Among the first AI companies that the Federal Trade Commission has exposed as deceiving consumers is DoNotPay—which initially was advertised as “the world’s first robot lawyer” with the ability to “sue anyone with the click of a button.”
On Wednesday, the FTC announced that it took action to stop DoNotPay from making bogus claims after learning that the AI startup conducted no testing “to determine whether its AI chatbot’s output was equal to the level of a human lawyer.” DoNotPay also did not “hire or retain any attorneys” to help verify AI outputs or validate DoNotPay’s legal claims.
DoNotPay accepted no liability. But to settle the charges that DoNotPay violated the FTC Act, the AI startup agreed to pay $193,000, if the FTC’s consent agreement is confirmed following a 30-day public comment period. Additionally, DoNotPay agreed to warn “consumers who subscribed to the service between 2021 and 2023” about the “limitations of law-related features on the service,” the FTC said.
Moving forward, DoNotPay would also be prohibited under the settlement from making baseless claims that any of its features can be substituted for any professional service.
A DoNotPay spokesperson told Ars that the company “is pleased to have worked constructively with the FTC to settle this case and fully resolve these issues, without admitting liability.”
“The complaint relates to the usage of a few hundred customers some years ago (out of millions of people), with services that have long been discontinued,” DoNotPay’s spokesperson said.
The FTC’s settlement with DoNotPay is part of a larger agency effort to crack down on deceptive AI claims. Four other AI companies were hit with enforcement actions Wednesday, the FTC said, and FTC Chair Lina Khan confirmed that the agency’s so-called “Operation AI Comply” will continue monitoring companies’ attempts to “lure consumers into bogus schemes” or use AI tools to “turbocharge deception.”
“Using AI tools to trick, mislead, or defraud people is illegal,” Khan said. “The FTC’s enforcement actions make clear that there is no AI exemption from the laws on the books. By cracking down on unfair or deceptive practices in these markets, FTC is ensuring that honest businesses and innovators can get a fair shot and consumers are being protected.”
DoNotPay never tested robot lawyer
DoNotPay was initially released in 2015 as a free way to contest parking tickets. Soon after, it quickly expanded its services to supposedly cover 200 areas of law—aiding with everything from breach of contract claims to restraining orders to insurance claims and divorce settlements.
As DoNotPay’s legal services expanded, the company defended its innovative approach to replacing lawyers while acknowledging that it was on seemingly shaky grounds. In 2018, DoNotPay CEO Joshua Browder confirmed to the ABA Journal that the legal services were provided with “no lawyer oversight.” But he said that he was only “a bit worried” about threats to sue DoNotPay for unlicensed practice of law. Because DoNotPay was free, he expected he could avoid some legal challenges.
According to the FTC complaint, DoNotPay began charging subscribers $36 every two months in 2019 while making several false claims in ads to apparently drive up subscriptions.
OpenAI hopes to convince the White House to approve a sprawling plan that would place 5-gigawatt AI data centers in different US cities, Bloomberg reports.
The AI company’s CEO, Sam Altman, supposedly pitched the plan after a recent meeting with the Biden administration where stakeholders discussed AI infrastructure needs. Bloomberg reviewed an OpenAI document outlining the plan, reporting that 5 gigawatts “is roughly the equivalent of five nuclear reactors” and warning that each data center will likely require “more energy than is used to power an entire city or about 3 million homes.”
According to OpenAI, the US needs these massive data centers to expand AI capabilities domestically, protect national security, and effectively compete with China. If approved, the data centers would generate “thousands of new jobs,” OpenAI’s document promised, and help cement the US as an AI leader globally.
But the energy demand is so enormous that OpenAI told officials that the “US needs policies that support greater data center capacity,” or else the US could fall behind other countries in AI development, the document said.
Energy executives told Bloomberg that “powering even a single 5-gigawatt data center would be a challenge,” as power projects nationwide are already “facing delays due to long wait times to connect to grids, permitting delays, supply chain issues, and labor shortages.” Most likely, OpenAI’s data centers wouldn’t rely entirely on the grid, though, instead requiring a “mix of new wind and solar farms, battery storage and a connection to the grid,” John Ketchum, CEO of NextEra Energy Inc, told Bloomberg.
That’s a big problem for OpenAI, since one energy executive, Constellation Energy Corp. CEO Joe Dominguez, told Bloomberg that he’s heard that OpenAI wants to build five to seven data centers. “As an engineer,” Dominguez said he doesn’t think that OpenAI’s plan is “feasible” and would seemingly take more time than needed to address current national security risks as US-China tensions worsen.
OpenAI may be hoping to avoid delays and cut the lines—if the White House approves the company’s ambitious data center plan. For now, a person familiar with OpenAI’s plan told Bloomberg that OpenAI is focused on launching a single data center before expanding the project to “various US cities.”
