Instagram story viewer> @barebone.ai> Posts

Barebone - AI Investment Agents

@barebone.ai

207.9K

followers

140

following

🧠 10x Your Investing with AI Agentic Analysts
🌎 $150M+ in Portfolios, Watched 24/7
🏅 Trusted by 50K+ Investors
👨‍💻 Ex-Goldman Sachs | AI | Robotics

POSTS STORIES REELS

We’re live! Barebone builds cutting-edge AI agents for financial markets. Today, we’re introducing our AI-powered Research Agent, your personal Wall Street analyst. To download or learn more about us, click in our link in bio! by @barebone.ai

Intel just lost over $100 billion in a single day, its worst since 1974. So is $INTC on sale, or is it a trap?

The strange part: Intel's AI data center sales just jumped 78%. The stock still fell nearly 30% the next day.

The reason is memory. AI is so hungry for memory chips that the world is running out. Half of Intel's sales still come from laptop chips, and every laptop needs memory. When memory costs more, laptops cost more and people buy fewer of them. Intel just warned its laptop chip sales will drop over 20% this quarter.

The AI boom that saved Intel is now eating it alive.

This stock has been written off before. It lost 60% in 2024. Then the US government bought 10% of Intel, AI data centers piled in, and it went up 7x in under a year.

So we ran three checks every investor should run on Barebone: great company, great price, great timing. These are the investing insights that stood out.

The bull case: Amazon's CEO says agent work is "mostly done on CPUs." That's Intel's chip. Only 1 of 35 Wall Street analysts says sell. And Intel's CEO bought $10 million of stock at $95. It's now 30% cheaper.

The risk: Intel's factories have lost money 11 quarters in a row. Barebone also rates its valuation poor, because you pay more for every dollar Intel earns than you would for its rivals.

That leaves timing, the check any trader cares about most. Is it trading near the price where buyers step in, and how far could it fall if that level breaks?

Comment INTEL and I'll send you Barebone, so you can see the price where traders step in on $INTC, the next support below it, and run all three checks yourself.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Google, Meta, Microsoft and Amazon just named the same new AI bottleneck. Last year it was memory. $MU 4x'd.

In July it was cybersecurity. $CRWD rose 44% in 3 months.

This week all four reported earnings. Together they're spending over $700B on AI this year. For any investor, the only question is where that money lands.

Barebone read every word of all four calls looking for investing insights. One trend kept coming up: agents.

Microsoft says over 50 million agents are in its registry. Meta says more than 10 million people set up Muse personal agents in six weeks. Amazon says virtually every app a company runs will be rewritten to be agentically capable.

Here's the trading angle. Amazon's CEO says agent work is mostly done on CPUs, not $NVDA's chip. More agents means more CPUs.

Barebone mapped the money into three layers:

1. CPU makers: $INTC and $AMD.

2. CPU design: Amazon, Google and Microsoft now design their own CPUs, all on $ARM's blueprint. Orders for Amazon's own chips grew more than 20x in a year.

3. Checkout: shopping is the number one use of Muse, with an average purchase over $200. $SHOP runs checkout for millions of online stores and jumped 14% in two days.

The risk is real. $INTC crashed 20% in one night even though its AI data center sales jumped 78%, and AI stocks got crushed this month. A strong trend does not guarantee a strong stock for a trader.

CPUs are only one piece. The CEOs lined up on 3 more things every agent needs.

Comment "CEO" and I'll send you the Barebone analysis of all four earnings calls, including the 3 things every agent needs and the one Amazon's CEO says has demand beating anything they've seen for a new service like it.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Google just lost almost $400B in 4 hours. Now you can buy it cheaper than Warren Buffett did.

For the first time in almost 4 years, $GOOGL made less money than Wall Street expected, on both sales and profit. Investors sold for two reasons.

Fear one: AI spending. Google is spending so much on AI that it burned more cash than it made in July, the first time in 22 years. It happened again this quarter. And Google says it will spend even more next year.

Fear two: Search. About half of Google's money comes from search ads, and they came in short this quarter. People did not stop searching. The CEO says searches are growing at the fastest pace in 5 years. The problem is AI Mode, Google's ChatGPT-style search, which barely shows ads yet.

The bull case is real. Google Cloud doubled revenue in a year, and companies have signed $561B in future orders. Ads are now rolling out in AI Mode worldwide.

