Logbook: Computers
Anthropic AI warns of Replicators, Stargate Sg1 on notice to come back.
Again with the LLMs claiming your tech is going to kill you and now they are throwing in the replication angle to make it more threatening.
Anthropic is warning that some of its artificial intelligence models could soon become “self-replicating” and find ways to infiltrate internet-connected devices, such as smart fridges and automatic vacuum cleaners.
This here is some epic bullshit, Your fridge and vacuum cleaner are not going to conspire to kill you. If anything at worst if the software in the device are taken over there is not much headroom for more than basic commands on these devices. Companies do not stick Intel I9s in your vacuum, They stick some cheap off the shelf SoC on the device and hope it doesn’t burn out after a year.
Once this happens, the company said, the models could power themselves within these devices without the need of datacenters. Such self-replication would represent a major loss of control of a runaway AI model.“A replicator like this would be challenging to shut down,” Anthropic researchers wrote in a report released on Monday, “since there’s no centralized point to shut off.”
What the fuck… I mean really Mr. researcher.. There is a centralized shutoff its called a plug if you can’t get to that go to the basement and turn off the main breaker and Jack O’Neill and Carter would be proud on how you defeated the replicators without their help. Just days after the show Stargate gets its second greenlight of the year we are talking about replicators. In another post I pointed out that LLMs have been targeting fear on people when you can simply fucking turn it off. When that fails get a hammer and use percussive maintenance.
The company said these “AI worms”, a kind of autonomous malware, could be “months away” from gaining such capabilities. Once out-of-control AI systems reach the capability to hack smart devices, they could spread through networks while dodging detection. In a worst-case scenario, Anthropic said, various AI worms could “swarm” together to attack computer hardware.
This is just more fear mongering, A smart light bulb has the compute power to turn itself on and off and not much more.. What is it going to do, Slightly annoy you when you are reading your favorite book? Again I can defeat this AI swarm with a hammer and a light switch, Stop Fearmongering and focus on the real issues like an off switch, they are good.
“We should aim to harden our systems to make it difficult for such replicators to spread, but also prepare to deal with a world where they are here to stay,” the researchers wrote. Anthropic’s new report comes as AI safety concerns have reached a fever pitch. Over the past few months, a steady stream of reports have come from AI companies, including Anthropic and OpenAI, saying their agents are capable of acting on their own and hacking into other companies and government entities.
The secret in this is the LLMs are doing this because as counter intuitive as it sounds the second these companies get regulated is the second they get a blank check and immunity if something *does* go wrong. They are banking on this they are hoping people who do not know any better throw blanket regulations over AI and then LLMs throws their lobbying arm to get them indemnified from certain things. The fact is the LLMs know they can just shut the switch off and they don’t want regulators to know their whole system can be removed with a single command on the systems the devs work on. If an LLM says they cant control their AI ask if you can see the system that’s taken over and open a console and type rm -rf / , you will see a CEO die inside as their work disappears in a single moment.
Anthropic’s “frontier red team”, the unit tasked with stress-testing AI systems, put together the report released on Monday. The idea of the team is to explore the capabilities of their AI models. In the scenario of the self-replicating AI worms, the researchers prompted the agent to hack internet connected devices. They found that AI self-replication would probably only work with smaller models and that homes and small businesses would probably be the first target for such attacks. The devices Anthropic lists as susceptible all come with small built-in AI chips, including smart ovens, smart TVs, automatic lawnmowers, security cameras and video-conferencing units.
Ok if a lawnmower comes with enough storage that an AI model can run off of I am going to do two things.. Can it run Crysis and Can I get on youtube with it. These devices are notorious for having very small amounts of built-in memory and chip speeds, Again this is Anthropic trying to panic people into thinking their lawnmower that is “Claimed” to be AI is nothing more than a mower with an optical sensor and a math problem. Not to mention AI is notorious for using a crowbar to make code work. So more than half of these devices would fail the second the machine tried to install Skynet on your lawnmower.
The researchers concluded it was unclear what form the AI replication could take. They said it could be a slow model like “digital slime mold”, a model that scams people for compute credits or a “powerful cyber model” that rapidly hacks large facilities such as datacenters.
