 |
|
|
|||||
The AI industry hit a significant bend in the road toward artificial general intelligence in 2024. Previously, the stunning intelligence gains that led to chatbots such ChatGPT and Claude had come from supersizing models and the data and computing power used to train them. When the progress from massive scaling leveled off, researchers knew they would need a new strategy, beyond training, to keep moving toward AGI models that are broadly smarter than humans. Starting with OpenAIs pivotal o1 model, researchers began to apply more computing power to the real-time reasoning a model does just after a user prompts it with a problem or question. o1 required more time to produce answers than other models, but its answers were clearly better than those of non-reasoning models. The o1 model rose quickly to the top of the rankings in common benchmark tests, and soon Google DeepMind, Anthropic, DeepSeek and others were training their models for real-time reasoning. The big AI labs would now need even more of the Nvidia GPUs theyd been using for training to support all the real-time reasoning their models would be doing. Thats part of the reason Nvidia saw its stock price rise 171% in 2024. The chip company had been reading the tea leaves and built new features into its GPUs that made them even better suited to real-time inference than before. The result was a new architecture called Blackwell, and new chips called the B100 and B200. Nvidia unveiled the new GPUs in March 2024 and quickly sold its entire 2024 supply to the largest data center operators. Even before the appearance of new reasoning models, some of AIs hottest companies produced state-of-the-art new AI systems. Google DeepMind broke through with a family of natively multi-modal models called Gemini that understand imagery and audio as well as they do language. Anthropic continued to put intense pressure on OpenAI, and was first to release an AI model that could perform task on a users computer. Mistral released impressive new small language models that can run on laptops and even phones with its Ministral 3B and Ministral 8B, as did Microsoft with its Phi-3 and Phi-4 models. And Runway again made its case for the state-of-the-art with its new Gen-3 Alpha video generation models. And all the while a small AI lab in China was quietly developing new AI models, including reasoning models, that would begin sending shockwaves through the AI industry by the end of 2024. DeepSeek#1 on our list of the most innovative companies in Asia-Pacifictrained state-of-the-art models at far lower cost and with far less GPU power than anyone thought possible. It even showed its work through research papers and by open-sourcing its models. DeepSeeks breakthroughs caused some angst, but its fresh thinking and openness will likely spur bigger and faster innovations from the worlds top AI companies by this time next year.1. NvidiaFor arming employees with the tools to get their jobs doneLarge-language model makers continue to pour money into realizing the lofty ambitions of their AI systems. But one company is already reaping the rewards. Nvidia kicked off the AI race by providing the computing power with its market-dominating graphics processing units (GPUs). Now, as its customers chase human-level intelligence, its groundbreaking Blackwell processor and platform is ready. For general model training tasks, the Blackwell processor is up to 2.5 times more powerful than its predecessor, the H100, and requires significantly less energy to operate. The largest data center operators and AI labs, including Google, Meta, Microsoft, OpenAI, Tesla and xAI, are buying or planning to buy Blackwell GPUs by the hundreds of thousands.While recent models from Chinas DeepSeek and Alibaba have made big strides in wringing cutting-edge AI out of older, less powerful Nvidia GPUs, the company isnt just cranking out processors and waiting to see what the world does with them. Instead, its actively building platforms for everything from drug discovery (Clara for Biopharma) to autonomous vehicles (Drive AGX) to video production (Holoscan) to digital twins (Omniverse). By driving AI progress in an ever-growing array of real-world scenarios, its positioning itself for continued growth, even if future models dont need all the computational muscle they can get.Read more about Nvidia, honored as No. 2 on Fast Companys list of the Worlds 50 Most Innovative Companies of 2025.2. OpenAIFor improving AI by giving it more time to thinkSince 2019, OpenAI has continually improved its models by giving them more training data and computing power. The rest of the industry has followed its lead. Now that researchers are seeing diminishing returns from this scaling strategy, OpenAI needed an alternative route toward its goal of creating AI models that are smarter than humans in most tasksin other words, models that can reach artificial general intelligence (AGI).The company found it, and the resulting model is called o1. Instead of scaling up training data and compute time while the model is pre-trained, OpenAI researchers built o1 to spend more time and computing power at inference, when the model is deployed and answering user prompts. While doing this, the model collects and remembers contextual data, both from the user and from relevant data sources. It takes a trial-and-error approach to figuring out the best route to an answer. The model generates PhD-level answers to complex questions, landing it on top of performance benchmark