Robots equipped with artificial intelligence capable of separating a water bottle from a diaper could bring the world closer to a cooler planet, or so Matanya Horowitz believes.
Horowitz, an engineer, founded AMP Robotics after completing a PhD in robotics at Caltech. The Colorado company, founded in 2014, designs robots with artificial intelligence to identify recyclable materials such as office paper, aluminum cans and plastic bottles in waste piles. The goal is to automate the recycling process, which often relies on manual labor.
The benefits are two-fold: more recyclable materials that can be sold to processors and less recyclable materials for the landfill.
“Ultimately, we’re trying to make recycling a more profitable business with the goal of more widespread adoption of recycling around the world,” Horowitz told Business Insider.
AMP Robotics is much more than that 44,500 climate technology startups have sprung up since 2010. Investors invested $70.1 billion in climate technology last year, up 89% from 2021. According to HolonIQ Global Impact Intelligence. Ever since AI became the darling of the tech world, some of these startups have embraced it.
While AI tools like Horowitz’s waste-sorting robots have been around for years, the advent of generative AI tools like OpenAI’s ChatGPT has rusted away. A discussion on using AI to solve social problems.
He 2023 Report of the UN Intergovernmental Panel on Climate Change It says cutting-edge technology such as AI can “increase energy efficiency” and “encourage the adoption” of renewable energy, which, in theory, could reduce carbon emissions and spur economic growth.
However, some machine learning researchers and climate activists point out that Business Insider If the tools are not used carefully, using AI to solve the climate crisis could do more harm than good. AI tools are complex, error-prone, and can require significant amounts of natural resources to operate, factors that slow development. Achieving net zero emissions by 2050It aims to limit the global average temperature to a critical limit of 1.5°C.
AMP Robotics raised more than this 176 million dollars in early May, according to PitchBook, and Horowitz says he’s deployed “hundreds of robots” around the world. According to him, the robots have helped his customers sort through large amounts of waste quickly. AMP Robotics representative Carling Spellhawk explains Business Insider From late 2021 to May 2023, robots have saved “1.6 million metric tons of greenhouse gas emissions,” the equivalent of “taking 326,000 cars off the road” over the same period.
However, Horowitz admits that his AI robots aren’t perfect.
“We expect there will be a learning process over the next 10 years,” Horowitz says.
AI is the best way to collect quality climate data
Climate change has already “caused significant damage” to some ecosystems, with “increasingly irreversible losses”. says the IPCC in its 2023 report summary. It has left millions of people food and water insecure, disproportionately harming low-income households, indigenous peoples and vulnerable communities in places like Africa and South America, the UN agency said.
Sasha Lucioni, Research Scientist and Head of Climate at the Hugging Face AI Platform, explains Business Insider AI can be used to understand the extent of the most affected.
“Artificial intelligence is a very powerful tool to process that data and help society say, ‘Okay, let’s focus on this particular aspect as a short-term solution,'” explains Lucioni, a founding member of the nonprofit Enterprise Profit. Climate Change AI.
Lucioni stands for Rainforest Link, a bio-sound monitoring tool that deploys AI-powered sensors in rainforests to detect sounds like chainsaw sounds. Park rangers can use this data to investigate noise, which could prevent illegal deforestation, the organization says Contributes to climate change.
David Rolnick, a computer science professor at McGill University who studies AI applications to the climate crisis, agrees. According to The investigation he cooperated withAI can “distill data into actionable information” to calculate the financial impact of environmental damage, identify the most carbon-emitting parts of supply chains, and forecast a company’s energy needs.
“Utilizing that data, even in relatively simple ways, can amplify the action in ways that manual actions sometimes can’t,” he explains. Business Insider Rolnick, co-founder and president of Climate Change AI.
Banks, manufacturers and healthcare companies have turned to Persephone, a carbon management platform that uses AI to measure its customers’ carbon footprint.
Demand for Persephone’s services is likely due to pressure from governments, investors and corporate stakeholders. to reduce their carbon footprintAs explained Business Insider James Newsom, the company’s chief data officer.
Newsom says many companies don’t have the tools or know-how to manually identify emission sources in their operations, which he says is complicated and time-consuming.
By applying AI to the process, customers can identify outliers in their data and help them “recommend changes” about which areas of the business need to reduce their emissions. For example, an airline can use Persephone’s software to calculate the amount of carbon emitted by burning jet fuel during flights and use this information to adjust its energy habits.
In turn, companies can develop climate action plans cheaper and faster.
“Technology has really democratized and helped people in organizations who didn’t have knowledge,” Newsome says.
But researchers warn that AI could do more harm than good
Researchers, activists and climate technology managers agree that AI alone cannot cool the planet.
Rolnick says the hype surrounding the technology may lead companies to use advanced AI tools, when in reality “low-tech” options could be used, such as sensors that detect energy levels or insulation. Green to reduce overall carbon emissions.
“There may be situations where AI is overkill,” Rolnick says. “It’s very easy for people to go for a more sophisticated technology instead of a less technological option, even if the latter is more useful.”
Some climate activists agree.
The founder of 350.org, a global environmental nonprofit, points out that tools like solar panels and wind turbines can reduce carbon emissions. Calculating emissions “isn’t as complex as it needs AI,” McKibben summarizes. Business Insider.
At the same time, focusing solely on artificial intelligence as a solution to climate change can obscure a global problem: Climate change is linked to global economic growth, climate activist Helena Norberg-Hodge, who criticizes market solutions for artificial intelligence, told Business Insider. Climate crisis.
“Before looking at the benefits of measuring AI, we need to look at what we’re measuring and why,” says Norberg-Hodge.
Researchers agree that AI could be used in ways that could harm the climate. Oil and gas companies, for example, are using AI and advanced analytics to “accelerate exploration and extraction,” which could create up to $425 billion in added value for the sector by 2025, according to one estimate. A Greenpeace report.
Deployment of AI tools also consumes a lot of energy. Large linguistic models like GPT, such as OpenAI’s ChatGPT-based models, are “super computationally intensive” and less stable than simpler technologies, Luccioni says, because training requires a large amount of water.
Even AI climate technology executives see limits to their impact
AI has its limitations, even for those responsible for climate technologies.
Horowitz, CEO of AMP Robotics, says his robots don’t always sort recyclables accurately, and Newsom, chief data officer at Persefoni, says the quality of environmental data varies from customer to customer, making it difficult to accurately understand the full scale. Your carbon footprint.
Despite these limitations, Horowitz, Newome, and other researchers agree that AI tools can be useful as part of a broader, comprehensive climate solution involving multiple stakeholders, including governments, local communities, and climate scientists from various disciplines.
Without the participation of different groups, AI developers can “work in silos,” running the risk of models being misunderstood, misused and inaccessible, especially to those most vulnerable to the climate crisis, Luccioni says.
“You have to have people who understand AI, who understand its purpose, and who are often affected by the technology,” Rolnick says.