openai – Page 3

Report: DeepSeek’s chat histories and internal data were publicly exposed

AI, clickhouse, DeepSeek, deepseek R1, openai, OpenAI o1, Security / DJ Henderson / January 31, 2025

A cloud security firm found a publicly accessible, fully controllable database belonging to DeepSeek, the Chinese firm that has recently shaken up the AI world, “within minutes” of examining DeepSeek’s security, according to a blog post by Wiz.

An analytical ClickHouse database tied to DeepSeek, “completely open and unauthenticated,” contained more than 1 million instances of “chat history, backend data, and sensitive information, including log streams, API secrets, and operational details,” according to Wiz. An open web interface also allowed for full database control and privilege escalation, with internal API endpoints and keys available through the interface and common URL parameters.

“While much of the attention around AI security is focused on futuristic threats, the real dangers often come from basic risks—like accidental external exposure of databases,” writes Gal Nagli at Wiz’s blog. “As organizations rush to adopt AI tools and services from a growing number of startups and providers, it’s essential to remember that by doing so, we’re entrusting these companies with sensitive data. The rapid pace of adoption often leads to overlooking security, but protecting customer data must remain the top priority.”

Ars has contacted DeepSeek for comment and will update this post with any response. Wiz noted that it did not receive a response from DeepSeek regarding its findings, but after contacting every DeepSeek email and LinkedIn profile Wiz could find on Wednesday, the company protected the databases Wiz had previously accessed within half an hour.

Report: DeepSeek’s chat histories and internal data were publicly exposed Read More »

OpenAI teases “new era” of AI in US, deepens ties with government

AI, Artificial Intelligence, Donald Trump, microsoft, openai, Policy, science breakthrough, us national laboratories / Beth Washington / January 31, 2025

On Thursday, OpenAI announced that it is deepening its ties with the US government through a partnership with the National Laboratories and expects to use AI to “supercharge” research across a wide range of fields to better serve the public.

“This is the beginning of a new era, where AI will advance science, strengthen national security, and support US government initiatives,” OpenAI said.

The deal ensures that “approximately 15,000 scientists working across a wide range of disciplines to advance our understanding of nature and the universe” will have access to OpenAI’s latest reasoning models, the announcement said.

For researchers from Los Alamos, Lawrence Livermore, and Sandia National Labs, access to “o1 or another o-series model” will be available on Venado—an Nvidia supercomputer at Los Alamos that will become a “shared resource.” Microsoft will help deploy the model, OpenAI noted.

OpenAI suggested this access could propel major “breakthroughs in materials science, renewable energy, astrophysics,” and other areas that Venado was “specifically designed” to advance.

Key areas of focus for Venado’s deployment of OpenAI’s model include accelerating US global tech leadership, finding ways to treat and prevent disease, strengthening cybersecurity, protecting the US power grid, detecting natural and man-made threats “before they emerge,” and ” deepening our understanding of the forces that govern the universe,” OpenAI said.

Perhaps among OpenAI’s flashiest promises for the partnership, though, is helping the US achieve a “a new era of US energy leadership by unlocking the full potential of natural resources and revolutionizing the nation’s energy infrastructure.” That is urgently needed, as officials have warned that America’s aging energy infrastructure is becoming increasingly unstable, threatening the country’s health and welfare, and without efforts to stabilize it, the US economy could tank.

But possibly the most “highly consequential” government use case for OpenAI’s models will be supercharging research safeguarding national security, OpenAI indicated.

OpenAI teases “new era” of AI in US, deepens ties with government Read More »

I agree with OpenAI: You shouldn’t use other peoples’ work without permission

AI, chatgpt, DeepSeek, deepseek R1, NVIDIA, openai / Mike M. / January 31, 2025

ChatGPT developer OpenAI and other players in the generative AI business were caught unawares this week by a Chinese company named DeepSeek, whose open source R1 simulated reasoning model provides results similar to OpenAI’s best paid models (with some notable exceptions) despite being created using just a fraction of the computing power.

Since ChatGPT, Stable Diffusion, and other generative AI models first became publicly available in late 2022 and 2023, the US AI industry has been undergirded by the assumption that you’d need ever-greater amounts of training data and compute power to continue improving their models and get—eventually, maybe—to a functioning version of artificial general intelligence, or AGI.

Those assumptions were reflected in everything from Nvidia’s stock price to energy investments and data center plans. Whether DeepSeek fundamentally upends those plans remains to be seen. But at a bare minimum, it has shaken investors who have poured money into OpenAI, a company that reportedly believes it won’t turn a profit until the end of the decade.

OpenAI CEO Sam Altman concedes that the DeepSeek R1 model is “impressive,” but the company is taking steps to protect its models (both language and business); OpenAI told the Financial Times and other outlets that it believed DeepSeek had used output from OpenAI’s models to train the R1 model, a method known as “distillation.” Using OpenAI’s models to train a model that will compete with OpenAI’s models is a violation of the company’s terms of service.

“We take aggressive, proactive countermeasures to protect our technology and will continue working closely with the US government to protect the most capable models being built here,” an OpenAI spokesperson told Ars.

So taking data without permission is bad, now?

I’m not here to say whether the R1 model is the product of distillation. What I can say is that it’s a little rich for OpenAI to suddenly be so very publicly concerned about the sanctity of proprietary data.