rankings.OpenAI o1 experimental and mini versions are available to ChatGPT Plus users. The company offers a new service called ChatGPT Pro that provides unlimited access to o1 proper for $200 per month. In December of 2024 OpenAI announced o1s successor, o3, and in February 2025 it gave paid users access to o3-mini, a smaller, faster version that excels in science, math, and coding. The biggest impact of OpenAIs new reasoning models is proving to the industry that scaling up computing at inference time may be a viable path to making new intelligence breakthroughs on the way to AGI.3. Google DeepMindFor delivering a truly multi-modal AI modelThe foundational research in AI model architecture and model training that led to the chatbots of today happened at Google in the late 2010s. The search giant developed a large language model-powered chatbot long before ChatGPT, but, as the story goes, it hesitated to expose the technology to the public because of concerns over safety, privacy, and legal risks. That hesitation put Google behind in the AI race that ensued with the launch of ChatGPT.The release of Google DeepMinds Gemini 2.0 in 2024 marked the moment when Google officially caught up. Gemini 2.0 is the first mass market AI model that is natively multimodal, meaning that it can process and generate images, video, audio, and computer code the same way it does text. This makes it possible for the model to watch a video clip, or live video from a phone camera, and quickly analyze and reason about itThe Gemini model is also notable for being able to control other Google services such as Maps and Search. This is Google beginning to flexcombining its AI research with its legacy information and productivity tools. Gemini is one of the first AI models capable of working autonomously and reasoning its way through complex problems on the behalf of the user. The Gemini 2.0 Flash Thinking Experimental model even shows the user the thought process it used to reach an answer. And in December, Google showed off Project Mariner, a Gemini-based agentic AI feature designed to perform tasks such as online grocery shopping on behalf of the user.4. AnthropicFor imbuing Claude with the ability to use a computerGenerative AI has so far been used mainly for a few tasks: writing and summarizing text as well as generating images. The next step is making large language models reason and use tools. Anthropics Computer Use model gave us our first look into that future.Starting with 2024s Claude 3.5 Sonnet, Anthropics model can perceive whats happening on the users screen, including internet content. It can operate a cursor, click buttons, and type in text. In a video, a Claude researcher demonstrated how Claude can fill out a form based on information available at the websites open in the browser tabs. It can complete tasks such as building a personal website or sorting the logistics of a day trip. Its wild to watch the AI work, clicking open new tabs, running searches, and filling in data fields.Right now the model runs slowly and doesnt always get the right answer, but thats likely to change quickly as Anthropic finds and fixes the models pitfalls. Googles aforementioned Project Mariner followed Anthropics lead in December, and OpenAI introduced its own computer use model, Operator, in January 2025. In February 2025, Anthropic moved on to its next major iteration of Claude, Claude 3.7 Sonnet, a larger model that can automatically switch into reasoning mode for hard questions.5. MicrosoftFor pushing the state-of-the-art in small language modelsThe development of Microsofts Phi models started out with a question posed by the companys researchers in 2023: How small can we make a model that shows signs of emerging intelligence? This turns out to be an important moment in the evolution of small language models, or models that perform well in scenarios where memory, processing power, or connectivity are limited, and response speed is important.During 2024, Microsoft released two generations of small models that exhibited reasoning and logic capabilities that hadnt been explicitly trained into them. In April the company released a series of Phi-3 models that excel in language, reasoning, coding, and math benchmarksprobably owing to the fact that Microsoft trained them using synthetic data generated by much larger and more capable LLMs. People downloaded variants of the open-source Phi-3 more than 4.5 million times on HuggingFace during 2024.In late 2024, Microsoft released its Phi-4 small language models, which improved on the Phi-3 models in reasoning-focused tasks and even outperformed OpenAIs GPT-4o on the GPQA (scientific questions) and MATH benchmarks. Microsoft released the model under an open-source and open-weights license, so developers can use it to develop edge models or apps that run on phones or laptops. In less than a month, Phi-4 was downloaded 375,000 times on HuggingFace.6. AmazonFor building a cost-effective rival to Nvidias GPUAmazon AWS recently released a new version of its Tranium processor for AI, called Trainium2, whichin some settingscould challenge the dominance of Nvidia GPUs. The Trainium2 chip is built to supply the massive computing power used in training the largest generative AI models, and for inference time work after the model has been deployed. AWS says Trainium is 30% to 40% more cost-effective in jobs normally done by GPUs.Trainium2 cures the power and software integration deficiencies seen in the first Trainium chip, says Dylan Patel of