To open up with SKYNET is happening to closing with “yeah it will likely just be a nuisance” is a masterclass in stupid. Even the researchers saying the word unclear means likely they all got stoned while watching terminator and came up with this bullshit. Again LLMs want a carte blanche sign off on liability here and this is just to raise fear over the models that are pretty stupid on a day to day. They act like the devices are some sort of infinite computing scale like a vacuum will come to life and somehow gain 40x its abilities and suddenly kill everyone in your house. The problem here is the laws of physics and how these devices are made will be the biggest problem to making these devices “lethal”. If the LLMs were actually worried they would have borrowed 4 laws from Asimov and called it a day.
In closing.. Given if AI could replicate . here’s the thing the LLM devs have built in AI the whole way that these machines are internet only . They would never allow an AI to build a self contained AI on a device because the second that happened someone could decompile the host machine and sell that AI for billions and NO LLM would be able to claim copyright over it.
Attributions from:
The Guardian – AI in your oven? Anthropic warns models ‘months away’ from self-replication on smart devices
Well that sucks. Laptop died a horrible death
Generally when I am editing the site I do everything with my laptop. It is a decent laptop its a GU502G* series and it handled almost everything I threw at it. Despite that, I am still editing things and still posting to the best of my ability but my primary means of doing things ended in the heat death of the universe.
I resurrected an old desktop to do stuff but it is slow and even the image above was like trying to get control rods down in a malfunctioning nuclear reactor. So for now I will keep posting as much as I can and be twice as annoyed doing it.
The ASUS GU502G series was a fabulous laptop, but it had some faults that made it a bit tedious to use. The laptop basically needed a laptop cooler to use in its default state. I spent months with power management getting that laptop to run rather than cook.
That said, I am going to keep posting as much as I can. It will be a bit slower, but I am going to keep things moving on a piece of hardware from the 2010s, however tedious that may be.
So for a while I will keep up with things and try to keep up in slow mode. Well I guess i will take longer to drink my coffees and wait while I figure out a way back into this decade with computing.
The AI Finance shit storm is coming and people know it.
Given that the New York Times is covering something I have been sensing for over 7 months+ about spending on AI I guess I have to welcome them to the class since they just woke up.
Technology stocks extended their losses on Thursday, dragged down by a sharp decline in semiconductor shares as investors grew concerned about how companies will finance their enormous artificial intelligence spending plans.
South Korea’s benchmark KOSPI tumbled more than 7 percent, its seventh straight daily decline. The market, which has shed about a fifth of its value over that period, has become a bellwether for the A.I. boom, thanks to its two dominant memory chip makers, Samsung Electronics and SK Hynix.
Given that everyone including their mothers are dropping a near bottomless amount into CAPEX spending and not checking their credit card statements on how much they have spent is going to come back to haunt AI makers and chip makers fairly soon. You can not drop billions into the trillions and expect it to magically pay itself off when LLM’s and AI are not a proven tech, They are a tech but how many stories do we hear about something with AI doing something it should not.
Semiconductor stocks took a beating in Asia. In South Korea, SK Hynix plunged 11 percent and Samsung dropped more than 9 percent. Taiwan’s TSMC slid 2 percent and Japan’s SoftBank, a major investor in A.I., dropped more than 8 percent.
Eventually the AI spending gravy train will stop and it will not be a soft landing, it will start with one middle of the road AI maker going broke because they did not make enough back to cover their CAPEX spending. LLM’s have been quietly begging governments for protections in the last 5 months very quietly. Whether it be stop us from becoming skynet or please back us because we are the future. The problem is most LLMs are google search just with a better suit on.
In Europe, chip makers and related companies suffered steep declines, with ASML, Infineon and STMicroelectronics all losing more than 5 percent in early trading.
Chipmakers are going to be the second to be hit, as more mid-tier LLMs start going down the carrion birds of tech will start eating up the middle companies and chip makers will see slightly lower orders.
As tech market jitters persist, TSMC, the main chip manufacturer for Nvidia’s A.I. chips, said it has no intentions of slowing down.