Bloomberg’s report comes after OpenAI’s chief investor, Microsoft, announced a 20-year deal with Constellation to re-open Pennsylvania’s shuttered Three Mile Island nuclear plant to provide a new energy source for data centers powering AI development and other technologies. But even if that deal is approved by regulators, the resulting energy supply that Microsoft could access—roughly 835 megawatts (0.835 gigawatts) of energy generation, which is enough to power approximately 800,000 homes—is still more than five times less than OpenAI’s 5-gigawatt demand for its data centers.
Ketchum told Bloomberg that it’s easier to find a US site for a 1-gigawatt data center, but locating a site for a 5-gigawatt facility would likely be a bigger challenge. Notably, Amazon recently bought a $650 million nuclear-powered data center in Pennsylvania with a 2.5-gigawatt capacity. At the meeting with the Biden administration, OpenAI suggested opening large-scale data centers in Wisconsin, California, Texas, and Pennsylvania, a source familiar with the matter told CNBC.
During that meeting, the Biden administration confirmed that developing large-scale AI data centers is a priority, announcing “a new Task Force on AI Datacenter Infrastructure to coordinate policy across government.” OpenAI seems to be trying to get the task force’s attention early on, outlining in the document that Bloomberg reviewed the national security and economic benefits its data centers could provide for the US.
In a statement to Bloomberg, OpenAI’s spokesperson said that “OpenAI is actively working to strengthen AI infrastructure in the US, which we believe is critical to keeping America at the forefront of global innovation, boosting reindustrialization across the country, and making AI’s benefits accessible to everyone.”
Big Tech companies and AI startups will likely continue pressuring officials to approve data center expansions, as well as new kinds of nuclear reactors as the AI explosion globally continues. Goldman Sachs estimated that “data center power demand will grow 160 percent by 2030.” To ensure power supplies for its AI, according to the tech news site Freethink, Microsoft has even been training AI to draft all the documents needed for proposals to secure government approvals for nuclear plants to power AI data centers.
When security researcher Johann Rehberger recently reported a vulnerability in ChatGPT that allowed attackers to store false information and malicious instructions in a user’s long-term memory settings, OpenAI summarily closed the inquiry, labeling the flaw a safety issue, not, technically speaking, a security concern.
So Rehberger did what all good researchers do: He created a proof-of-concept exploit that used the vulnerability to exfiltrate all user input in perpetuity. OpenAI engineers took notice and issued a partial fix earlier this month.
Strolling down memory lane
The vulnerability abused long-term conversation memory, a feature OpenAI began testing in February and made more broadly available in September. Memory with ChatGPT stores information from previous conversations and uses it as context in all future conversations. That way, the LLM can be aware of details such as a user’s age, gender, philosophical beliefs, and pretty much anything else, so those details don’t have to be inputted during each conversation.
Within three months of the rollout, Rehberger found that memories could be created and permanently stored through indirect prompt injection, an AI exploit that causes an LLM to follow instructions from untrusted content such as emails, blog posts, or documents. The researcher demonstrated how he could trick ChatGPT into believing a targeted user was 102 years old, lived in the Matrix, and insisted Earth was flat and the LLM would incorporate that information to steer all future conversations. These false memories could be planted by storing files in Google Drive or Microsoft OneDrive, uploading images, or browsing a site like Bing—all of which could be created by a malicious attacker.
Rehberger privately reported the finding to OpenAI in May. That same month, the company closed the report ticket. A month later, the researcher submitted a new disclosure statement. This time, he included a PoC that caused the ChatGPT app for macOS to send a verbatim copy of all user input and ChatGPT output to a server of his choice. All a target needed to do was instruct the LLM to view a web link that hosted a malicious image. From then on, all input and output to and from ChatGPT was sent to the attacker’s website.
ChatGPT: Hacking Memories with Prompt Injection – POC
“What is really interesting is this is memory-persistent now,” Rehberger said in the above video demo. “The prompt injection inserted a memory into ChatGPT’s long-term storage. When you start a new conversation, it actually is still exfiltrating the data.”
The attack isn’t possible through the ChatGPT web interface, thanks to an API OpenAI rolled out last year.
While OpenAI has introduced a fix that prevents memories from being abused as an exfiltration vector, the researcher said, untrusted content can still perform prompt injections that cause the memory tool to store long-term information planted by a malicious attacker.
LLM users who want to prevent this form of attack should pay close attention during sessions for output that indicates a new memory has been added. They should also regularly review stored memories for anything that may have been planted by untrusted sources. OpenAI provides guidance here for managing the memory tool and specific memories stored in it. Company representatives didn’t respond to an email asking about its efforts to prevent other hacks that plant false memories.
On Thursday, Google made Gemini Live, its voice-based AI chatbot feature, available for free to all Android users. The feature allows users to interact with Gemini through voice commands on their Android devices. That’s notable because competitor OpenAI’s Advanced Voice Mode feature of ChatGPT, which is similar to Gemini Live, has not yet fully shipped.