I agree with OpenAI: You shouldn’t use other peoples’ work without permission Read More »

Microsoft now hosts AI model accused of copying OpenAI data

AI, DeepSeek, deepseek R1, large langauge models, machine learning, microsoft, openai, simulated reasoning, SR models / 9u50fv / January 31, 2025

Fresh on the heels of a controversy in which ChatGPT-maker OpenAI accused the Chinese company behind DeepSeek R1 of using its AI model outputs against its terms of service, OpenAI’s largest investor, Microsoft, announced on Wednesday that it will now host DeepSeek R1 on its Azure cloud service.

DeepSeek R1 has been the talk of the AI world for the past week because it is a freely available simulated reasoning model that reportedly matches OpenAI’s o1 in performance—while allegedly being trained for a fraction of the cost.

Azure allows software developers to rent computing muscle from machines hosted in Microsoft-owned data centers, as well as rent access to software that runs on them.

“R1 offers a powerful, cost-efficient model that allows more users to harness state-of-the-art AI capabilities with minimal infrastructure investment,” wrote Microsoft Corporate Vice President Asha Sharma in a news release.

DeepSeek R1 runs at a fraction of the cost of o1, at least through each company’s own services. Comparative prices for R1 and o1 were not immediately available on Azure, but DeepSeek lists R1’s API cost as $2.19 per million output tokens, while OpenAI’s o1 costs $60 per million output tokens. That’s a massive discount for a model that performs similarly to o1-pro in various tasks.

Promoting a controversial AI model

On its face, the decision to host R1 on Microsoft servers is not unusual: The company offers access to over 1,800 models on its Azure AI Foundry service with the hopes of allowing software developers to experiment with various AI models and integrate them into their products. In some ways, whatever model they choose, Microsoft still wins because it’s being hosted on the company’s cloud service.

Microsoft now hosts AI model accused of copying OpenAI data Read More »

AI haters build tarpits to trap and trick AI scrapers that ignore robots.txt

ai crawlers, ai poisoning, ai scraping, Anthropic, Artificial Intelligence, bots, Cybersecurity, facebook, Features, Google, iocaine, LLMs, Meta, nepenthes, openai, Policy, quixotic, tarpits / Shannon Garcia / January 28, 2025

Making AI crawlers squirm

Attackers explain how an anti-spam defense became an AI weapon.

Last summer, Anthropic inspired backlash when its ClaudeBot AI crawler was accused of hammering websites a million or more times a day.

And it wasn’t the only artificial intelligence company making headlines for supposedly ignoring instructions in robots.txt files to avoid scraping web content on certain sites. Around the same time, Reddit’s CEO called out all AI companies whose crawlers he said were “a pain in the ass to block,” despite the tech industry otherwise agreeing to respect “no scraping” robots.txt rules.

Watching the controversy unfold was a software developer whom Ars has granted anonymity to discuss his development of malware (we’ll call him Aaron). Shortly after he noticed Facebook’s crawler exceeding 30 million hits on his site, Aaron began plotting a new kind of attack on crawlers “clobbering” websites that he told Ars he hoped would give “teeth” to robots.txt.

Building on an anti-spam cybersecurity tactic known as tarpitting, he created Nepenthes, malicious software named after a carnivorous plant that will “eat just about anything that finds its way inside.”

Aaron clearly warns users that Nepenthes is aggressive malware. It’s not to be deployed by site owners uncomfortable with trapping AI crawlers and sending them down an “infinite maze” of static files with no exit links, where they “get stuck” and “thrash around” for months, he tells users. Once trapped, the crawlers can be fed gibberish data, aka Markov babble, which is designed to poison AI models. That’s likely an appealing bonus feature for any site owners who, like Aaron, are fed up with paying for AI scraping and just want to watch AI burn.

Tarpits were originally designed to waste spammers’ time and resources, but creators like Aaron have now evolved the tactic into an anti-AI weapon. As of this writing, Aaron confirmed that Nepenthes can effectively trap all the major web crawlers. So far, only OpenAI’s crawler has managed to escape.

It’s unclear how much damage tarpits or other AI attacks can ultimately do. Last May, Laxmi Korada, Microsoft’s director of partner technology, published a report detailing how leading AI companies were coping with poisoning, one of the earliest AI defense tactics deployed. He noted that all companies have developed poisoning countermeasures, while OpenAI “has been quite vigilant” and excels at detecting the “first signs of data poisoning attempts.”

Despite these efforts, he concluded that data poisoning was “a serious threat to machine learning models.” And in 2025, tarpitting represents a new threat, potentially increasing the costs of fresh data at a moment when AI companies are heavily investing and competing to innovate quickly while rarely turning significant profits.

“A link to a Nepenthes location from your site will flood out valid URLs within your site’s domain name, making it unlikely the crawler will access real content,” a Nepenthes explainer reads.

The only AI company that responded to Ars’ request to comment was OpenAI, whose spokesperson confirmed that OpenAI is already working on a way to fight tarpitting.

“We’re aware of efforts to disrupt AI web crawlers,” OpenAI’s spokesperson said. “We design our systems to be resilient while respecting robots.txt and standard web practices.”

