 |
|
|
|||||
Welcome to AI Decoded, Fast Companys weekly newsletter that breaks down the most important news in the world of AI. You can sign up to receive this newsletter every week here. Why OpenAIs new open-weight models matter OpenAI is opening up again. The companys release of two open-weight modelsgpt-oss-120b and gpt-oss-20bthis month marks a major shift from its 2019 pivot away from transparency, when it began keeping its most advanced research under wraps after a breakthrough in model scaling and compute. Now, with GPT-5 on the horizon, OpenAI is signaling a returnat least in partto its original ethos. These new models come with all their internal weights exposed, meaning developers can inspect and fine-tune how they work. That doesnt make them open-source in the strictest sensethe training data and source code remain closedbut it does make them more accessible and adaptable than anything OpenAI has offered in years. The move matters, not just because of the models themselves, but because of whos behind them. OpenAI is still the dominant force in generative AI, with ChatGPT as its flagship consumer product. When a leader of that stature starts releasing open models, it sends a signal across the industry. Open models are here to stay, says Anyscale cofounder Robert Nishihara. Now that OpenAI is competing on this front, the landscape will get much more competitive and we can expect to see better open models. Enterprisesespecially ones in regulated industries like healthcare or financiallike to build on open-source models so that they can tailor them for their needs, and so they can run the models on in-house servers or in private clouds rather than undertaking the high cost and security risks of sending their (possibly sensitive or proprietary) data out to a third-party LLM such as OpenAIs GPT-4.5, Anthropics Claude, or Googles Gemini. OpenAIs oss models are licensed under Apache 2.0, meaning developers can use, modify, and even commercialize them, as long as they credit OpenAI and waive any patent claims. None of that would matter if the models werent state of the art, but they are. The larger gpt-oss-120b (120 billion parameters) model matches OpenAIs o4-mini on core reasoning benchmarks while running on a single graphics processing unit (GPU), OpenAI says. The smaller gpt-oss-20b model performs on par with the companys o3-mini, and is compact enough to run on edge devices with just 16 GB of memory (like a high-end laptop). That small size matters a lot. Many in the industry believe that small models running on personal devices could be the wave of the future. On-device models, after all, dont have to connect to the cloud to process data, so they are more secure and can keep data private more easily. Small models are also often trained to do a relatively narrow task (like quality inspection in a factory or language translation from a phone). The release could also accelerate the broader ecosystem of open AI infrastructure. The more popular open models become, the more important open-source infrastructure for deploying those models becomes, Nishihara says. Were seeing the rise of open models complemented by the emergence of high-quality open-source infrastructure for training and serving those modelsthis includes projects like Ray and vLLM. Theres also a geopolitical subtext. The Trump administration has increasingly framed AI as a strategic asset in its rivalry with China, pushing American companies to shape global norms and infrastructure. Open-weight models from a top U.S. labbuilt to run on Nvidia chipscould spread quickly across regions like Africa and the Middle East, countering the rise of free Chinese models tuned for Huawei hardware. Its a soft-power play, not unlike the U.S. dollars dominance as a global currency. Googles new Genie 3 world models could wild new forms of gaming, entertainment With the right prompt, AI models can generate words, voices, music, images, video, and other things. And the quality of those generations continues to grow. Google DeepMind has pushed the boundaries even further with its world models, capable of generating live, interactive environments that users can navigate and modify in real time. Words alone dont fully capture the capabilities of DeepMinds new Genie 3 model. A demo video shows a number of lifelike worlds (a desert, a scuba diving scene, a living room, a canal city, etc.) At one point, the user adds a whimsical element to the canal city world by writing the prompt: A man in a chicken suit emerges from the left of the shot and runs down the towpath hugging the wall. And the man in the chicken suit immediately appears in the world. Then, the user drops a dinosaur into the nearby canal. Splash. The most obvious application of this kind of AI is in gaming, where a model could generate an endless stream of environments and game scenarios for the gamer. Its a natural research focus for DeepMind, which focused its early AI research on video game environments. The potential for world modeling is enormous. Future versions of the Genie model could enable choose your adventure experiences in video or AR formats, where storytelling adapts dynamically to the viewers preferences, interests, and impulses. As Google notes, companies working on self-driving cars or robotics could also benefit, using these models to simulate real-world conditions that would be costly or impractical to recreate physically. The AI industry responds to AI tool abuse by students As the new school year approaches, educators and parents continue to worry that students are using AI tools to do their schoolwork for them. The danger is that students can rely heavily on AI t generate answers to questions, while failing to learn all the contextual stuff they would encounter during the process of finding answers on their own. A growing body of research suggests that relying on AI harms overall academic performance. Now OpenAI and Google have each responded to this worrisome situation by releasing special study modes inside their respective AI chatbots. OpenAIs tool is called ChatGPT study mode, while Google offers a similar feature within its Gemini chatbot called Guided Learning. The tools format and features seem remarkably similar. Both break down complex problems into smaller chunks and then walk the student through them using a question-and-answer approach. Google says its questions are designed to teach students the how and why behind a topic, encouraging learning throughout the exchange. OpenAI says its tool uses Socratic questioning, hints, and self-reflection prompts to guide understanding and promote active learning. Both OpenAI and Google say that the teaching approach and format are based on research by learning experts. Still, the student is ultimately in control of what AI tools they use. OpenAI says that users can easily toggle between regular chatbot mode and the study mode. Google says it believes students need AI for both traditional question searches and for guided study. So these new learning tools may provide an alternative mode of learning using AI, but theyre not likely to significantly shift the argument around AIs threat to real learning. More AI coverage from Fast Company: Google wants you to be a citizen data scientist Reviving this government agency could be the key to U.S. tech dominance Cloudflare vs. Perplexity: a web scraping war with big implications for AI What the White House Action Plan on AI gets right and wrong about bias Want exclusive reporting and trend analysis on technology, business innovation, the future of work, and design? Sign up for Fast Company Premium.