SemiAnalysis, and puts Amazon on a path toward catching up to Nvidia. (Note that AWS itself remains very reliant on NVIDIA for GPUs.) Unseating Nvidia is difficult because customers get locked in to building on top of Nvidias CUDA software layer, which lets researchers control how their models use the chips resources. Amazon offers its own kernel control software layer called Neuron Kernel Interface (NKI), which, like CUDA, gives researchers granular control over how the chip kernels work together.Patel points out that Trainium2 is still untested at scale. AWS is now building a server cluster with 400,000 Trainium2 chips for Anthropic, which could teach it a lot about making its AI chips work well together in big numbers.7. ArmFor making AI data centers more cost effective and efficientThe British semiconductor designer Arm has long provided the tech industry with the architecture used in chips that power small devices such as phones, sensors, and IoT hardware. This takes on new importance as a new era arrives in which edge device chips will run AI models. But data centers will play a huge role in this evolution as well, often handling some or all of the heaviest AI processing and delivering the results down to the edge device.As data centers expand around the world, the amount of electrical power they use will become a pressing consideration. Thats part of the reason Arms latest Neoverse CPU architecture focuses on efficiency. It offers a 50% performance improvement over earlier generations, and 20% better performance per watt than processors using competing x86 architectures, the company says.Arm says that Amazon, Microsoft, Google, and Oracle have all now adopted Arm Neoverse for both general-purpose computing and CPU-based AI inference and training. For example, in 2024 Micerosoft announced its first custom silicon built for the cloud, the Cobalt 100 processor, was built on Arm Neoverse. Some of the largest AI data centers will rely on NVIDIAs Grace Hopper Superchip, which contains a Hopper GPU and a Grace CPU that uses Neoverse. Arm is reportedly planning to launch its own CPU this year, with Meta as one of its first customers.8. GretelFor feeding the generative AI boom with synthetic dataOver the past year, AI companies have seen diminishing returns from training their models with ever larger amounts of data scraped from the web. Theyve begun focusing less on the quantity of their training data, andmore on its quality. Thats why the theyre are spending more and more on non-public and specialty content that the license from publisher partners. AI researchers also must fill gaps or blind spots within their human-generated or human-annotated training data. For this theyve increasingly turned to synthetic training data generated by special AI models.Gretel gained a higher profile during 2024 by specializing in creating and curating synthetic training data. The company announced the general availability of its flagship product, Gretel Navigator, which lets developers use natural language or SQL prompts to create, augment, edit, and curate synthetic training datasets, or their fine-tuning and testing datasets. The platform has already attracted a community of more than 150,000 developers who have synthesized more than 350 billion pieces of training data.Other industry players have taken notice. Gretel partnered with Google to make its synthetic training data easily available to Google Cloud customers. The company announced a similar partnership with Databricks in June, which allows that companys enterprise customers to access synthetic training data for their models running within the Databricks cloud.9. Mistral AIFor creating small AI models that run in laptops and phonesMistral AI, Frances entry into the generative AI race, has constantly put pressure on OpenAI, Anthropic, and Google at the leading edge of frontier AI model development. Mistral AI released a string of new models containing significant technological advances in 2024, and demonstrated rapid growth in its business, both through direct marketing of its APIs and through strategic partnerships. Early in the year the company produced a pair of open-source models called Mixtral, notable for its innovative use of the mixture of experts architecture, in which only a specialized subset of the models parameters are put in play to handle a query, improving efficiency. In July 2024 Mistral announced Mistral Large 2, which, at 123 billion parameters, showed significant improvements in code generation, maths, reasoning, and function calling. The French company also released Ministral 3B and Ministral 8Bboth of which are smaller models that can run in laptops or phones and store about 50 text pages of context information provided by the user.Mistral has seen success in Europe by marketing itself as a low-cost and flexible alternative to U.S. AI companies like OpenAI. It also continued its expansion into the U.S. enterprise market during 2024. In June, the company raised a $640 million funding round, led by the venture capital firm General Catalyst. The round increased Mistrals valuation to roughly $6.2 billion.10. Fireworks AIFor bringing a plug-and-play model to AI development and deploymentFireworks provides a custom runtime environment that removes some of the considerable engineering work typically associated with building infrastructure for AI deployments. Using the Fireworks platform, enterprises can plug in any of more than 100 AI models, then customize and fine-tune them for their