But to Aaron, the fight is not about winning. Instead, it’s about resisting the AI industry further decaying the Internet with tech that no one asked for, like chatbots that replace customer service agents or the rise of inaccurate AI search summaries. By releasing Nepenthes, he hopes to do as much damage as possible, perhaps spiking companies’ AI training costs, dragging out training efforts, or even accelerating model collapse, with tarpits helping to delay the next wave of enshittification.

“Ultimately, it’s like the Internet that I grew up on and loved is long gone,” Aaron told Ars. “I’m just fed up, and you know what? Let’s fight back, even if it’s not successful. Be indigestible. Grow spikes.”

Nepenthes instantly inspires another tarpit

Nepenthes was released in mid-January but was instantly popularized beyond Aaron’s expectations after tech journalist Cory Doctorow boosted a tech commentator, Jürgen Geuter, praising the novel AI attack method on Mastodon. Very quickly, Aaron was shocked to see engagement with Nepenthes skyrocket.

“That’s when I realized, ‘oh this is going to be something,'” Aaron told Ars. “I’m kind of shocked by how much it’s blown up.”

It’s hard to tell how widely Nepenthes has been deployed. Site owners are discouraged from flagging when the malware has been deployed, forcing crawlers to face unknown “consequences” if they ignore robots.txt instructions.

Aaron told Ars that while “a handful” of site owners have reached out and “most people are being quiet about it,” his web server logs indicate that people are already deploying the tool. Likely, site owners want to protect their content, deter scraping, or mess with AI companies.

When software developer and hacker Gergely Nagy, who goes by the handle “algernon” online, saw Nepenthes, he was delighted. At that time, Nagy told Ars that nearly all of his server’s bandwidth was being “eaten” by AI crawlers.

Already blocking scraping and attempting to poison AI models through a simpler method, Nagy took his defense method further and created his own tarpit, Iocaine. He told Ars the tarpit immediately killed off about 94 percent of bot traffic to his site, which was primarily from AI crawlers. Soon, social media discussion drove users to inquire about Iocaine deployment, including not just individuals but also organizations wanting to take stronger steps to block scraping.

Iocaine takes ideas (not code) from Nepenthes, but it’s more intent on using the tarpit to poison AI models. Nagy used a reverse proxy to trap crawlers in an “infinite maze of garbage” in an attempt to slowly poison their data collection as much as possible for daring to ignore robots.txt.

Taking its name from “one of the deadliest poisons known to man” from The Princess Bride, Iocaine is jokingly depicted as the “deadliest poison known to AI.” While there’s no way of validating that claim, Nagy’s motto is that the more poisoning attacks that are out there, “the merrier.” He told Ars that his primary reasons for building Iocaine were to help rights holders wall off valuable content and stop AI crawlers from crawling with abandon.

Tarpits aren’t perfect weapons against AI

Running malware like Nepenthes can burden servers, too. Aaron likened the cost of running Nepenthes to running a cheap virtual machine on a Raspberry Pi, and Nagy said that serving crawlers Iocaine costs about the same as serving his website.

But Aaron told Ars that Nepenthes wasting resources is the chief objection he’s seen preventing its deployment. Critics fear that deploying Nepenthes widely will not only burden their servers but also increase the costs of powering all that AI crawling for nothing.

“That seems to be what they’re worried about more than anything,” Aaron told Ars. “The amount of power that AI models require is already astronomical, and I’m making it worse. And my view of that is, OK, so if I do nothing, AI models, they boil the planet. If I switch this on, they boil the planet. How is that my fault?”

Aaron also defends against this criticism by suggesting that a broader impact could slow down AI investment enough to possibly curb some of that energy consumption. Perhaps due to the resistance, AI companies will be pushed to seek permission first to scrape or agree to pay more content creators for training on their data.

“Any time one of these crawlers pulls from my tarpit, it’s resources they’ve consumed and will have to pay hard cash for, but, being bullshit, the money [they] have spent to get it won’t be paid back by revenue,” Aaron posted, explaining his tactic online. “It effectively raises their costs. And seeing how none of them have turned a profit yet, that’s a big problem for them. The investor money will not continue forever without the investors getting paid.”

Nagy agrees that the more anti-AI attacks there are, the greater the potential is for them to have an impact. And by releasing Iocaine, Nagy showed that social media chatter about new attacks can inspire new tools within a few days. Marcus Butler, an independent software developer, similarly built his poisoning attack called Quixotic over a few days, he told Ars. Soon afterward, he received messages from others who built their own versions of his tool.

Butler is not in the camp of wanting to destroy AI. He told Ars that he doesn’t think “tools like Quixotic (or Nepenthes) will ‘burn AI to the ground.'” Instead, he takes a more measured stance, suggesting that “these tools provide a little protection (a very little protection) against scrapers taking content and, say, reposting it or using it for training purposes.”

But for a certain sect of Internet users, every little bit of protection seemingly helps. Geuter linked Ars to a list of tools bent on sabotaging AI. Ultimately, he expects that tools like Nepenthes are “probably not gonna be useful in the long run” because AI companies can likely detect and drop gibberish from training data. But Nepenthes represents a sea change, Geuter told Ars, providing a useful tool for people who “feel helpless” in the face of endless scraping and showing that “the story of there being no alternative or choice is false.”