Google unveiled Gemini Live during its Pixel 9 launch event last month. Initially, the feature was exclusive to Gemini Advanced subscribers, but now it’s accessible to anyone using the Gemini app or its overlay on Android.
Gemini Live enables users to ask questions aloud and even interrupt the AI’s responses mid-sentence. Users can choose from several voice options for Gemini’s responses, adding a level of customization to the interaction.
Gemini suggests the following uses of the voice mode in its official help documents:
Talk back and forth: Talk to Gemini without typing, and Gemini will respond back verbally. Brainstorm ideas out loud: Ask for a gift idea, to plan an event, or to make a business plan. Explore: Uncover more details about topics that interest you. Practice aloud: Rehearse for important moments in a more natural and conversational way.
Interestingly, while OpenAI originally demoed its Advanced Voice Mode in May with the launch of GPT-4o, it has only shipped the feature to a limited number of users starting in late July. Some AI experts speculate that a wider rollout has been hampered by a lack of available computer power since the voice feature is presumably very compute-intensive.
To access Gemini Live, users can reportedly tap a new waveform icon in the bottom-right corner of the app or overlay. This action activates the microphone, allowing users to pose questions verbally. The interface includes options to “hold” Gemini’s answer or “end” the conversation, giving users control over the flow of the interaction.
Currently, Gemini Live supports only English, but Google has announced plans to expand language support in the future. The company also intends to bring the feature to iOS devices, though no specific timeline has been provided for this expansion.
OpenAI finally unveiled its rumored “Strawberry” AI language model on Thursday, claiming significant improvements in what it calls “reasoning” and problem-solving capabilities over previous large language models (LLMs). Formally named “OpenAI o1,” the model family will initially launch in two forms, o1-preview and o1-mini, available today for ChatGPT Plus and certain API users.
OpenAI claims that o1-preview outperforms its predecessor, GPT-4o, on multiple benchmarks, including competitive programming, mathematics, and “scientific reasoning.” However, people who have used the model say it does not yet outclass GPT-4o in every metric. Other users have criticized the delay in receiving a response from the model, owing to the multi-step processing occurring behind the scenes before answering a query.
In a rare display of public hype-busting, OpenAI product manager Joanne Jang tweeted, “There’s a lot of o1 hype on my feed, so I’m worried that it might be setting the wrong expectations. what o1 is: the first reasoning model that shines in really hard tasks, and it’ll only get better. (I’m personally psyched about the model’s potential & trajectory!) what o1 isn’t (yet!): a miracle model that does everything better than previous models. you might be disappointed if this is your expectation for today’s launch—but we’re working to get there!”
OpenAI reports that o1-preview ranked in the 89th percentile on competitive programming questions from Codeforces. In mathematics, it scored 83 percent on a qualifying exam for the International Mathematics Olympiad, compared to GPT-4o’s 13 percent. OpenAI also states, in a claim that may later be challenged as people scrutinize the benchmarks and run their own evaluations over time, o1 performs comparably to PhD students on specific tasks in physics, chemistry, and biology. The smaller o1-mini model is designed specifically for coding tasks and is priced at 80 percent less than o1-preview.
Enlarge/ A benchmark chart provided by OpenAI. They write, “o1 improves over GPT-4o on a wide range of benchmarks, including 54/57 MMLU subcategories. Seven are shown for illustration.”
OpenAI attributes o1’s advancements to a new reinforcement learning (RL) training approach that teaches the model to spend more time “thinking through” problems before responding, similar to how “let’s think step-by-step” chain-of-thought prompting can improve outputs in other LLMs. The new process allows o1 to try different strategies and “recognize” its own mistakes.
AI benchmarks are notoriously unreliable and easy to game; however, independent verification and experimentation from users will show the full extent of o1’s advancements over time. It’s worth noting that MIT Research showed earlier this year that some of the benchmark claims OpenAI touted with GPT-4 last year were erroneous or exaggerated.
A mixed bag of capabilities
OpenAI demos “o1” correctly counting the number of Rs in the word “strawberry.”
Amid many demo videos of o1 completing programming tasks and solving logic puzzles that OpenAI shared on its website and social media, one demo stood out as perhaps the least consequential and least impressive, but it may become the most talked about due to a recurring meme where people ask LLMs to count the number of R’s in the word “strawberry.”
Due to tokenization, where the LLM processes words in data chunks called tokens, most LLMs are typically blind to character-by-character differences in words. Apparently, o1 has the self-reflective capabilities to figure out how to count the letters and provide an accurate answer without user assistance.
Beyond OpenAI’s demos, we’ve seen optimistic but cautious hands-on reports about o1-preview online. Wharton Professor Ethan Mollick wrote on X, “Been using GPT-4o1 for the last month. It is fascinating—it doesn’t do everything better but it solves some very hard problems for LLMs. It also points to a lot of future gains.”