specific use cases.The company introduced new products during 2024 that will equip it to ride some important waves in the AI industry. For one, developers have become more concerned about how quickly AI-powered models and apps respond to user requests. Fireworks debuted its FireAttentionV2, an optimization and quantization software that speeds up the work of models and reduces network latency. Secondly, AI systems are increasingly not single models but pipelines that call different models and tools using APIs. A new FireFunction V2 software works as an orchestrator of all the components in these increasingly complex systems, especially as enterprises launch more autonomous AI apps.Fireworks says it saw a 600% rise in revenue growth in 2024. Its customer base includes Verizon, DoorDash, Uber, Quora, Upwork, and other known names.11. Snorkel AIFor helping businesses prepare their data for use in AI modelsEnterprises have learned that their AI systems are only as good as their data. And Snorkel AI has made an impressive business out of helping enterprises get their proprietary data ready for use in AI models. The companys Snorkel Flow AI data development platform gives companies a cost-efficient way to label and curate their proprietary data so that it can be used to customize and evaluate AI models for their unique business purposes. In 2024, Snorkel added support for images, letting companies train multimodal AI models and image generators using their own proprietary images. It also added retrieval augmented generation (RAG) to its platform, which allowed its customers to retrieve only the most relevant chunks of information from long documentssuch as proprietary knowledge base contentfor use in training the AI. Snorkel Custom, a higher-touch new service level, puts the companys machine learning experts to work on projects directly with the customer.Snorkel says its year-over-year annual bookings doubled during 2024, with triple-digit growth in annual bookings in each of the last three years. Six of the largest banks now use Snorkel Flow, the company says, as well as brands such as Chubb, Wayfair, and Experian.12. CalypsoAIFor letting companies see the logic behind AI decisions in real timeAs AI becomes more instrumental in critical decision-making processes, enterprises are looking for ways to gain more visibility into the workings of the models. That goes double for companies in regulated industries that must constantly watch for bias and other unintended outputs. CalypsoAI was among the first to recognize this emerging need, and it quickly responded with enhanced explainability features in its AI infrastructure platform.Whats surprising about Calypso is the reach of its observability technology. In 2024 the company launched its AI Security Platform, which protects enterprise data by securing, auditing, and monitoring all active generative AI models a company may be using, regardless of the model vendor or whether the model is hosted internally or externally. Calypso also introduced new visualization tools that allow users to see the logic behind AI decisions in real time.The market is responding to Calypsos shift toward AI observability. The company says it saw a 10x inclease in its revenues during 2024, and expects its revenues to grow by another 5x in 2025.13. GalileoFor creating an AI model that can spot hallucinations in other modelsOn the whole, AI systems hallucinate facts and show biases less than they did a year ago. But theyre still prone to these issues, a worrisome situation for any business using AIespecially those in regulated industries such as healthcare and banking. AI development teams use Galileos AI platform to measure, optimize, and monitor the accuracy of their models and apps.In early 2024, after two years of research, Galileo released a suite of evaluation models called Luna that are trained to recognize harmful outputs. The models enable Galileos platform to quickly scrutinize and score the work of an LLM while its in the process of stringing together the tokens that will form its response. This takes the system about only about 200 milliseconds, leaving time to flag and prevent the AIs output from being seen by a user. Its possible to use a standard LLM to perform this task, but its expensive. Galileos purpose-built models are more accurate, more cost-efficient, and, importantly, faster.Galileo says it quadrupled its customer count in 2024, and that it counts Twilio, Reddit, Chegg, Comcast, and JP Morgan Chase as clients. The startup also raised a $68 milion funding round from investors such as HuggingFace CEO Clement Delangue.14. RunawayFor pairing with Lionsgate to create Hollywood-ready AI toolsOne of the great hopesand fearsabout AI is that itll soon be able to generate video thats good enough to change the art and economics of filmmaking forever. The technology took some large steps in the direction of that future during 2024, and one of the main companies pushing the state-of-the-art is the New York-based video generation startup Runway. When the company released its new Gen-3 Alpha model in June 2024, many in the AI community took to X to croon about the much improved believability of the generated video.Runway also made major improvements to its tools for controlling the look of AI video. The model was trained on both images and video, and can create video based on either text or image inputs. The company followed up with Gen-3 Alpha Turbo, a more cost-efficient and speedy version of Gen-3.Hollywood