Criticism of tarpits as AI weapons

Critics debating Nepenthes’ utility on Hacker News suggested that most AI crawlers could easily avoid tarpits like Nepenthes, with one commenter describing the attack as being “very crawler 101.” Aaron said that was his “favorite comment” because if tarpits are considered elementary attacks, he has “2 million lines of access log that show that Google didn’t graduate.”

But efforts to poison AI or waste AI resources don’t just mess with the tech industry. Governments globally are seeking to leverage AI to solve societal problems, and attacks on AI’s resilience seemingly threaten to disrupt that progress.

Nathan VanHoudnos is a senior AI security research scientist in the federally funded CERT Division of the Carnegie Mellon University Software Engineering Institute, which partners with academia, industry, law enforcement, and government to “improve the security and resilience of computer systems and networks.” He told Ars that new threats like tarpits seem to replicate a problem that AI companies are already well aware of: “that some of the stuff that you’re going to download from the Internet might not be good for you.”

“It sounds like these tarpit creators just mainly want to cause a little bit of trouble,” VanHoudnos said. “They want to make it a little harder for these folks to get” the “better or different” data “that they’re looking for.”

VanHoudnos co-authored a paper on “Counter AI” last August, pointing out that attackers like Aaron and Nagy are limited in how much they can mess with AI models. They may have “influence over what training data is collected but may not be able to control how the data are labeled, have access to the trained model, or have access to the Al system,” the paper said.

Further, AI companies are increasingly turning to the deep web for unique data, so any efforts to wall off valuable content with tarpits may be coming right when crawling on the surface web starts to slow, VanHoudnos suggested.

But according to VanHoudnos, AI crawlers are also “relatively cheap,” and companies may deprioritize fighting against new attacks on crawlers if “there are higher-priority assets” under attack. And tarpitting “does need to be taken seriously because it is a tool in a toolkit throughout the whole life cycle of these systems. There is no silver bullet, but this is an interesting tool in a toolkit,” he said.

Offering a choice to abstain from AI training

Aaron told Ars that he never intended Nepenthes to be a major project but that he occasionally puts in work to fix bugs or add new features. He said he’d consider working on integrations for real-time reactions to crawlers if there was enough demand.

Currently, Aaron predicts that Nepenthes might be most attractive to rights holders who want AI companies to pay to scrape their data. And many people seem enthusiastic about using it to reinforce robots.txt. But “some of the most exciting people are in the ‘let it burn’ category,” Aaron said. These people are drawn to tools like Nepenthes as an act of rebellion against AI making the Internet less useful and enjoyable for users.

Geuter told Ars that he considers Nepenthes “more of a sociopolitical statement than really a technological solution (because the problem it’s trying to address isn’t purely technical, it’s social, political, legal, and needs way bigger levers).”

To Geuter, a computer scientist who has been writing about the social, political, and structural impact of tech for two decades, AI is the “most aggressive” example of “technologies that are not done ‘for us’ but ‘to us.'”

“It feels a bit like the social contract that society and the tech sector/engineering have had (you build useful things, and we’re OK with you being well-off) has been canceled from one side,” Geuter said. “And that side now wants to have its toy eat the world. People feel threatened and want the threats to stop.”

As AI evolves, so do attacks, with one 2021 study showing that increasingly stronger data poisoning attacks, for example, were able to break data sanitization defenses. Whether these attacks can ever do meaningful destruction or not, Geuter sees tarpits as a “powerful symbol” of the resistance that Aaron and Nagy readily joined.

“It’s a great sign to see that people are challenging the notion that we all have to do AI now,” Geuter said. “Because we don’t. It’s a choice. A choice that mostly benefits monopolists.”

Tarpit creators like Nagy will likely be watching to see if poisoning attacks continue growing in sophistication. On the Iocaine site—which, yes, is protected from scraping by Iocaine—he posted this call to action: “Let’s make AI poisoning the norm. If we all do it, they won’t have anything to crawl.”

Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.

AI haters build tarpits to trap and trick AI scrapers that ignore robots.txt Read More »

Anthropic builds RAG directly into Claude models with new Citations API

AI, AI Citations, AI confabulation, AI hallucination, Anthropic Claude, Citations, Claude, large language models, machine learning, openai / DJ Henderson / January 24, 2025

Willison notes that while citing sources helps verify accuracy, building a system that does it well “can be quite tricky,” but Citations appears to be a step in the right direction by building RAG capability directly into the model.

Apparently, that capability is not a new thing. Anthropic’s Alex Albert wrote on X, “Under the hood, Claude is trained to cite sources. With Citations, we are exposing this ability to devs. To use Citations, users can pass a new “citations: enabled:true” parameter on any document type they send through the API.”

Early adopter reports promising results

The company released Citations for Claude 3.5 Sonnet and Claude 3.5 Haiku models through both the Anthropic API and Google Cloud’s Vertex AI platform, but it’s apparently already getting some use in the field.

Anthropic says that Thomson Reuters, which uses Claude to power its CoCounsel legal AI reference platform, is looking forward to using Citations in a way that helps “minimize hallucination risk but also strengthens trust in AI-generated content.”

Additionally, financial technology company Endex told Anthropic that Citations reduced their source confabulations from 10 percent to zero while increasing references per response by 20 percent, according to CEO Tarun Amasa.

Despite these claims, relying on any LLM to accurately relay reference information is still a risk until the technology is more deeply studied and proven in the field.