Category:
E-Commerce
The Trump administration has been talking to drugmakers about ways to raise prices of medicines in Europe and elsewhere in order to cut drug costs in the United States, according to a White House official and three pharmaceutical industry sources. U.S. officials told drug companies it would support their international negotiations with governments if they adopt “most favored nation” pricing under which U.S. drug costs match the lower rates offered to other wealthy countries, the White House official said. The U.S. is currently negotiating bilateral trade deals and setting tariff rates on the sector. The Trump administration has asked some companies for ideas on raising prices abroad, two of the sources said, describing multiple meetings over several months aimed at lowering U.S. prices without triggering cuts to research and development spending drugmakers insist would result. The White House official called the effort collaborative, saying both sides were seeking advice from each other. The U.S. pays more for prescription drugs than any other country, often nearly three times as much as other developed nations. President Donald Trump has repeatedly said he wants to narrow this gap to stop Americans from being “ripped off.” The previously unreported discussions reflect the challenges Trump faces to achieve that goal, and are the backdrop to the letters he sent last week to CEOs of 17 major drugmakers, urging them to cut U.S. prices to match those paid overseas. Unlike in the U.S., where market forces determine drug prices, European governments typically negotiate directly with companies to set prices for their national healthcare systems. Anna Kaltenboeck, a health economist at Verdant Research, said European nations have leverage to drive pricing and are sometimes willing to walk away from purchasing medicines they deem too expensive. Drugmakers generate most of their sales in the U.S. The Pharmaceutical Research and Manufacturers of America the industry’s main lobby group has always argued that cutting U.S. prices would stifle innovation by lowering R&D spending. PhRMA declined to comment on the private meetings. Kaltenboeck said past studies had shown that drugmakers made enough money in the U.S. to more than fund their entire global R&D spends. “Prices can come down in the United States without being increased in other countries, and we can still get innovation,” she said. TOP PRIORITY Despite the Trump administration’s tariff threats and pressure to move more manufacturing to the U.S., the push to raise European drug prices is its top priority in discussions with industry, according to a senior executive at a European drugmaker, who spoke on condition of anonymity about the confidential meetings. “This is the key conversation right now with PhRMA and every company getting that message from Pennsylvania Avenue to a point that we are already executing on it,” the executive said, referring to the White House address. The company had already met with European governments on the issue, the executive added. An E.U. Commission spokesperson said it is in regular contact with the pharma industry and pointed to an agreement with the U.S. that should it impose tariffs on pharmaceuticals, they would be capped at 15%. When asked how the administration would support international drug price negotiations, the White House official referred Reuters to Trump’s most favored nation executive order from May. That order directed trade officials to pursue trade and legal action against countries keeping drug prices below fair market value. In last week’s letters, Trump complained that since the May executive order, most industry proposals had simply shifted blame for high prices or requested policy changes that would result in billions in industry handouts. A second source, a pharmaceutical executive who was not authorized to speak on the matter, said the Trump administration has been continually meeting with representatives of his company and had discussed strategies for raising drug prices internationally. “There’s a big push from the administration to drive up prices outside the U.S.,” the executive said. The executive said the Trump administration had been looking at using trade talks with the UK and EU as leverage, and considered pressuring countries to spend a higher percentage of GDP on new medicines or offering tariff breaks in exchange for higher drug spending. It was understood that the UK deal specifically aims to get the country to ramp up investment in branded medicines over time, the executive said. A spokesperson for the UK government said it would continue to work closely with the U.S. and its own pharmaceutical industry to understand the possible impact of any changes to drug pricing, without commenting on the trade talks. In April, over 30 industry CEOs including those from AstraZeneca, Bayer and Novo Nordisk signed a letter to European Union President Ursula von der Leyen saying Europe needed to rethink its pricing policies. “It’s going to be very difficult for a country that already has the ability to control what it spends to go in the other direction,” Kaltenboeck said, “and it doesn’t make much sense for them politically.” Patrick Wingrove and Maggie Fick, Reuters
Category:
E-Commerce
Japanese technology conglomerate SoftBank Group Corp. posted a 421.8 billion yen ($2.9 billion) profit in the April-June quarter, rebounding from a loss a year earlier as its investments benefited from the craze for artificial intelligence. Quarterly sales at Tokyo-based SoftBank Group, which invests heavily in AI companies like Nvidia and OpenAI, rose 7% to 1.8 trillion yen ($12 billion), the company said Thursday. SoftBank’s loss in April-June 2024 was 174 billion yen. The company’s fortunes tend to fluctuate because it invests in a range of ventures through its Vision Funds, a move that carries risks. The groups founder, Masayoshi Son, has emphasized that he sees a vibrant future in AI. SoftBank has also invested in Arm Holdings and Taiwan Semiconductor Manufacturing Co. Both companies, which produce computer chips, have benefitted from the growth of AI. The era is definitely AI, and we are focused on AI, SoftBank senior executive Yoshimitsu Goto told reporters. An investment company goes through its ups and downs, but we are recently seeing steady growth. Some of SoftBank’s other investments also have paid off big. An example is Coupang, an e-commerce company known as the Amazon of South Korea, because it started out in Seoul. Coupang now operates in the U.S. and other Asian nations. Goto said preparations for an IPO for PayPay, a kind of cashless payment system, were going well. The company has already held IPOs for Chime, a U.S. neobank that provides banking services for low-credit consumers, and for Etoro, a personal investment platform. SoftBank Group stock, which has risen from a year ago, finished 1.3% higher on the Tokyo Stock Exchange after its earnings results were announced. Yuri Kageyama, AP business writer
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||