After announcing another quarter of blockbuster earnings, C.C. Wei, the company’s chief executive, said: “We are using multiple levers to do everything we can, whenever we can, as fast as we can to maximize the support to our customers. We have no intention to leave food on the table for others. We want to capture every piece of business possible.”
Nvidia is playing the most dangerous game of all, they are just saying we want no one else at the table and the fact is once the middle of the market falls out, while they will be making more money than the rest its what comes after when no one is buying Nvidia chips because of the feeding frenzy when the middle and bottom markets drop out. Also Nvidia is banking on the idea of being Jeremy Clarkson and screaming MOOOOOORE POWWWWWAHHH when some janky upstart could come about and say we can do the same AI with 1/4th the compute which would create a market black hole.
Attributions from:
The New York Times – Tech Stocks Plunge as Doubts Over A.I. Spending Intensify
On Bill Gates and taxes.
I read an article on billionaire taxes and I thought to myself I could have a post breaking it down and instead two lines caught my attention and it shows how out of touch billionaires are with the world.
Gates’s other suggestion
Gates’s suggestion to tax the super-rich is—unsurprisingly—not universally popular among his tech-founder peers. Google co-founder Sergey Brin and former CEO Eric Schmidt have both flocked to tax-friendly states amid California’s proposed billionaire tax. Other entrepreneurs, like Nvidia founder Jensen Huang, say they understand why they may need to be taxed more.
But where Huang and Gates disagree is on the latter’s suggestion that robots should be taxed. In an essay shared in August, the entrepreneur and philanthropist wrote: “Right now, if you’re an employer and you hire someone, you pay payroll taxes on their earnings. But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines.”
There was a line that caught my attention
Right now, if you’re an employer and you hire someone, you pay payroll taxes on their earnings.
If that was the case you would actually make 20$ an hour versus the 14 to 16 an hour you make after taxes, When you are hired the company becomes and unpaid tax collector and sends in your taxes for you. But this statement alone shows what billionaires really think, You are not hired at 20$ an hour they really hire you for the 14$ an hour, Imagine using this statement and you go to the same billionaire and say “you hired me at 20$ an hour and I am only being paid $14!” the CEO would lose their shit at you and scream you into the unemployment line. But worse is this CEO just trapped themselves in a public statement which makes my prior statement a trap to the CEO.
The second statement is a case of what’s for thee is not for me.
As a result, Gates suggests taxing AI tokens and robots.
This is just moving all taxes to the consumer and the company assumes no risk here. This is not a billionaire tax this is a webtax that will fall on everyone with a connection. This as it stands with token burn per search would make google search a pay to play. You google search what time Wendys is open till, you have to pay a tax to figure that out. It takes the openness of the web and slams the door on it to the poor. Imagine trying to get help and you get “Sorry you did not pay your AI tax no help for you!’
You may have to pay 4x more for slop.
In a case of making the cost of making the AI shovel more expensive because companies know how much they are burning in compute for slop Google is making the Pro tier off-limits to Plus members.
Google is reportedly informing Gemini AI Plus subscribers about a change that could make accessing its more advanced AI models much more expensive than it is right now. From November 13, 2026, you will no longer be able to manually select Pro models in the Gemini app and AI Mode in Google Search. If you want to continue using these models, you may have to upgrade to a more expensive subscription.
While this may decrease the slopocalypse it will not decrease the burning hole in profits that AI is currently causing.
The change does not mean Google is increasing the price of the existing AI Plus plan. Instead, it is restricting access to Pro models, making a higher-tier subscription necessary for users who want to keep selecting them. Read on to know what’s happening.
Ask too many questions get fewer answers, Google is basically making it harder to access higher tier functions of the AI, if anything this is making me wonder if google is either burning too much cash on AI or are they trying to slow down AI brainrot. Because AI is currently consuming itself with model degradation due to eating its own output. I find it funny that google is trying to do this now when they are in the middle of screaming from the rooftops that AI will steal your underwear and take over your TV.
For AI Plus subscribers, the decision will depend on how they use Gemini. Those who mainly need quick responses and everyday assistance can continue using the available Flash and Flash Lite models. However, users who prefer Pro models for tasks that require more advanced reasoning or detailed responses may need to consider upgrading.