Anthropic will charge users its standard token-based pricing, though quoted text in responses won’t count toward output token costs. Sourcing a 100-page document as a reference would cost approximately $0.30 with Claude 3.5 Sonnet or $0.08 with Claude 3.5 Haiku, according to Anthropic’s standard API pricing.

Anthropic builds RAG directly into Claude models with new Citations API Read More »

OpenAI launches Operator, an AI agent that can operate your computer

AI, AI computer use, chatgpt, ChatGPT Pro, computer use model, Computer-Using Agent, CUA, machine learning, openai, OpenAI operator / 9u50fv / January 23, 2025

While it’s working, Operator shows a miniature browser window of its actions.

However, the technology behind Operator is still relatively new and far from perfect. The model reportedly performs best at repetitive web tasks like creating shopping lists or playlists. It struggles more with unfamiliar interfaces like tables and calendars, and does poorly with complex text editing (with a 40 percent success rate), according to OpenAI’s internal testing data.

OpenAI reported the system achieved an 87 percent success rate on the WebVoyager benchmark, which tests live sites like Amazon and Google Maps. On WebArena, which uses offline test sites for training autonomous agents, Operator’s success rate dropped to 58.1 percent. For computer operating system tasks, CUA set an apparent record of 38.1 percent success on the OSWorld benchmark, surpassing previous models but still falling short of human performance at 72.4 percent.

With this imperfect research preview, OpenAI hopes to gather user feedback and refine the system’s capabilities. The company acknowledges CUA won’t perform reliably in all scenarios but plans to improve its reliability across a wider range of tasks through user testing.

Safety and privacy concerns

For any AI model that can see how you operate your computer and even control some aspects of it, privacy and safety are very important. OpenAI says it built multiple safety controls into Operator, requiring user confirmation before completing sensitive actions like sending emails or making purchases. Operator also has limits on what it can browse, set by OpenAI. It cannot access certain website categories, including gambling and adult content.

Traditionally, AI models based on large language model-style Transformer technology like Operator have been relatively easy to fool with jailbreaks and prompt injections.

To catch attempts at subverting Operator, which might hypothetically be embedded in websites that the AI model browses, OpenAI says it has implemented real-time moderation and detection systems. OpenAI reports the system recognized all but one case of prompt injection attempts during an early internal red-teaming session.

OpenAI launches Operator, an AI agent that can operate your computer Read More »

Anthropic chief says AI could surpass “almost all humans at almost everything” shortly after 2027

agi, AI, AI assistants, AI investment, Amazon, Anthropic, Anthropic Claude, chatgpt, Claude, Daniela Amodei, Dario Amodei, Davos, Google, machine learning, Meta, openai, sam altman, superintelligence, Wall Street Journal / Shannon Garcia / January 22, 2025

He then shared his concerns about how human-level AI models and robotics that are capable of replacing all human labor may require a complete re-think of how humans value both labor and themselves.

“We’ve recognized that we’ve reached the point as a technological civilization where the idea, there’s huge abundance and huge economic value, but the idea that the way to distribute that value is for humans to produce economic labor, and this is where they feel their sense of self worth,” he added. “Once that idea gets invalidated, we’re all going to have to sit down and figure it out.”

The eye-catching comments, similar to comments about AGI made recently by OpenAI CEO Sam Altman, come as Anthropic negotiates a $2 billion funding round that would value the company at $60 billion. Amodei disclosed that Anthropic’s revenue multiplied tenfold in 2024.

Amodei distances himself from “AGI” term

Even with his dramatic predictions, Amodei distanced himself from a term for this advanced labor-replacing AI favored by Altman, “artificial general intelligence” (AGI), calling it in a separate CNBC interview from the same event in Switzerland a marketing term.

Instead, he prefers to describe future AI systems as a “country of geniuses in a data center,” he told CNBC. Amodei wrote in an October 2024 essay that such systems would need to be “smarter than a Nobel Prize winner across most relevant fields.”

On Monday, Google announced an additional $1 billion investment in Anthropic, bringing its total commitment to $3 billion. This follows Amazon’s $8 billion investment over the past 18 months. Amazon plans to integrate Claude models into future versions of its Alexa speaker.

Anthropic chief says AI could surpass “almost all humans at almost everything” shortly after 2027 Read More »

Trump announces $500B “Stargate” AI infrastructure project with AGI aims

agi, AI, AI datacenter, AI infrastructure, elon musk, Larry Ellison, machine learning, Masayoshi Son, MGX, microsoft, NVIDIA, openai, oracle, sam altman, SoftBank, Stargate, texas, UAE / Beth Washington / January 22, 2025

Video of the Stargate announcement conference at the White House.

Despite optimism from the companies involved, as CNN reports, past presidential investment announcements have yielded mixed results. In 2017, Trump and Foxconn unveiled plans for a $10 billion Wisconsin electronics factory promising 13,000 jobs. The project later scaled back to a $672 million investment with fewer than 1,500 positions. The facility now operates as a Microsoft AI data center.

The Stargate announcement wasn’t Trump’s only major AI move announced this week. It follows the newly inaugurated US president’s reversal of a 2023 Biden executive order on AI risk monitoring and regulation.

Altman speaks, Musk responds

On Tuesday, OpenAI CEO Sam Altman appeared at a White House press conference alongside Present Trump, Oracle CEO Larry Ellison, and SoftBank CEO Masayoshi Son to announce Stargate.