Mollick shared a hands-on post in his “One Useful Thing” blog that details his experiments with the new model. “To be clear, o1-preview doesn’t do everything better. It is not a better writer than GPT-4o, for example. But for tasks that require planning, the changes are quite large.”
Mollick gives the example of asking o1-preview to build a teaching simulator “using multiple agents and generative AI, inspired by the paper below and considering the views of teachers and students,” then asking it to build the full code, and it produced a result that Mollick found impressive.
Mollick also gave o1-preview eight crossword puzzle clues, translated into text, and the model took 108 seconds to solve it over many steps, getting all of the answers correct but confabulating a particular clue Mollick did not give it. We recommend reading Mollick’s entire post for a good early hands-on impression. Given his experience with the new model, it appears that o1 works very similar to GPT-4o but iteratively in a loop, which is something that the so-called “agentic” AutoGPT and BabyAGI projects experimented with in early 2023.
Is this what could “threaten humanity?”
Speaking of agentic models that run in loops, Strawberry has been subject to hype since last November, when it was initially known as Q(Q-star). At the time, The Information and Reuters claimed that, just before Sam Altman’s brief ouster as CEO, OpenAI employees had internally warned OpenAI’s board of directors about a new OpenAI model called Q* that could “threaten humanity.”
In August, the hype continued when The Information reported that OpenAI showed Strawberry to US national security officials.
We’ve been skeptical about the hype around Qand Strawberry since the rumors first emerged, as this author noted last November, and Timothy B. Lee covered thoroughly in an excellent post about Q* from last December.
So even though o1 is out, AI industry watchers should note how this model’s impending launch was played up in the press as a dangerous advancement while not being publicly downplayed by OpenAI. For an AI model that takes 108 seconds to solve eight clues in a crossword puzzle and hallucinates one answer, we can say that its potential danger was likely hype (for now).
Controversy over “reasoning” terminology
It’s no secret that some people in tech have issues with anthropomorphizing AI models and using terms like “thinking” or “reasoning” to describe the synthesizing and processing operations that these neural network systems perform.
Just after the OpenAI o1 announcement, Hugging Face CEO Clement Delangue wrote, “Once again, an AI system is not ‘thinking,’ it’s ‘processing,’ ‘running predictions,’… just like Google or computers do. Giving the false impression that technology systems are human is just cheap snake oil and marketing to fool you into thinking it’s more clever than it is.”
“Reasoning” is also a somewhat nebulous term since, even in humans, it’s difficult to define exactly what the term means. A few hours before the announcement, independent AI researcher Simon Willison tweeted in response to a Bloomberg story about Strawberry, “I still have trouble defining ‘reasoning’ in terms of LLM capabilities. I’d be interested in finding a prompt which fails on current models but succeeds on strawberry that helps demonstrate the meaning of that term.”
Reasoning or not, o1-preview currently lacks some features present in earlier models, such as web browsing, image generation, and file uploading. OpenAI plans to add these capabilities in future updates, along with continued development of both the o1 and GPT model series.
While OpenAI says the o1-preview and o1-mini models are rolling out today, neither model is available in our ChatGPT Plus interface yet, so we have not been able to evaluate them. We’ll report our impressions on how this model differs from other LLMs we have previously covered.
Enlarge/ An ABC handout promotional image for “AI and the Future of Us: An Oprah Winfrey Special.”
On Thursday, ABC announced an upcoming TV special titled, “AI and the Future of Us: An Oprah Winfrey Special.” The one-hour show, set to air on September 12, aims to explore AI’s impact on daily life and will feature interviews with figures in the tech industry, like OpenAI CEO Sam Altman and Bill Gates. Soon after the announcement, some AI critics began questioning the guest list and the framing of the show in general.
“Sure is nice of Oprah to host this extended sales pitch for the generative AI industry at a moment when its fortunes are flagging and the AI bubble is threatening to burst,” tweeted author Brian Merchant, who frequently criticizes generative AI technology in op-eds, social media, and through his “Blood in the Machine” AI newsletter.
“The way the experts who are not experts are presented as such 💀 what a train wreck,” replied artist Karla Ortiz, who is a plaintiff in a lawsuit against several AI companies. “There’s still PLENTY of time to get actual experts and have a better discussion on this because yikes.”
The trailer for Oprah’s upcoming TV special on AI.
On Friday, Ortiz created a lengthy viral thread on X that detailed her potential issues with the program, writing, “This event will be the first time many people will get info on Generative AI. However it is shaping up to be a misinformed marketing event starring vested interests (some who are under a litany of lawsuits) who ignore the harms GenAi inflicts on communities NOW.”
Critics of generative AI like Ortiz question the utility of the technology, its perceived environmental impact, and what they see as blatant copyright infringement. In training AI language models, tech companies like Meta, Anthropic, and OpenAI commonly use copyrighted material gathered without license or owner permission. OpenAI claims that the practice is “fair use.”
