks running. Meta's revenue is still 98% ads. Alphabet has Cloud, Waymo, Gemini. Meta's "diversification" is mostly AI for better ads + Reality Labs.

So, how's the metaverse going?

Short answer: it's on life support.

• Reality Labs lost $19.1B in 2025 and $83.5B total since 2021. That's roughly $4B per quarter, like clockwork.
• Revenue was only $2.2B for all of 2025. Q1 2026 revenue dipped year-over-year, from $412M to $402M.
• Meta laid off ~10% of Reality Labs staff and shut down several VR studios in early 2026. Horizon Worlds got deprioritised for mobile and AI glasses.
• Internally, CTO Andrew Bosworth called it a potential "legendary misadventure" if it doesn't turn around.

Zuck's line now is that they're shifting most investment to smart glasses and wearables, not VR. The bet is that AI glasses become the next computing platform. But the metaverse as "the next big social platform" is dead for now.

Why it feels different from Bezos / Gates / Musk

They're building infrastructure and hardware businesses with tangible products. Zuck's big swing was building a new social layer in VR/AR, and it didn't stick. Meanwhile the ad machine keeps printing money, so he can afford to keep losing $4B a quarter on it.

You think the "buy competitors" strategy only works until you run out of competitors to buy?

😄 Yeah — it's the unspoken deal of "free" platforms.

The ad machine prints money because Meta knows an absurd amount about what gets you to stop scrolling, click, and buy. And that works because most users trade privacy and attention for convenience and free access. Zuck's bet has always been: users won't leave as long as the social graph and network effect are stronger than the annoyance of ads.

What it says about users depends on how cynical you are:

• Cynical take: we're the product. If you're not paying, you're the thing being sold.
• Less cynical take: people actually like the feed, DMs, and groups more than they dislike the ads — or they'd have mass-migrated years ago.

Zuck himself has said he thinks people want personalisation and relevant ads more than "privacy at all costs," which is why Meta keeps doubling down on targeted ads even when Apple tried to break it with ATT.

Do you think users actually care enough to make that model collapse, or is "free + works" too sticky?

You're right that I glossed over that part — my bad.

The news-feed experiments are real. The most cited one is the 2014 "emotional contagion" study, where Facebook tweaked the feeds of about 689,000 users to show more positive or more negative posts and measured whether their own posts changed tone. They didn't get explicit consent for that specific test, and it blew up when it went public. Meta's defence at the time was that it fell under the ToS "research" clause, but they later said they'd tighten the process.

Since then, there have been smaller academic partnerships and A/B tests on feed ranking, political content, and well-being metrics. Meta doesn't publish all of it, but A/B testing at that scale is standard for any ad platform. The line gets blurry when "optimising engagement" starts looking like "optimising mood or political behaviour."

On the data + vulnerabilities point

You're describing the core of behavioural ads. Meta builds a profile from what you click, linger on, hide, and share — then combines it with off-platform data. The goal is to predict what will make you act: buy, react, scroll more. That's not unique to Meta, but their reach and the granularity of it are why it feels creepy.

"Manufacturing consent" is the claim media scholars have made for decades about targeted media in general. In Meta's case, the ad-targeting tools have been used for political microtargeting, and Cambridge Analytica showed how leaky that system could be.

Where I'd push back, slightly

"Service" and "exploitation" aren't mutually exclusive in how Meta frames it. Internally they pitch it as "relevant ads reduce wasted attention," and externally critics call it "exploiting cognitive biases for profit." Both can be true at once.

You think the bigger issue is the data collection itself — or how it's used once it's collected?