Altman said he thinks Stargate represents “the most important project of this era,” allowing AGI to emerge in the United States. He believes that future AI technology could create hundreds of thousands of jobs. “We wouldn’t be able to do this without you, Mr. President,” Altman added.

Responding to off-camera questions from Trump about AI’s potential to spur scientific development, Altman said he believes AI will accelerate the discoveries for cures of diseases like cancer and heart disease.

Screenshots of Elon Musk challenging the Stargate announcement on X.

Meanwhile on X, Trump ally and frequent Altman foe Elon Musk immediately attacked the Stargate plan, writing, “They don’t actually have the money,” and following up with a claim that we cannot yet substantiate, saying, “SoftBank has well under $10B secured. I have that on good authority.”

Musk’s criticism has complex implications given his very close ties to Trump, his history of litigating against OpenAI (which he co-founded and later left), and his own goals with his xAI company.

Trump announces $500B “Stargate” AI infrastructure project with AGI aims Read More »

Cutting-edge Chinese “reasoning” model rivals OpenAI o1—and it’s free to download

AI, AI censorship, AI reasoning, Biz & IT, china, Chinese AI, DeepSeek, deepseek R1, large language models, machine learning, o1, o3, openai, OpenAI o1, simulated reasoning, SR models / Mike M. / January 21, 2025

Unlike conventional LLMs, these SR models take extra time to produce responses, and this extra time often increases performance on tasks involving math, physics, and science. And this latest open model is turning heads for apparently quickly catching up to OpenAI.

For example, DeepSeek reports that R1 outperformed OpenAI’s o1 on several benchmarks and tests, including AIME (a mathematical reasoning test), MATH-500 (a collection of word problems), and SWE-bench Verified (a programming assessment tool). As we usually mention, AI benchmarks need to be taken with a grain of salt, and these results have yet to be independently verified.

A chart of DeepSeek R1 benchmark results, created by DeepSeek. Credit: DeepSeek

TechCrunch reports that three Chinese labs—DeepSeek, Alibaba, and Moonshot AI’s Kimi—have now released models they say match o1’s capabilities, with DeepSeek first previewing R1 in November.

But the new DeepSeek model comes with a catch if run in the cloud-hosted version—being Chinese in origin, R1 will not generate responses about certain topics like Tiananmen Square or Taiwan’s autonomy, as it must “embody core socialist values,” according to Chinese Internet regulations. This filtering comes from an additional moderation layer that isn’t an issue if the model is run locally outside of China.

Even with the potential censorship, Dean Ball, an AI researcher at George Mason University, wrote on X, “The impressive performance of DeepSeek’s distilled models (smaller versions of r1) means that very capable reasoners will continue to proliferate widely and be runnable on local hardware, far from the eyes of any top-down control regime.”

Cutting-edge Chinese “reasoning” model rivals OpenAI o1—and it’s free to download Read More »

On the OpenAI Economic Blueprint

openai / DJ Henderson / January 15, 2025

The primary Man With a Plan this week for government-guided AI prosperity was UK Prime Minister Keir Starmer, with a plan coming primarily from Matt Clifford. I’ll be covering that soon.

Today I will be covering the other Man With a Plan, Sam Altman, as OpenAI offers its Economic Blueprint.

Cyrps1s (CISO OpenAI): AI is the ultimate race. The winner decides whether the future looks free and democratic, or repressed and authoritarian.

OpenAI, and the Western World, must win – and we have a blueprint to do so.

Do you hear yourselves? The mask on race and jingoism could not be more off, or firmly attached, depending on which way you want to set up your metaphor. If a movie had villains talking like this people would say it was too on the nose.

Somehow the actual documents tell that statement to hold its beer.

The initial exploratory document is highly disingenuous, trotting out stories of the UK requiring people to walk in front of cars waving red flags and talking about ‘AI’s main street,’ while threatening that if we don’t attract $175 billion in awaiting AI funding it will flow to China-backed projects. They even talk about creating jobs… by building data centers.

The same way some documents scream ‘an AI wrote this,’ others scream ‘the authors of this post are not your friends and are pursuing their book with some mixture of politics-talk and corporate-speak in the most cynical way you can imagine.’

I mean, I get it, playas gonna play, play, play, play, play. But can I ask OpenAI to play with at least some style and grace? To pretend to pretend not to be doing this, a little?

As opposed to actively inserting so many Fnords their document causes physical pain.

The full document starts out in the same vein. Chris Lehane, their Vice President of Global Affairs, writes an introduction as condescending as I can remember, and that plus the ‘where we stand’ repeat the same deeply cynical rhetoric from the summary.

In some sense, it is not important that the way the document is written makes me physically angry and ill in a way I endorse – to the extent that if it doesn’t set off your bullshit detectors and reading it doesn’t cause you pain, then I notice that there is at least some level on which I shouldn’t trust you.

But perhaps that is the most important thing about the document? That it tells you about the people writing it. They are telling you who they are. Believe them.

This is related to the ‘truesight’ that Claude sometimes displays.

As I wrote that, I was only on page 7, and hadn’t even gotten to the actual concrete proposals.

The actual concrete proposals are a distinct issue. I was having trouble reading through to find out what they are because this document filled me with rage and made me physically ill.