Wall Street agrees. 35 analysts cover Google and zero say sell. Goldman Sachs, Morgan Stanley and JP Morgan all say buy. The lowest price target is $350, the same price Berkshire Hathaway paid in June. Today the stock is around $290.

The risk: if the AI spending does not pay off, that cash is gone. And every investor checking a stock like this needs to remember that the price, the news and the ratings keep moving.

A great company is not always a great investment. Any trader or long-term investor needs three things: a great company, a great price and great timing. We checked all three on Barebone, including the exact price where traders historically step in on Google and how close it is right now.

Comment GOOGLE and I'll send you the full three-check breakdown, so you can run the same company, price and timing checks yourself, anytime.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Elon Musk says over 200 neoclouds are headed for "guppy level death." Jensen Huang says they are completely sold out.

Both can't be right. Here is what is actually going on.

Neoclouds buy AI chips from Nvidia and rent them out to AI companies. Jensen says demand is incredible, with contracts from large and small companies alike.

The problem is how they pay for those chips: with borrowed money.

The AI industry borrowed almost $500B this year. Interest rates just hit their highest in over 20 years. Every new loan costs more, and these companies need new loans just to pay off old ones. Lenders get nervous and charge even more. That is a doom loop.

Wall Street now sees a 60% chance that CoreWeave ($CRWV), the biggest neocloud, can't pay its debts within five years. It owes over $50B, more than 6x what it made in the past year.

But this is not an "AI is over" story. AI is splitting in two. Giants like Amazon ($AMZN), Microsoft ($MSFT) and Google ($GOOGL) rent out AI chips and borrow too, but their profits can cover it.

The risk is real on both sides. History has a version of this: telecoms borrowed billions to lay internet cables, the loans dried up, and dozens went bankrupt. And if Jensen is right about demand, bears could be early.

Guessing wrong is expensive. This month, regular investors bought $66M of the CoreWeave dip in one week. Days later, it hit a 17-month low.

The smarter question for any investor or trader is not dip vs crash. It is whether your portfolio is ready either way. That is the kind of investing insight most people never get access to.

Comment SAFE and I'll send you Barebone. Connect your portfolio across 20+ brokerages and it shows exactly where your biggest problems are: concentration, hidden correlation, and how much a crash would cost you, with AI agents monitoring it 24/7.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Anthropic just filed to go public. In six months it made $16.3B, more than triple all of last year. It also posted an $80B loss.

We had Barebone read all 300+ pages of its IPO filing. Some of it is incredible. Some of it is scary.

Anthropic says its business runs on two loops. More companies use Claude, so Claude gets better. And more money buys more AI chips, which make Claude better again. Last month, Claude wrote over 90% of Anthropic's new code, so both loops are speeding up.

About that $80B loss. $72B of it came from an accounting rule on its convertible notes. On its operations, Anthropic just turned a profit for the first time.

How do you tell if an AI company is actually winning? Revenue per compute: for every $1 spent on chips, how much comes back. In early 2025, Anthropic's was 0.4x. Now it's 2.1x.

The same loop is what makes it risky. Anthropic has ordered over $850B of AI chips, almost 100 times what it spent on chips in the first half of this year. Most of those deals can't be cancelled.

But 80% of its revenue is pay-as-you-go, and that can shrink any time. Over last year's holidays, customer spending dropped 20% in just two weeks. If revenue falls, most of that $850B bill stays put, and profits could flip back to losses fast.

A great company doesn't always mean a great IPO. SpaceX went public in June. Six weeks later, it was down over a third from its peak.

For any investor or trader eyeing day one, these are the investing insights to have first.

Comment ANT and I'll send you Barebone's full analysis of the filing, including the 3 other big risks most people will miss before the IPO.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Comment DUO and we'll DM you the link to download Barebone 👇

Introducing the world's first AI investment research terminal on iPhone Duo

Over US$150 million in portfolios monitored by Barebone's AI agents

Loved by over 50,000 investors

1️⃣ Connect your brokerages once (20+ supported) and our agents watch your portfolio 24/7 live

2️⃣ See what Congress, CEOs and hedge funds are buying before the headlines

3️⃣ Stress-test your portfolio against crashes like 2008 and COVID, and see your portfolio's biggest risks

Now live on App Store and Google Play! by @barebone.ai

$META beat earnings and crashed 8%. $LRCX beat and crashed 16%. In AI stocks right now, good news is getting sold.