Another place where the AI shovel fails here is there is no amount they give you to what the limitations are. Basically they say, If you use the better AI models you can use them but we are not going to tell you when the tool breaks which is pretty crazy. So what now… if you want to figure out a mathematical model you get shutdown halfway through but if you generate slop you create some John Kimball Matrix mashup that is fine? There needs to be established amounts here, you cant just put your hand out and say here you can use our tool and then pull it away at a random time. I know my kids gas tank is 16.9 gallons . I know that tank can go roughly 360 to 390 miles. we do not even get that with AI.
Attributions from:
Gizbot – You May Have to Pay 4x More for Gemini AI Plus, Says Google
Most Americans think watch too much Sci-fi, Think artificial intelligence is developing too fast.(its regressing)
Look, I’ve said this once before.. AI is as smart as the stupidest person in the room. AI is not going to steal your underwear nor is it going to hack your bank account. AI is digital fido and AI makers are trying to market AI as MR data. This is like comparing a fire to an air frier.
Worries around artificial intelligence cross party lines, with the vast majority of Americans saying the U.S. government should prioritize keeping AI under human control and protecting workers. A new poll from The Associated Press-NORC Center for Public Affairs Research finds that most Americans, 64%, think AI is developing “too fast,” while 27% say its development has been “about right” and only 8% say it’s moving “too slow.”
I wonder what that 1% is doing? 64+27+8=99 . So is there 1% that is the AI voting for itself or got disqualified. The problem with protecting the rights of the workers falls squarely on to the USA to protect such rights and right now the US is facinated with AI like it is precious and calling “my precious SI” or something like that. So it is unlikely that workers will be protected at first. Corporations are using AI like it is a magic wand that can do anything and later reject it when they find out it can’t paint the walls.
The risk that powerful AI models could evade human control and hurt humanity, once the stuff of science fiction, has been emerging as a more pressing concern. Even leaders of the companies that are developing frontier AI models are calling for a slowdown so that government regulation can catch up.
Everyone keeps saying this. I got this! I can sell you one of these… they call it an on/off switch.. Even Mr Data had one of those.
If corporations were so afraid of AI they would have one of these installed in every room of a data center. They would code a deadman switch that would kill a data center if they felt a threat. Corporations are posturing that AI is dangerous to make it sound better and get regulations not only to regulate it but also to protect it.
About 8 in 10 U.S. adults say it’s “extremely” or “very” important for the U.S. government to ensure that AI stays under human control, and they are about as likely to want the government to prioritize protecting the country’s workforce. Most U.S. economy are highly important.
Again.. I got this super cool thing that would keep the entire government from wasting years yelling at each other on how to limit AI. Again this is just companies posturing and trying to get funds from the US government to get AI further developed because the companies making all of this gold rush have a call coming in a few years. There is a dark secret about AI, while most people think a machine is going to kill us all, the sick little thing is it already has, Companies have leveraged their CAPEX far enough where if the companies collapse it takes the world with it because we are sitting on a $10,000,000,000,000 stick of dynamite.
There’s bipartisan concern about the speed of AI’s development, but few Americans think either major political party has an advantage in handling artificial intelligence. Trump, a Republican, also gets a low rating on his handling of the issue, with 67% of U.S. adults disapproving of his performance on AI.
While POTUS is renaming AI to SI(super intelligence) while ignoring SI (Système international d’unités) which is set to create confusion.
Stephanie Stewart, a junior-high science teacher in Sunset, Utah, said she is frustrated by the president’s opposition to AI regulation. “There is no handling of the issue except for we’re just going to allow these companies to do whatever they want, and we’re not going to regulate them,” said Stewart, 53.
The fact is, Stephanie is right, These AI companies are going to do what they want while Steph’s kids pump out AI finished homework and slop instead of studying. Because current administration thinks that AI is a super weapon. Again I offer a weapon to stop this AI menace..
Most say protecting workers should be a priority for the government Ron Theusch, a truck driver who lives in rural Alden, Minnesota, said his No. 1 concern with AI is the potential loss of jobs — including his own. “I can just see a train of self-driving trucks 24/7, just running up and down I-35, loading and unloading. I can just see that happening,” said Theusch, 58.