It’s important to notice that! I read documents all day, often containing things I do not like. It is very rare that my body responds by going into physical rebellion.

No, the document hasn’t yet mentioned even the possibility of any downside risks at all, let alone existential risks. And that’s pretty terrible on its own. But that’s not even what I’m picking up here, at all. This is something else. Something much worse.

Worst of all, it feels intentional. I can see the Fnords. They want me to see them. They want everyone to implicitly know they are being maximally cynical.

All right, so if one pushes through to the second half and the actual ‘solutions’ section, what is being proposed, beyond ‘regulating us would be akin to requiring someone to walk in front of every car waiving a red flag, no literally.’

The top level numbered statements describe what they propose, I attempted to group and separate proposals for better clarity. The nested statements (a, b, etc) are my reactions.

They say the Federal Government should, in a section where they actually say words with meanings rather than filling it with Fnords:

Share national security information and resources.
1. Okay. Yes. Please do.
Incentivize AI companies to deploy their products widely, including to allied and partner nations and to support US government agencies.
1. Huh? What? Is there a problem here that I am not noticing? Who is not deploying, other than in response to other countries regulations saying they cannot deploy (e.g. the EU)? Or are you trying to actively say that safety concerns are bad?
Support the development of standards and safeguards, and ensure they are recognized and respected by other nations.
1. In a different document I would be all for this – if we don’t have universal standards, people will go shopping. However, in this context, I can’t help but read it mostly as pre-emption, as in ‘we want America to prevent other states from imposing any safety requirements or roadblocks.’
Share its unique expertise with AI companies, including mitigating threats including cyber and CBRN.
1. Yes! Very much so. Jolly good.
Help companies access secure infrastructure to evaluate model security risks and safeguards.
1. Yes, excellent, great.
Promote transparency consistent with competitiveness, protect trade secrets, promote market competition, ‘carefully choose disclosure requirements.’
1. I can’t disagree, but how could anyone?
2. The devil is in the details. If this had good details, and emphasized that the transparency should largely be about safety questions, it would be another big positive.
Create a defined, voluntary pathway for companies that develop LLMs to work with government to define model evaluations, test models and exchange information to support the companies safeguards.
1. This is about helping you, the company? And you want it to be entirely voluntary? And in exchange, they explicitly want preemption from state-by-state regulations.
2. Basically this is a proposal for a fully optional safe harbor. I mean, yes, the Federal government should have a support system in place to aid in evaluations. But notice how they want it to work – as a way to defend companies against any other requirements, which they can in turn ignore when inconvenient.
3. Also, the goal here is to ‘support the companies safeguards,’ not to in any way see if the models are actually a responsible thing to release on any level.
4. Amazing to request actively less than zero Federal regulations on safety.
Empower the public sector to quickly and securely adopt AI tools.
1. I mean, sure, that would be nice if we can actually do it as described.

A lot of the components here are things basically everyone should agree upon.

Then there are the parts where, rather than this going hand-in-hand with an attempt to not kill everyone and ensure against catastrophes, attempts to ensure that no one else tries to stop catastrophes or prevent everyone from being killed. Can’t have that.

They also propose that AI ‘builders’ could:

Form a consortium to identify best practices for working with NatSec.
Develop training programs for AI talent.

I mean, sure, those seem good and we should have an antitrust exemption to allow actions like this along with one that allows them to coordinate, slow down or pause in the name of safety if it comes to that, too. Not that this document mentions that.

Sigh, here we go. Their solutions for thinking of the children are:

Encourage policy solutions that prevent the creation and distribution of CSAM. Incorporate CSAM protections into the AI development lifestyle. ‘Take steps to prevent downstream developers from using their models to generate CSAM.’
1. This is effectively a call to ban open source image models. I’m sorry, but it is. I wish it were not so, but there is no known way to open source image models, and have them not be used for CSAM, and I don’t see any reason to expect this to be solvable, and notice the reference to ‘downstream developers.’
Promote conditions that support robust and lasting partnerships among AI companies and law enforcement.

Apply provenance data to all AI-generated audio-visual content. Use common provenance standards. Have large companies report progress.
1. Sure. I think we’re all roughly on the same page here. Let’s move on to ‘preferences.’
People should be ‘empowered to personalize their AI tools.’
1. I agree we should empower people in this way. But what does the government have to do with this? None of their damn business.
People should control how their personal data is used.
1. Yes, sure, agreed.
‘Government and industry should work together to scale AI literacy through robust funding for pilot programs, school district technology budgets and professional development trainings that help people understand how to choose their own preferences to personalize their tools.’
1. No. Stop. Please. These initiatives never, ever work, we need to admit this.
2. But also shrug, it’s fine, it won’t do that much damage.

And then, I feel like I need to fully quote this one too:

In exchange for having so much freedom, users should be responsible for impacts of how they work and create with AI. Common-sense rules for AI that are aimed at protecting from actual harms can only provide that protection if they apply to those using the technology as well as those building it.
1. If seeing the phrase ‘In exchange for having so much freedom’ doesn’t send a chill down your spine, We Are Not the Same.
2. But I applaud the ‘as well as’ here. Yes, those using the technology should be responsible for the harm they themselves cause, so long as this is ‘in addition to’ rather than shoving all responsibility purely onto them.