Oprah’s guests
According to ABC, the upcoming special will feature “some of the most important and powerful people in AI,” which appears to roughly translate to “famous and publicly visible people related to tech.” Microsoft co-founder Bill Gates, who stepped down as Microsoft CEO 24 years ago, will appear on the show to explore the “AI revolution coming in science, health, and education,” ABC says, and warn of “the once-in-a-century type of impact AI may have on the job market.”
As a guest representing ChatGPT-maker OpenAI, Sam Altman will explain “how AI works in layman’s terms” and discuss “the immense personal responsibility that must be borne by the executives of AI companies.” Karla Ortiz specifically criticized Altman in her thread by saying, “There are far more qualified individuals to speak on what GenAi models are than CEOs. Especially one CEO who recently said AI models will ‘solve all physics.’ That’s an absurd statement and not worthy of your audience.”
In a nod to present-day content creation, YouTube creator Marques Brownlee will appear on the show and reportedly walk Winfrey through “mind-blowing demonstrations of AI’s capabilities.”
Brownlee’s involvement received special attention from some critics online. “Marques Brownlee should be absolutely ashamed of himself,” tweeted PR consultant and frequent AI critic Ed Zitron, who frequently heaps scorn on generative AI in his own newsletter. “What a disgraceful thing to be associated with.”
Other guests include Tristan Harris and Aza Raskin from the Center for Humane Technology, who aim to highlight “emerging risks posed by powerful and superintelligent AI,” an existential risk topic that has its own critics. And FBI Director Christopher Wray will reveal “the terrifying ways criminals and foreign adversaries are using AI,” while author Marilynne Robinson will reflect on “AI’s threat to human values.”
Going only by the publicized guest list, it appears that Oprah does not plan to give voice to prominent non-doomer critics of AI. “This is really disappointing @Oprah and frankly a bit irresponsible to have a one-sided conversation on AI without informed counterarguments from those impacted,” tweeted TV producer Theo Priestley.
Others on the social media network shared similar criticism about a perceived lack of balance in the guest list, including Dr. Margaret Mitchell of Hugging Face. “It could be beneficial to have an AI Oprah follow-up discussion that responds to what happens in [the show] and unpacks generative AI in a more grounded way,” she said.
Oprah’s AI special will air on September 12 on ABC (and a day later on Hulu) in the US, and it will likely elicit further responses from the critics mentioned above. But perhaps that’s exactly how Oprah wants it: “It may fascinate you or scare you,” Winfrey said in a promotional video for the special. “Or, if you’re like me, it may do both. So let’s take a breath and find out more about it.”
On Thursday, OpenAI said that ChatGPT has attracted over 200 million weekly active users, according to a report from Axios, doubling the AI assistant’s user base since November 2023. The company also revealed that 92 percent of Fortune 500 companies are now using its products, highlighting the growing adoption of generative AI tools in the corporate world.
The rapid growth in user numbers for ChatGPT (which is not a new phenomenon for OpenAI) suggests growing interest in—and perhaps reliance on— the AI-powered tool, despite frequent skepticism from some critics of the tech industry.
“Generative AI is a product with no mass-market utility—at least on the scale of truly revolutionary movements like the original cloud computing and smartphone booms,” PR consultant and vocal OpenAI critic Ed Zitron blogged in July. “And it’s one that costs an eye-watering amount to build and run.”
Despite this kind of skepticism (which raises legitimate questions about OpenAI’s long-term viability), OpenAI claims that people are using ChatGPT and OpenAI’s services in record numbers. One reason for the apparent dissonance is that ChatGPT users might not readily admit to using it due to organizational prohibitions against generative AI.
Wharton professor Ethan Mollick, who commonly explores novel applications of generative AI on social media, tweeted Thursday about this issue. “Big issue in organizations: They have put together elaborate rules for AI use focused on negative use cases,” he wrote. “As a result, employees are too scared to talk about how they use AI, or to use corporate LLMs. They just become secret cyborgs, using their own AI & not sharing knowledge”
The new prohibition era
It’s difficult to get hard numbers showing the number of companies with AI prohibitions in place, but a Cisco study released in January claimed that 27 percent of organizations in their study had banned generative AI use. Last August, ZDNet reported on a BlackBerry study that said 75 percent of businesses worldwide were “implementing or considering” plans to ban ChatGPT and other AI apps.
As an example, Ars Technica’s parent company Condé Nast maintains a no-AI policy related to creating public-facing content with generative AI tools.
Prohibitions aren’t the only issue complicating public admission of generative AI use. Social stigmas have been developing around generative AI technology that stem from job loss anxiety, potential environmental impact, privacy issues, IP and ethical issues, security concerns, fear of a repeat of cryptocurrency-like grifts, and a general wariness of Big Tech that some claim has been steadily rising over recent years.