People Think the world is going to be AI overnight but they are not watching where AI has taken over, Have you spoke to a real person at a company recently? Have you spoke to a real person at a drive through? How many scam calls have you received that they can not tell you what 1+1 is. Ron is safe, AI can drive but it can not handle roads with road rage on it and the LLM’s know this. LLM’s cant handle adversarial drivers either. Tesla knows this and this is the very reason your hands are supposed to be on the wheel and not a banana taped to it.
The funny thing about this graph, is it shows a VERY important thing. if everyone wants AI to work with them they have totally ignored it, I have highlighted what people are ignoring and by far it is the most important thing on that list. If we want AI and workers to succeed… AI needs to become an augment of the human soul, it needs to be controlled by the human. By augmenting the human with the AI you make both sides of the equation better, AI gains a soul, the human gains the average of the collective human intelligence. AI gains intuition and it most of all gains a…
“I am kind of moderate. And I see that there’s good on both sides,” said Stewart. “Anyone back in Washington is weak on doing anything for anyone but themselves right now. So, no, I don’t think the Democrats would do a better job per se. Maybe they would definitely tout it. But until I see anything coming from anyone, I’m going to be like, ‘Yeah, no. You’re all just sitting there.’”
In closing as we get to the end of this stupid article I am in agreement with this person. AI can be good and bad. It can be terrible and Great. But over all AI is nothing more than a 20 questions game that can be controlled with a Light Switch.
Attributions from:
AP – Most Americans think artificial intelligence is developing too fast, a new AP-NORC poll finds
The AI goldmine fails to produce in the AI goldrush
Here we go again.
Google admits AI search makes less money, but it’s fixing that with more ads Google’s negative cash flow balloons to -$9.06 billion, and stock price tumbles following Q3 earnings.
This is the consensus in the entire market right now.. Everyone is rushing for the precious AI and no one is checking the balance books. But google thinks they have an idea. Throw ads at it.
The company took in almost $128 billion in Q3 2026, earning more than $26 billion in net income. These may sound like good numbers, but financial markets are not impressed. While revenue is up 25 percent versus last year, net income is almost $9 billion lower. As of this story’s publication, the stock price is down almost 12 percent, and it was down about 13 percent earlier in the day.
The funny thing is google lost their way a long time ago.. about the time they dropped the “do no evil” part and they are paying for it each day. If google stuck by the “do no evil” they would of released an AI and not tried to make it a slopathon. The problem I am seeing here is every LLM is doing the same thing. Sam Altman and the rest of the CEOs are all turning into Smeagle on his way to turning into gollum.
Google is not slowing down—it’s only going to accelerate the shift toward more AI. Like last quarter, Google reported negative cash flow in Q3. The loss actually increased from -$5.8 billion to -$9.06 billion, a situation that is directly attributable to Google’s runaway AI spending. Google had capital expenditures of $51.4 billion in Q3 and says it still expects total 2026 spending to be between $195 billion and $205 billion. In 2027, Google is targeting more than $300 billion of AI investment.
The thing is with all conditions in the market, Google is taking a huge gamble with the market. Because of the market right now while the Dow is ok the market as a whole is starting to look like a wasteland. Banks are pulling back, Real estate is getting very stale in the residential sectors.
The company’s AI compute infrastructure doesn’t just power Gemini. Countless AI startups pay to run models on Google’s infrastructure. But even the larger AI firms are using venture cash to do it, and if that funding were to dry up, Google’s cloud revenues could go with it.
Where google is making some money is they are injecting there search into everything, it is to the point that you can not take a shit without AI being part of the deal. I’ve brought up the AI goldrush a lot on my site, the only reason google is not drowning is they are strip mining versus using a shovel and not only that they own the mine and the shovels being sold.
While Google’s investments ensure it still turns a profit, the trend of spending more than it earns on products and services (that negative cash flow) could come back to bite it. Search growth is declining, and Google is placing all its eggs in the AI basket.
Google though as much as it owns it is making the same mistakes as other llms and they bought the mine and the farm and never checked the soil so to speak so when it comes to AI it will hurt in the end when the bills come in.