has been watching generative AIs progress closely, and Runway says its now begun producing custom versions of its models for entertainment industry players. It entered into a formal partnership with one such player, Lionsgate Studios, in September 2024. Runway built a custom model for the production company and trained it on Lionsgates film catalog. Runway says the model is meant to help Lionsgates filmmakers, directors and other creatives augment their work while saving time, money and resources. Runway believes its arrangement with Lionsgate could serve as a template for similar deals with other production companies.15. Cerebras SystemsFor taking a bigger-is-better approach to AI chipsAI systems, especially large frontier models, require tremendous amounts of computing power to run at scale. That means that thousands or millions of chips must be wired together to share the workload, but the network connections between the chips can slow things down. Cerebras Systems technology is designed to reap the speed and efficiency benefits of putting a lot of computing power in one really big chip.The companys latest WS-3 (third-generation Wafer Scale Engine) chip, for example, is the size of a dinner plate at 814 square millimeters, 56 times larger than NVIDIAs market-leading H100 chips. The chip packs an astonishing 4 trillion transistors, offers 44 gigabits of memory, and can be clustered together to form supercomputers, such as Condor Galaxy, a constellation of interconnected supercomputers Cerebras is developing with its biggest customer, UAE-based AI and cloud computing company G42.So far Cerebras has found a sweet spot in large research organizations such as Mayo Clinic, Sandia National Laboratories, Lawrence Livermore, and the Los Alamos National Laboratory. The company filed papers for an IPO in September 2024. The prospectus says the companys sales more than tripled to $78.7 million in 2023 and climbed to $136.4 million in the first half of 2024.Explore the full 2025 list of Fast Companys Most Innovative Companies, 609 organizations that are reshaping industries and culture. Weve selected the companies making the biggest impact across 58 categories, including advertising, applied AI, biotech, retail, sustainability, and more.
Category:
E-Commerce
Have you ever finished off your last pickle spear and, craving a little more of that vinegary punch, taken a couple of sips of brine straight from the jar? Or maybe youre more open about your pickle juice habits and like to mix up a pickle martini in the light of day, rather than hunched over your fridge light at 2 a.m. Whatever you prefer, now theres a product designed for exactly those kinds of moments. Claussen, the Chicago-based pickle purveyor, has picked up on the TikTok trend of using pickle brine as a mixer for everything from Diet Coke to pickle cereal, and theyre meeting customers where theyre at with a new drink called Just the Brine. As the name suggests, Just the Brine is an eight-ounce bottle of juice-sans-pickle. The limited-edition product comes in a six-pack, and it debuted for a short time on GoPuff over the weekend in honor of St. Patricks Day (for those who missed out, it’s now available to win on Claussens website while supplies last.) Just the Brine is the latest evolution of a pickle craze that started back in 2022 (remember Sonics pickle slushie?) and has shown a shockingly strong staying power in the cultural zeitgeist. [Photo: Claussen] Care for some pickles with that brine? Since 2022, weve gone from pickle pizza and potato chips to Grillos pickle toothpasteand, judging by TikToks ongoing pickle obsession, it seems like the trend has yet to run its course. Users are finding ways to use the preserved vegetables that even the most ardent pickle fans never couldve imagined, like a pickle fountain or a fried pickle board. The next evolution of the trend, it seems, is to just lose the pickles altogether. Last October, Dua Lipas viral TikTok video mixing Diet Coke with pickle juice sparked a cultural moment, amassing over 12 million views, says Caroline Sheehey, Claussens brand manager. Inspired by her mixture, Claussen responded by seeding a product concept, Just The Brine, on Instagram. The post received nearly 70,000 likes and thousands of comments from fans sharing how they already love Claussens beloved brine and use it in a variety of ways such as after a sports workout, as a brine for their chicken, to help with dehydration as a morning after cure, cocktail mixer, and more. After seeing the fan response, Sheehy says, the team knew they had to make Just the Brine a reality. Claussen is marketing its brine bottles as a kind of dual-purpose product: a mixer to pregame your night out, and an electrolyte beverage for your inevitable hangover the next day. One serving size is two ounces, which contains 630 mg of sodium (about half the sodium content of a standard instant ramen pack.) [Just the Brine] is perfect for pickling at night and using as a mixer in your cocktails or soda, and perfect for unpickling the next morning as a refreshing electrolyte boost, Sheehy says. Its a strange marketing tactic, given that curing your pickle-induced hangover with more pickles seems like the quickest way to never want to set eyes on the color green again. But, lets be honest, the chances that Claussen ever actually adds this stunt product to its permanent line-up are slim to noneso the lucky few who get their hands on it might as well enjoy it via a pickle-fueled rager while it lasts.