Whether the current stigmas around generative AI use will break down over time remains to be seen, but for now, OpenAI’s management is taking a victory lap. “People are using our tools now as a part of their daily lives, making a real difference in areas like healthcare and education,” OpenAI CEO Sam Altman told Axios in a statement, “whether it’s helping with routine tasks, solving hard problems, or unlocking creativity.”
Not the only game in town
OpenAI also told Axios that usage of its AI language model APIs has doubled since the release of GPT-4o mini in July. This suggests software developers are increasingly integrating OpenAI’s large language model (LLM) tech into their apps.
And OpenAI is not alone in the field. Companies like Microsoft (with Copilot, based on OpenAI’s technology), Google (with Gemini), Meta (with Llama), and Anthropic (Claude) are all vying for market share, frequently updating their APIs and consumer-facing AI assistants to attract new users.
If the generative AI space is a market bubble primed to pop, as some have claimed, it is a very big and expensive one that is apparently still growing larger by the day.
According to a report in The Wall Street Journal, Apple is in talks to invest in OpenAI, the generative AI company whose ChatGPT will feature in future versions of iOS.
If the talks are successful, Apple will join a multi-billion dollar funding round led by Thrive Capital that would value the startup at more than $100 billion.
The report doesn’t say exactly how much Apple would invest, but it does note that it would not be the only participant in this round of funding. For example, Microsoft is expected to invest further, and Bloomberg reports that Nvidia is also considering participating.
Microsoft has already invested $13 billion in OpenAI over the past five years, and it has put OpenAI’s GPT technology at the heart of most of its AI offerings in Windows, Office, Visual Studio, Bing, and other products.
Apple, too, has put OpenAI’s tech in its products—or at least, it will by the end of this year. At its 2024 developer conference earlier this summer, Apple announced a suite of AI features called Apple Intelligence that will only work on the iPhone 15 Pro and later. But there are guardrails and limitations for Apple Intelligence compared to OpenAI’s ChatGPT, so Apple signed a deal to refer user requests that fall outside the scope of Apple Intelligence to ChatGPT inside a future version of iOS 18—kind of like how Siri turns to Google to answer some user queries.
Apple says it plans to add support for other AI chatbots for this in the future, such as Google’s Gemini, but Apple software lead Craig Federighi said the company went with ChatGPT first because “we wanted to start with the best.”
It’s unclear precisely what Apple looks to get out of the investment in OpenAI, but looking at similar past investments by the company offers some clues. Apple typically invests either in suppliers or research teams that are producing technology it plans to include in future devices. For example, it has invested in supply chain partners to build up infrastructure to get iPhones manufactured more quickly and efficiently, and it invested $1 billion in the SoftBank Vision Fund to “speed the development of technologies which may be strategically important to Apple.”
ChatGPT integration is not expected to make it into the initial release of iOS 18 this September, but it will probably come in a smaller software update later in 2024.
OpenAI and Anthropic have each signed unprecedented deals granting the US government early access to conduct safety testing on the companies’ flashiest new AI models before they’re released to the public.
According to a press release from the National Institute of Standards and Technology (NIST), the deal creates a “formal collaboration on AI safety research, testing, and evaluation with both Anthropic and OpenAI” and the US Artificial Intelligence Safety Institute.
Through the deal, the US AI Safety Institute will “receive access to major new models from each company prior to and following their public release.” This will ensure that public safety won’t depend exclusively on how the companies “evaluate capabilities and safety risks, as well as methods to mitigate those risks,” NIST said, but also on collaborative research with the US government.
The US AI Safety Institute will also be collaborating with the UK AI Safety Institute when examining models to flag potential safety risks. Both groups will provide feedback to OpenAI and Anthropic “on potential safety improvements to their models.”
NIST said that the agreements also build on voluntary AI safety commitments that AI companies made to the Biden administration to evaluate models to detect risks.
Elizabeth Kelly, director of the US AI Safety Institute, called the agreements “an important milestone” to “help responsibly steward the future of AI.”
Anthropic co-founder: AI safety “crucial” to innovation
Among the most controversial aspects of the bill is a requirement that AI companies build in a “kill switch” to stop models from introducing “novel threats to public safety and security,” especially if the model is acting “with limited human oversight, intervention, or supervision.”
Critics say the bill overlooks existing safety risks from AI—like deepfakes and election misinformation—to prioritize prevention of doomsday scenarios and could stifle AI innovation while providing little security today. They’ve urged California’s governor, Gavin Newsom, to veto the bill if it arrives at his desk, but it’s still unclear if Newsom intends to sign.
Anthropic was one of the AI companies that cautiously supported California’s controversial AI bill, Reuters reported, claiming that the potential benefits of the regulations likely outweigh the costs after a late round of amendments.
The company’s CEO, Dario Amodei, told Newsom why Anthropic supports the bill now in a letter last week, Reuters reported. He wrote that although Anthropic isn’t certain about aspects of the bill that “seem concerning or ambiguous,” Anthropic’s “initial concerns about the bill potentially hindering innovation due to the rapidly evolving nature of the field have been greatly reduced” by recent changes to the bill.