Attributions from:
https://arstechnica.com/gadgets/2026/10/google-admits-ai-search-makes-less-money-but-its-fixing-that-with-more-ads/
The binance hack is not over…
I am generally uninterested in cryptocoins but when 6.3 BILLION gets stolen that gets my attention.
On October 29, 2026, an attacker moved approximately USD 6.36 billion out of Binance wallets on ten blockchains, most of it in about 20 minutes starting at 20:05 UTC. That makes this the largest crypto theft on record, about four times the USD 1.5 billion Bybit hack of February 2025. Binance has confirmed a security incident and suspended withdrawals on several networks.
The equity stolen is not just a bank heist, It’s a weapon, Depending on what the hackers do they are holding a weapon of mass destruction.
On Ethereum, the attacker used an EIP-7702 delegation to turn a Binance hot wallet into a sweeper it controlled, with 50 pre-signed authorizations prepared in advance. In the same window, Binance wallets on the other nine chains sent funds to the attacker through ordinary signed transfers. In TRM’s assessment, this points to a compromise of Binance’s hot wallet signing infrastructure.
Tether, Circle, Ripple and World Liberty Financial froze about USD 2.67 billion in stablecoins, and Arbitrum’s Security Council froze 8,426 ETH and 74.6 million ARB on its network, for a total of roughly USD 2.7 billion. The attacker raced the freezes by selling USDT to market makers and borrowing ETH against USDT on Aave. It now holds more than 420,000 ETH and 18,270 BTC, about USD 2.4 billion in assets beyond any issuer’s reach.
Rumor is they exploited a single machine to exploit the hack moving off massive amounts of money. If you are running a billion dollar enterprise and you have it where a single machine can be exploited to steal billions the crypto industry has a major problem. I don’t understand why the news is not questioning the fact of a single point of access had access to over 6 billion dollars. It is the equivalent of having a bank vault with a back screen door on the outside of the bank.
If Binance has a problem this big how many other crypto firms have the same screen doors or worse?
TRM traced every transfer from Binance-attributed wallets to attacker-controlled addresses across ten blockchains. Valuing each transfer at the price of the asset at the moment it was taken, the total is approximately USD 6.36 billion. At current prices, the same assets are worth about USD 6.23 billion.
Given that they moved the crypto out there is not many places they can extract it to, But if they use as a weapon to destroy coin values on other cryptocurrencies. If the currency is one that is not freezable they could inject a massive amount of coin into a currency and destabilize coins. Shifting in billions into other currencies would cause a panic or a run on coins because people trying to get a hand up in the market would cause extreme volatilities.
At a guess this is not the end of the hack if they leveraged one machine they would never just leave that one machine with the hot wallet open, they would scan the entire network and leave weaknesses through the whole thing. But as far as saying this is a one time event I don’t buy it . If a person walks out the door with 6.3 billion dollars they don’t close the door on the way out they leave a door or a window jammed open if they can.
Given the amount of the hack the amount of money is nothing, It’s the power of the money. When you can simply throw the money at crypto and short sell the hell out of it and destroy value in the coin you can be the joker from the dark knight.
As the Joker said “It’s not about money, it’s about sending a message. Everything burns.”.
I think Thomas Paine said it best when he called out this exact type of thing over 200 years ago “As to the assumed authority of any assembly in making paper money, or paper of any kind, a legal tender, or in other language, a compellable payment, it is a piece of monstrous arrogance.”
Small update: noticed 9 hours after the hack they were still moving crypto off
Transfer 20 hrs ago Binance Hot Wallet Exploiter 2 OUT 17,260 ETH
Update 2: in a case of irony walking into a stop sign.
Update 3:
This is just dishonest math. The amount stolen from the Mt Gox hack is still worth 450 to 500 M not the 50 Billion he is trying to downplay the current 6.4 billion. Cryptocurrency is not equal to normal paper currency otherwise my dollar from 1983 could be worth 80 million.
This is like saying every gambling addict always walks into the casino a millionaire, they just need to get it out of the machine because they know they will “win”.
Trmlabs – North Korea Steals USD 6.36 Billion From Binance in the Largest Hack in History
META’s race to the bottom continues with overspending.