Category:
E-Commerce
Her voice cracking with emotion as she stood under the fluorescent lights, Janice Blanock asked her local legislators in southwestern Pennsylvania to take a moment and leaf through the photos of her son that shed handed them. Theres really nothing different that I can say to you that I havent said already over the last several months, she told supervisors for the tiny township of Cecil outside Pittsburgh. I can, however, share these photographs. These are just a few of the many pictures we have of our son Luke, from the time he became ill until before he died. The supervisors were gathered to vote on a zoning ordinance amendment that would greatly increase the required buffer zone between oil and gas drilling operations and homes and schools. The proposed rule mandating a setback of 2,500 feetfive times the distance of the current lawhad originally been proposed as a statewide requirement by Governor Josh Shapiro when he was Pennsylvanias attorney general. A bill based on that recommendation later stalled out when introduced in the state House of Representatives. Blanock, a 30-year resident of Cecil, had a reason to take the issue seriously. Her son waged a three-year battle with a rare type of bone cancer known as Ewings sarcoma and died in 2016 at age 19. Many believe, though theres yet to be demonstrable proof, that his cancer could be tied to oil and gas drillings many carcinogenic pollutants, some that are radioactive. In 2019, a cluster of Ewings sarcoma cases was identified in Washington County, where Cecil is located. Cecils school district was hit particularly hard. The county is home to more than 2,000 natural gas wells and was the 2004 birthplace of the states fracking industry. (Fracking is a process in which sand, water, and chemicals are blasted into the earth to free fossil fuel.) A growing body of peer-reviewed research has linked living near natural gas drilling operations to cancers and respiratory, reproductive, and neurological damage. In 2023, researchers at the University of Pittsburgh and the state Department of Health linked fracking exposure in the region encompassing Cecil to increased risk of asthma and lymphoma. Will you look at the damn picture, Darlene, Blanock urged one supervisor after handing her a photo of Luke. Around an hour later, the zoning ordinance passed and the room erupted with applause. With that, Cecila town of just 15,000 residents and no outsize political powerbecame the first jurisdiction in Pennsylvania to adopt such a restrictive measure, even as similar efforts at the state level have failed. But already it is facing legal challenges from two natural gas companies active in the areaTexas-based fracking company Range Resources, and Colorado-based gas pipeline company MarkWest Liberty Midstream. The Cecil Township Board of Supervisors meets monthly at the towns Municipal Building. [Photo: Audrey Carleton] Under current requirements, natural gas wells in Pennsylvania must be at least 500 feet from buildings and water wells, which environmentalists and medical experts say is not far enough. In 2023, a bill that would have required all new natural gas wells in the state to be located at least 2,500 feetnearly half a milefrom buildings and water wells was slated for a committee vote, but was abruptly killed at the request of Democratic leadership in the state House of Representatives. Three years before that, then-Attorney General Josh Shapiro released a grand jury report calling for a statewide 2,500-foot buffer between human activity and natural gas production. There is one point that is impossible to deny, the grand jury report stated. The closer people happen to live to a massive, industrial drilling complex, the worse it is likely to be for them. While that plea failed to get political traction, environmental groups continue to urge action. For their part, natural gas industry groups have minimized concerns about health risks associated with fracking exposure and have resisted proposals for setbacks or no-drill zones. But despite industry efforts, Cecil has gone its own way. The townships updated oil and gas ordinance prohibits new oil and gas wells from being drilled within 2,500 feet of protected structures, which includes homes, businesses, and religious institutions, and within 5,000 feet of schools and hospitals. Though the ordinance does not call for an outright ban on new drilling, Range Resources contends it would limit fracking in Cecil in such a way that it violates state law. The township argues otherwise: Wells located outside Cecil can still be drilled under the town. The ordinance also imposes additional restrictions on the industry that have generated less debate: It prohibits retention ponds for water used in the fracking process, places new noise restrictions on drilling, and limits work hours on well pads. I was not sure for the longest time that this was going to go this way, said Sarah Martik, a Cecil resident and executive director of the Center for Coalfield Justice, a southwestern Pennsylvania-based nonprofit environmental