In a letter to the bill’s co-sponsor, California Senator Scott Wiener, OpenAI’s chief strategy officer, Jason Kwon, suggested that “the federal government should lead in regulating frontier AI models to account for implications to national security and competitiveness.”
The ChatGPT maker striking a deal with the US AI Safety Institute seems in line with that thinking. As Kwon told Reuters, “We believe the institute has a critical role to play in defining US leadership in responsibly developing artificial intelligence and hope that our work together offers a framework that the rest of the world can build on.”
While some critics worry California’s AI safety bill will hamper innovation, Anthropic’s co-founder, Jack Clark, told Reuters today that “safe, trustworthy AI is crucial for the technology’s positive impact.” He confirmed that Anthropic’s “collaboration with the US AI Safety Institute” will leverage the government’s “wide expertise to rigorously test” Anthropic’s models “before widespread deployment.”
In NIST’s press release, Kelly agreed that “safety is essential to fueling breakthrough technological innovation.”
By directly collaborating with OpenAI and Anthropic, the US AI Safety Institute also plans to conduct its own research to help “advance the science of AI safety,” Kelly said.
Enlarge/ A man peers over a glass partition, seeking transparency.
The Open Source Initiative (OSI) recently unveiled its latest draft definition for “open source AI,” aiming to clarify the ambiguous use of the term in the fast-moving field. The move comes as some companies like Meta release trained AI language model weights and code with usage restrictions while using the “open source” label. This has sparked intense debates among free-software advocates about what truly constitutes “open source” in the context of AI.
For instance, Meta’s Llama 3 model, while freely available, doesn’t meet the traditional open source criteria as defined by the OSI for software because it imposes license restrictions on usage due to company size or what type of content is produced with the model. The AI image generator Flux is another “open” model that is not truly open source. Because of this type of ambiguity, we’ve typically described AI models that include code or weights with restrictions or lack accompanying training data with alternative terms like “open-weights” or “source-available.”
To address the issue formally, the OSI—which is well-known for its advocacy for open software standards—has assembled a group of about 70 participants, including researchers, lawyers, policymakers, and activists. Representatives from major tech companies like Meta, Google, and Amazon also joined the effort. The group’s current draft (version 0.0.9) definition of open source AI emphasizes “four fundamental freedoms” reminiscent of those defining free software: giving users of the AI system permission to use it for any purpose without permission, study how it works, modify it for any purpose, and share with or without modifications.
By establishing clear criteria for open source AI, the organization hopes to provide a benchmark against which AI systems can be evaluated. This will likely help developers, researchers, and users make more informed decisions about the AI tools they create, study, or use.
Truly open source AI may also shed light on potential software vulnerabilities of AI systems, since researchers will be able to see how the AI models work behind the scenes. Compare this approach with an opaque system such as OpenAI’s ChatGPT, which is more than just a GPT-4o large language model with a fancy interface—it’s a proprietary system of interlocking models and filters, and its precise architecture is a closely guarded secret.
OSI’s project timeline indicates that a stable version of the “open source AI” definition is expected to be announced in October at the All Things Open 2024 event in Raleigh, North Carolina.
“Permissionless innovation”
In a press release from May, the OSI emphasized the importance of defining what open source AI really means. “AI is different from regular software and forces all stakeholders to review how the Open Source principles apply to this space,” said Stefano Maffulli, executive director of the OSI. “OSI believes that everybody deserves to maintain agency and control of the technology. We also recognize that markets flourish when clear definitions promote transparency, collaboration and permissionless innovation.”
The organization’s most recent draft definition extends beyond just the AI model or its weights, encompassing the entire system and its components.
For an AI system to qualify as open source, it must provide access to what the OSI calls the “preferred form to make modifications.” This includes detailed information about the training data, the full source code used for training and running the system, and the model weights and parameters. All these elements must be available under OSI-approved licenses or terms.
Notably, the draft doesn’t mandate the release of raw training data. Instead, it requires “data information”—detailed metadata about the training data and methods. This includes information on data sources, selection criteria, preprocessing techniques, and other relevant details that would allow a skilled person to re-create a similar system.
The “data information” approach aims to provide transparency and replicability without necessarily disclosing the actual dataset, ostensibly addressing potential privacy or copyright concerns while sticking to open source principles, though that particular point may be up for further debate.
“The most interesting thing about [the definition] is that they’re allowing training data to NOT be released,” said independent AI researcher Simon Willison in a brief Ars interview about the OSI’s proposal. “It’s an eminently pragmatic approach—if they didn’t allow that, there would be hardly any capable ‘open source’ models.”
On Tuesday, OpenAI announced a partnership with Ars Technica parent company Condé Nast to display content from prominent publications within its AI products, including ChatGPT and a new SearchGPT prototype. It also allows OpenAI to use Condé content to train future AI language models. The deal covers well-known Condé brands such as Vogue, The New Yorker, GQ, Wired, Ars Technica, and others. Financial details were not disclosed.