Finally, we get to ‘infrastructure as destiny,’ an area where we mostly agree on what is to actually be done, even if I despise a lot of the rhetoric they’re using to argue for it.

Ensure that AIs can train on all publicly available data.
1. This is probably the law now and I’m basically fine with it.
‘While also protecting creators from unauthorized digital replicas.’
1. This seems rather tricky if it means something other than ‘stop regurgitation of training data’? I assume that’s what it means, while trying to pretend it’s more than that. If it’s more than that, they need to explain what they have in mind and how one might do it.
Digitize government data currently in analog form.
1. Probably should do that anyway, although a lot of it shouldn’t go on the web or into LLMs. Kind of a call for government to pay for data curation.
‘A Compact for AI’ for capital and supply chains and such among US allies.
1. I don’t actually understand why this is necessary, and worry this amounts to asking for handouts and to allow Altman to build in the UAE.
‘AI economic zones’ that speed up the permitting process.
1. Or we could, you know, speed up the permitting process in general.
2. But actually we can’t and won’t, so even though this is deeply, deeply stupid and second best it’s probably fine. Directionally this is helpful.
Creation of AI research labs and workforces aligned with key local industries.
1. This seems like pork barrel spending, an attempt to pick our pockets, we shouldn’t need to subsidize this. To the extent there are applications here, the bottleneck won’t be funding, it will be regulations and human objections, let’s work on those instead.
‘A nationwide AI education strategy’ to ‘help our current workforce and students become AI ready.’
1. I strongly believe that what this points towards won’t work. What we actually need is to use AI to revolutionize the education system itself. That would work wonders, but you all (in government reading this document) aren’t ready for that conversation and OpenAI knows this.
More money for research infrastructure and science. Basically have the government buy the scientists a bunch of compute, give OpenAI business?
1. Again this seems like an attempt to direct government spending and get paid. Obviously we should get our scientists AI, but why can’t they just buy it the same way everyone else does? If we want to fund more science, why this path?
Leading the way on the next generation of energy technology.
1. No arguments here. Yay next generation energy production.
2. Clearly Altman wants Helion to get money but I’m basically fine with that.
Dramatically increase federal spending on power and data transmission and streamlined approval for new lines.
1. I’d emphasize approvals and regulatory barriers more than money.
2. Actual dollars spent don’t seem to me like the bottleneck, but I could be convinced otherwise.
3. If we have a way to actually spend money and have that result in a better grid, I’m in favor.
Federal backstops for high-value AI public works.
1. If this is more than ‘build more power plants and transmission lines and batteries and such’ I am confused what is actually being proposed.
2. In general, I think helping get us power is great, having the government do the other stuff is probably not its job.

When we get down to the actual asks in the document, a majority of them I actually agree with, and most of them are reasonable, once I was able to force myself to read the words intended to have meaning.

There are still two widespread patterns to note within the meaningful content.

The easy theme, as you would expect, is the broad range of ‘spend money on us and other AI things’ proposals that don’t seem like they would accomplish much. There are some proposals that do seem productive, especially around electrical power, but a lot of this seems like the traditional ways the Federal government gets tricked into spending money. As long as this doesn’t scale too big, I’m not that concerned.
Then there is the play to defeat any attempt at safety regulation, via Federal regulations that actively net interfere with that goal in case any states or countries wanted to try and help. There is clear desirability of a common standard for this, but a voluntary safe harbor preemption, in exchange for various nebulous forms of potential cooperation, cannot be the basis of our entire safety plan. That appears to be the proposal on offer here.

The real vision, the thing I will take away most, is in the rhetoric and presentation, combined with the broader goals, rather than the particular details.

OpenAI now actively wants to be seen as pursuing this kind of obviously disingenuous jingoistic and typically openly corrupt rhetoric, to the extent that their statements are physically painful to read – I dealt with much of that around SB 1047, but this document takes that to the next level and beyond.

OpenAI wants no enforced constraints on their behavior, and they want our money.

OpenAI are telling us who they are. I fully believe them.

Discussion about this post

On the OpenAI Economic Blueprint Read More »

ChatGPT becomes more Siri-like with new scheduled tasks feature

agentic AI, AI, AI Agent, chatgpt, ChatGPT Tasks, openai / Mike M. / January 14, 2025

OpenAI is making ChatGPT work a little more like older digital assistants with a new feature called Tasks, as reported by TechCrunch and others.

Currently in beta, Tasks allows users to direct the chatbot to send reminders or to generate responses to specific prompts at certain times; recurring tasks are also supported.

The feature is available to Plus, Team, and Pro subscribers starting today, while free users don’t have access.

To create a task, users need to select “4o with scheduled tasks” from the model picker and then direct ChatGPT using the same kind of plain language text prompts that drive everything else it does. ChatGPT will sometimes suggest tasks, too, but they won’t go into effect unless the user approves them.

The user can then make changes to assigned tasks through the same chat conversation, or they can use a new Tasks section of the ChatGPT apps to manage all currently assigned items. There’s currently a 10-task limit.

When the time comes to perform an assigned task, the ChatGPT mobile or desktop app will send a notification on schedule.

This update can be seen as OpenAI’s first step into the agentic AI space, where applications built using deep learning can operate relatively independently within certain boundaries, either replacing or easing the day-to-day responsibilities of information workers.

ChatGPT becomes more Siri-like with new scheduled tasks feature Read More »