justice organization. This one thing is as far as weve ever gone, as far as regulating this industry in a way that is protective of our communities. But the road to this outcome was fraught. Documents obtained by Capital & Main through right-to-know requests reveal an up-close look at life in the shale fields, with citizens largely fed up with living alongside the natural gas industry. Noise, bright lights, and shaking at all hours were among the complaints emailed to supervisors in the months ahead of the vote. Here I am once again trying to prepare for another sleepless night, one resident wrote to the supervisors in May. My whole house shakes, my children are disturbed from sleep, my pets are afraid to be out in the yardcan you please help us. “I have SUFFERED from vertigo for years, another resident wrote in June, referring to vibrations from drilling at a nearby well pad that she felt in her home. You know in some places they torture people with this kind of low res hum and vibration. Torturebecause that is what it is. Documents also offer a look at the playbook the industry followed to curry favor among Cecil residents. Over the five years before the ordinance was adopted in 2024, Range Resources, the townships only active natural gas well operator with 34 active wells per state records, donated nearly to $300,000 to the community. The money was disbursed throughout the township, the encompassing school district, and local volunteer first responder organizations, and it was spent on festivals, childrens sports teams, a science fair, and CPR training sessions, according to a spreadsheet obtained by Capital & Main through a right-to-know request. Range Resources did not immediately respond to Capital & Mains request for comment. At least one township supervisor has financial ties to Range Resources. Records show Supervisor Darlene Barni has, for many years, maintained an oil and gas lease with the company; she ultimately recused herself from the final ordinance vote but participated in earlier stages of its development and routinely shares pro-oil and gas posts on Facebook. The company also weighed in at multiple stages during the drafting of the ordinance, using experts to testify against existing science that ties fracking to poor environmental and health outcomes and urging town leaders to refrain from enacting a setback as large as 2,500 feet. At least 92% of Cecil Townships surface property would be excluded from future oil and gas development, an attorney for Range Resources told supervisors in a letter. This would have the effect of limiting residents oil and gas royalty payments, he wrote. The attorney said the setbacks were exceedingly restrictive and inconsistent. Though the company currently has no permits under consideration for new well pads, Range Resources is challenging the ordinance with the townships Zoning Hearing Board. This process could take months, and the challenge is opposed by the township, residents, and several local environmental groups. At issue is whether Cecils ordinance is legal. Its a very, very specific question, said Kara Shirdon, who chairs the Cecil Zoning Hearing Board but recused herself for Range Resources legal challenge to eliminate the appearance of bias (Shirdon has been publicly supportive of the setback ordinance.) Though she said shes confident the ordinance will survive, she believes it will strain the townships resources. I think, honestly, the whole entire thing is because theyre pissed and they want to drain as much money as possible out of the township as punishment for not letting them do what they want to do. * * * Michelle Stonemark moved to Cecil township in 2012 after her parents bought around 30 acres there with the intent of housing their children and grandchildren. Her parents, sister, and family friends all built homes next to one another, in succession. And then it was my turn, Stonemark told Capital & Main. Just as I had gotten the drawing . . . we find out that Range Resources had applied to put a well pad in right behind my new house. With around 30 days notice, she recalled, Stonemark and her family found legal help and learned everything we could about fracking, in order to oppose the project. But their effort failed. We didnt have enough time. We were starting from nothing, she said. Drilling at the pad began in 2020 in the midst of the COVID-19 lockdown, as Stonemark, her husband and three children were stuck at home. Today, the well pad, known as Augustine George, sits just over 500 feet from her home, she said, and routinely rattles the walls and windows and sends fumes into the air. She said she and members of her family often experience headaches, nausea, nosebleeds, and earaches. They can feel vibrations from the well pad in their chests, she said. Flaring would go off at any and all times, during the day, at night, she said. Flaring, which involves burning off excess methane, has been linked to asthma and other respiratory conditions. In response, Stonemark launched a Facebook page she uses to serve as an industry watchdog: She