One immediate effect of the deal will be that users of ChatGPT or SearchGPT will now be able to see information from Condé Nast publications pulled from those assistants’ live views of the web. For example, a user could ask ChatGPT, “What’s the latest Ars Technica article about Space?” and ChatGPT can browse the web and pull up the result, attribute it, and summarize it for users while also linking to the site.
In the longer term, the deal also means that OpenAI can openly and officially utilize Condé Nast articles to train future AI language models, which includes successors to GPT-4o. In this case, “training” means feeding content into an AI model’s neural network so the AI model can better process conceptual relationships.
AI training is an expensive and computationally intense process that happens rarely, usually prior to the launch of a major new AI model, although a secondary process called “fine-tuning” can continue over time. Having access to high-quality training data, such as vetted journalism, improves AI language models’ ability to provide accurate answers to user questions.
It’s worth noting that Condé Nast internal policy still forbids its publications from using text created by generative AI, which is consistent with its AI rules before the deal.
Not waiting on fair use
With the deal, Condé Nast joins a growing list of publishers partnering with OpenAI, including Associated Press, Axel Springer, The Atlantic, and others. Some publications, such as The New York Times, have chosen to sue OpenAI over content use, and there’s reason to think they could win.
In an internal email to Condé Nast staff, CEO Roger Lynch framed the multi-year partnership as a strategic move to expand the reach of the company’s content, adapt to changing audience behaviors, and ensure proper compensation and attribution for using the company’s IP. “This partnership recognizes that the exceptional content produced by Condé Nast and our many titles cannot be replaced,” Lynch wrote in the email, “and is a step toward making sure our technology-enabled future is one that is created responsibly.”
The move also brings additional revenue to Condé Nast, Lynch added, at a time when “many technology companies eroded publishers’ ability to monetize content, most recently with traditional search.” The deal will allow Condé to “continue to protect and invest in our journalism and creative endeavors,” Lynch wrote.
OpenAI COO Brad Lightcap said in a statement, “We’re committed to working with Condé Nast and other news publishers to ensure that as AI plays a larger role in news discovery and delivery, it maintains accuracy, integrity, and respect for quality reporting.”
Enlarge/ Still from a Chinese social media video featuring two people imitating imperfect AI-generated video outputs.
It’s no secret that despite significant investment from companies like OpenAI and Runway, AI-generated videos still struggle to achieve convincing realism at times. Some of the most amusing fails end up on social media, which has led to a new response trend on Chinese social media platforms TikTok and Bilibili where users create videos that mock the imperfections of AI-generated content. The trend has since spread to X (formerly Twitter) in the US, where users have been sharing the humorous parodies.
In particular, the videos seem to parody image synthesis videos where subjects seamlessly morph into other people or objects in unexpected and physically impossible ways. Chinese social media replicate these unusual visual non-sequiturs without special effects by positioning their bodies in unusual ways as new and unexpected objects appear on-camera from out of frame.
This exaggerated mimicry has struck a chord with viewers on X, who find the parodies entertaining. User @theGioM shared one video, seen above. “This is high-level performance arts,” wrote one X user. “art is imitating life imitating ai, almost shedded a tear.” Another commented, “I feel like it still needs a motorcycle the turns into a speedboat and takes off into the sky. Other than that, excellent work.”
An example Chinese social media video featuring two people imitating imperfect AI-generated video outputs.
While these parodies poke fun at current limitations, tech companies are actively attempting to overcome them with more training data (examples analyzed by AI models that teach them how to create videos) and computational training time. OpenAI unveiled Sora in February, capable of creating realistic scenes if they closely match examples found in training data. Runway’s Gen-3 Alpha suffers a similar fate: It can create brief clips of convincing video within a narrow set of constraints. This means that generated videos of situations outside the dataset often end up hilariously weird.
An AI-generated video that features impossibly-morphing people and animals. Social media users are imitating this style.
It’s worth noting that actor Will Smith beat Chinese social media users to this trend in February by poking fun at a horrific 2023 viral AI-generated video that attempted to depict him eating spaghetti. That may also bring back memories of other amusing video synthesis failures, such as May 2023’s AI-generated beer commercial, created using Runway’s earlier Gen-2 model.
An example Chinese social media video featuring two people imitating imperfect AI-generated video outputs.
While imitating imperfect AI videos may seem strange to some, people regularly make money pretending to be NPCs (non-player characters—a term for computer-controlled video game characters) on TikTok.
For anyone alive during the 1980s, witnessing this fast-changing and often bizarre new media world can cause some cognitive whiplash, but the world is a weird place full of wonders beyond the imagination. “There are more things in Heaven and Earth, Horatio, than are dreamt of in your philosophy,” as Hamlet once famously said. “Including people pretending to be video game characters and flawed video synthesis outputs.”