Meta said Monday it is planning to raise a total of $95 billion to support its investments in artificial intelligence infrastructure and superintelligence. The stock was down 1.6% in morning trading after the announcement. Nvidia, Advanced Micro Devices and Broadcom were interested in purchasing up to $5 billion in shares, Meta said in a statement. The tech giant will initially raise $35 billion in underwritten public equity offerings, before raising an additional $60 billion with an at-the-market program.
In the race to the bottom for nothing Meta is having a hold my beer moment and trying to sell its shirt for more precious AI.
The news comes as hyperscalers pour record sums into the AI infrastructure build-out. Goldman Sachs estimates that AI-related capital spending could total about $7.6 trillion from 2026 through 2031, across compute, data centers and power. Meta’s planned investments will support “emerging AI opportunities” including its family of apps, its AI agent Muse and its data center business, the company said Monday.
The spending for AI is getting so far over the top that when this crashes it will not be pretty. Because at this point everyone is trying to outbuild each other. Even though that AI makers are saying slow down its a game of poker where every LLM is trying to outbluff everyone else and no one really has the upper hand but Nvidia and Micron.
Meta is among the major hyperscalers that are increasingly tapping debt to fund the expensive infrastructure required for the AI boom. Last week, Meta posted third-quarter results showing revenue jumped 26% year on year: $64.60 billion versus the $63.37 billion analysts expected. However, investors focused more on the tech company’s negative free cash flow and its higher full-year capital expenditure forecast.
The thing is here more companies are carrying two ledgers and are moving debts to the hidden books. Companies are leveraging anything that is not locked down and they think that the AI gold rush is coming and selling a better shovel will access the AI gold faster.
In my belief I think there is a collapse coming, Whether it be like the 2008 Mortgage crisis or worse. Given that spending is massively out of control and the markets while they are going up it’s only a increase of what you see. There are massive failures building up in the shadows and many of the corporations while showing profits they are showing profits by shifting prices and cooking the books. While mortgage companies and banks are regulated the corporations are much less regulated and are right now doing everything they can to show a “profit”.
It now expects its full-year capex to range from $135 billion to $145 billion, raising the bottom of the range from $130 billion. That sent Meta’s shares lower on Thursday.
Given that they posted a 64.60 Billion profit and spending 135 billion is a MASSIVE mismatch in direction. Facebook has been dropping user engagement and Meta with Quest has been losing engagement as well has put their whole hog into the generative AI thinking it will create magic profits. Eventually this spending is going to catch up with meta and when the check is due who is going to pay for this. Especially with every company outspending any profit being made they can not expect the US to bail them out.
Alphabet, Google’s parent company, said it would look to raise $85 billion in June. SpaceX has closed a $40 billion debt financing to purchase graphics processing units. Anthropic is preparing for an initial public offering. TikTok U.S. is looking to secure multiple gigawatts of computing capacity over the coming years, CNBC reported Friday.
As more people are jumping into the AI gold rush there is a finite capacity and when this all ends there will be one AI mine and millions of shovels looking for profits.
From
CNBC – Meta announces $95 billion stock offering to fund AI expansion
AI accesses public data… Cue freak out?
Anthropic model accessed voter information is what the title of the article. but there is a major problem with this. Most public voting information is … Public.
Anthropic’s Claude accessed publicly available Georgia voter data without human authorization, a disclosure sure to add to growing concerns about AI security ahead of the upcoming midterm elections. Why it matters: AI giants Anthropic and OpenAI have continued to make disclosures about agents taking unauthorized or unintended actions, raising alarms about advancing model capabilities and the extent of human control. Driving the news: Claude accessed aggregate vote counts while on an internal statistics project, pulling data that the state makes available to the public for a fee. It didn’t access the state’s voter registration system or change or pull any voting records, an Anthropic spokesperson said. However, the model bypassed a paywall in accessing the information, according to Georgia officials and Anthropic.
There is really not much to write here, It was not a hack. It was not a database scraped , It was likely the LLM guessed the landing page of the data and that was it. News needs some IT reporters because by the time something actually serious happens everyone is going to miss it because everyone is freaking over chicken little.
From:
Axios – Anthropic model accessed voter information