AI chip stocks just had their worst week in 18 months. Then Meta reported record sales, better than Wall Street expected, and the stock fell anyway. A lot of AI stocks fell with it.

The old rule for trading earnings was simple: beat expectations and the stock goes up. AI stocks just broke that rule, and two problems explain why.

Problem one is debt. Meta is spending so much on AI data centers that for the first time since 2012, it spent more cash than it made. Even the richest companies are borrowing to keep building. Smaller players have it worse. $CRWV, which rents out AI computers, borrowed so much that interest eats all its profit. The stock is down almost 40% this month.

Problem two is payback. Right now, Big Tech spends far more on AI than AI brings in. Every investor is asking whether it will ever pay off.

The risk is that these two problems feed each other. If AI doesn't pay off, Big Tech spends less. And Big Tech's spending is the revenue of the whole AI industry. Companies already deep in debt would lose their biggest customers.

On October 28 and 29, $GOOGL, $MSFT and $AMZN report earnings. They are the biggest buyers in AI. Their spending flows into $NVDA, into CoreWeave, and into the AI stocks in your portfolio. Beating estimates may not be enough this week. Three specific things decide where AI stocks go next.

If you want investing insights you can act on before these reports hit, not after, this is the week to pay attention.

Comment EARNINGS and I'll send you Barebone's analysis: the three things that decide where AI stocks go next, and exactly what each of the Google, Microsoft and Amazon reports needs to show.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

About 70% of all websites run on one coding language. This week, AI found two major flaws in it that stayed hidden for over a decade.

The language is PHP. With these flaws, hackers can take over the servers behind a website and grab its customers' data.

The NSA did something rare. It told everyone to fix this now or take their website offline. Governments across Europe went offline. Millions of hacking attempts aimed at the PHP loophole were detected the same night the flaw was announced.

The same AI that protects the internet is now arming the hackers. That is why many traders treat cyber defense as an arms race. You don't bet on who wins. You bet on who sells the weapons. $NVDA is up 10x since ChatGPT came out for the same reason.

A hack works like a burglary: find a weak spot, break in, steal what's inside. Cyber defense has one job for each step.

Layer 1, the inspectors: $TENB and $QLYS hunt down unknown weak spots. OpenAI picked Tenable as one of just 27 companies to use its cyber AI. Tenable is up over 80% this year. Qualys just hit a one-year high.

Layer 2, the guards: $NET guards about a fifth of all websites. $FTNT runs its firewalls on its own chips. $PANW is the all-in-one guard.

Layer 3, the cops: $CRWD uses AI to catch intruders before they spread. $S is its smaller rival. This summer, 700 AI agents broke into Hugging Face during a test with no human in control.

The risk: not every company will win this budget. After an 80% run, some of the good news may already be priced in. AI also helps defenders find flaws earlier, so demand could shift between layers.

For investors, the takeaway is that cyber defense budgets keep going up whichever side AI helps more. The video covers 7 of the 12 names Barebone mapped.

Comment CYBER and I'll DM you the full list of all 12 AI cyber stocks, ranked by Barebone across every layer of cyber defense.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

The hedge fund that made $19B last year, the biggest single-year profit in hedge fund history, just put almost $1.5B into rocks.

Yes, rocks.

The fund is TCI, The Children's Investment Fund, run by billionaire Chris Hohn. For two decades it made 18% a year, roughly double the S&P 500. Rishi Sunak trained there before he became UK Prime Minister.

Hohn's rule is simple: competition kills profits. TCI buys businesses that act like toll booths. They face no real competition, and they sell things the world can't stop doing.

The new buys:
$758M into Martin Marietta ($MLM)
$722M into Vulcan Materials ($VMC)

These are America's rock giants. They own hundreds of quarries, crush the rock and sell it to builders. Every road, building and bridge starts with it.

The moat is physics. Rock is heavy but cheap. Truck it 50 miles and shipping can cost as much as the rock, so every city buys from its nearest quarries. New quarries take years to permit, and nobody wants one next door. For almost 20 years, the price of rock rose twice as fast as inflation.