posts photos, videos, and documents relevant to the oil and gas industrys indiscretions, and publicly mourns the future she once envisioned for herself in Cecil. As I stand outside on this beautiful morning I cannot enjoy the day, she wrote in one post in May. A foul odor lingers in the air, and the constant low noises pulsate through my ears and head. Stonemark and her husband are also now attempting to intervene legally ad become a formal party against Range Resources challenge to the setback ordinance. Shirdon said she first caught wind of Range Resources plans for a well pad in 2017, less than a year after moving into her home. Since then, she said shes experienced headaches, sinus and respiratory issues, difficulty concentrating and sleeping, and irritability. The part that people underestimate, I think, is how much anxiety it causes, Shirdon said. Every time you feel the rumble, or every time you get stopped on the road, you start to worry, Are my kids being adversely affected by whats going on here? Merle Lesko has lived in his house nearby for nearly 30 years. Lesko said he and Stonemark often jokingly spar over who lives closer to the Augustine George pad. Salmon pink sound walls, dozens of feet high, poke through a line of trees behind his property. Lesko first urged the township to adopt a new buffer ordinance in early 2024, after regularly recording the decibel level emitted by the Augustine George pad at different locations in his house. He moved his bed and the desk where he works based on the lowest noise reading he found in his residencehis basementjust to escape the vibrations that would rattle his house. The noise was so bad, you could hear or feel the noise over a running lawnmower, he said. Theyve taken so many summers from me.” Though it took months of often impassioned debate, the adoption of Cecils ordinance has added fuel to a fight at the state level, where climate justice organizations are urging environmental regulators to increase the statewide oil and setback of 500 feet. In October, the Protective Buffers Pennsylvania campaign filed a petition with the states Environmental Quality Board, pushing for the adoption of an executive rule that would require a 3,281-foot buffer between fracking wells and buildings and water wellsa setback nearly 1,000 feet wider than in Cecils ordinance. There should be a baseline floor of protection for everybody in the commonwealth, said Lisa Hallowell, senior attorney at the Environmental Integrity Project, an environmental nonprofit that helped author the petition. More than 10% of Pennsylvanians lived within a half mile of an active oil and gas well as of 2022, the petition notes. Many share medical symptomsrashes, cancers, sleep disordersand have seen their water supplies affected by fracking, the petition states. Protective Buffers Pennsylvania has been involved in previous attempts to pass tougher statewide setback rules, including the 2023 bill that died in committee, Hallowell said. These efforts never got far. The Legislature has not had an appetite for that, she noted. Indeed, around the time that the 2023 setback bill was circulating through the Legislature, state Senator Gene Yaw of Williamsport, Republican chair of the Senate Environmental Resources and Energy Committee, questioned the need for the measure at all, saying in a public hearing that he had not heard of any links between fracking and cardiovascular, reproductive, or nervous system damage. Yaw has, separately, disclosed personal income from oil and gas companies EQT and Equinor, and won his reelection to the senate in November after accepting thousands of dollars in campaign donations from the oil and gas industry. A group of Democratic senators has announced that they soon plan to reintroduce the 2,500-foot setback proposal. But that bill will face an uphill battle in a divided Legislature. Janice Blanock at home [Photo: Audrey Carleton] After helping cement the setback proposal as law in Cecil, Blanock now wants to see other communities protected. Were hoping this movement goes far and wide, Blanock said the day after the ordinance passed. I think, just the fact that that happened last night, people will learn about it [and think], If they can do it, why cant we? Several months later, as legal challenges threaten Cecils hard-won victory, Blanock remains resolute. She still chokes back tears when she talks about Luke, and still resents having had her concerns about health risks associated with fracking exposure denied by the industry. Its not just about Luke, she said. This is about my other children, my grandchildren, my community, my family, friends, neighbors. Blanock shares photos and mementos of her son Luke. [Photo: Audrey Carleton] They can appeal it, she said of the natural gas companies challenging the ordinance. And then we can appeal it. Were as strong in our resolve to win this as they are. This piece was originally published by Capital & Main, which reports from California o economic, political, and social issues.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||