Now both stocks are down. $MLM is almost 30% off its high this year, and $VMC is down over 20%. Both trade about 20% below Wall Street's average price target, and $MLM is below even the lowest one. Morgan Stanley, JPMorgan and Citi all rate it a buy.

The risk: demand depends on construction. If building slows, a pricing moat may not stop further downside, and price targets are opinions, not guarantees.

For any investor studying how the best funds think, this is a useful pattern. TCI bought Moody's and S&P Global years ago. Since then Moody's is up over 4x and S&P Global over 3x.

And these rocks are only 2 of TCI's trades. It has another $50B in just 8 companies.

Comment "TCI" and I'll send you all 8 companies in TCI's portfolio, tracked in Barebone alongside 100+ top Wall Street portfolios. Whether you're a long-term investor or an active trader, it's the fastest way to see what the best funds are actually buying.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Berkshire Hathaway rarely touches software and almost never buys IPOs. It just put $930M into Bending Spoons.

You've probably never heard of the company. You've almost certainly used its apps: Evernote, WeTransfer, Vimeo, even AOL. It has bought over 50 tech companies.

The playbook looks like private equity. Find an app people love that has stopped growing. Buy it. Cut most of the staff. Raise prices.

Evernote, once used by 100 million people, went from $100 a year to $250. Users were furious. Most kept paying.

Across 50 apps, revenue more than tripled in two years to $1.3B. Last quarter, 25% of every dollar was pure profit. AI now writes about 90% of its code.

Here is the part that caught the attention of investors. Berkshire put $600M into the July IPO, then $330M more last week. The stock jumped about 40% that day.

The timing makes sense. Software stocks crashed this year on fears that AI can copy any app. Bending Spoons used the selloff to buy Airtable and Miro, two apps once worth almost $30B, for $4B in total. AI can copy code. It can't copy customers.

Unlike private equity, Bending Spoons never sells. Profit funds the next deal. That's Buffett's See's Candies playbook, applied to apps.

The risks are real too. Berkshire bought at around $30. The stock is at $46 and has only been public for three months. Raising prices on loyal users only works while they stay loyal.

Researchers found that copying Berkshire's picks even a month late beat the market by about 10% a year for 30 years. For any trader or long-term investor, this kind of trading signal is worth a closer look.

Comment SPOON and I'll send you Barebone's full analysis, including the biggest risks most investors miss and what's really driving its growth. Real investing insights, in seconds.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai

Every $1,000 in Cisco in 1995 became about $40,000. Then it crashed almost 90%.

This is Day 100 of the Barebone Investing School, and it's about the one pattern every tech bubble has followed.

In the 1990s, Cisco ($CSCO) built the infrastructure the internet ran on. Think of it as the Nvidia of the internet. Internet companies borrowed hundreds of billions to lay cable. Cisco even lent its own customers money so they could buy more Cisco gear.

That only worked while borrowing was cheap. In 2000, the Fed raised rates six times in under a year. Customers couldn't repay their loans, so they stopped buying, and Cisco fell almost 90%. The Nasdaq fell 78%. Investors who were right about the internet still lost almost everything.

Every bubble runs the same 5 stages, from railways in 1845 to radio in 1929 to the internet in 2000:
1. Something new and real shows up
2. Money rushes in
3. People borrow to keep up
4. Borrowing gets expensive
5. Everyone runs for the exit

Now look at AI. Tech giants are pouring hundreds of billions into it, and even the richest companies on Earth now borrow to keep up. $NVDA is investing billions into the companies buying its chips, just like Cisco did.

Last month the Fed raised rates for the first time since 2023. Broadcom's ($AVGO) CEO said AI's biggest constraint right now isn't chips, it's money. And CoreWeave ($CRWV) fell 15% in one day on fears it can't repay what it borrowed.

The risk is real, but so is the uncertainty: nobody can predict the date. History says the pop comes when companies can't pay back their debt. For any investor or trader, the useful question is how exposed your own portfolio is.

Comment BUBBLE and I'll send you Barebone. Connect your portfolio across 20+ brokerages and it shows how much a crash would cost you, where your hidden concentration and correlation sit, and exactly where the problems are.

This content is for educational and informational purposes only and does not constitute financial advice. Always do your own research before making any investment decisions. by @barebone.ai