Nature Climate Change, Published online: 14 March 2025; doi:10.1038/s41558-025-02281-6

Small step funding models fit better for climate research

This past January the new administration issued an executive order on Artificial Intelligence (AI), taking the place of the now rescinded Biden-era order, calling for a new AI Action Plan tasked with “unburdening” the current AI industry to stoke innovation and remove “engineered social agendas” from the industry. This new action plan for the president is currently being developed and open to public comments to the National Science Foundation (NSF).

EFF answered with a few clear points: First, government procurement of decision-making (ADM) technologies must be done with transparency and public accountability—no secret and untested algorithms should decide who keeps their job or who is denied safe haven in the United States. Second, Generative AI policy rules must be narrowly focused and proportionate to actual harms, with an eye on protecting other public interests. And finally, we shouldn't entrench the biggest companies and gatekeepers with AI licensing schemes.

Government Automated Decision Making

US procurement of AI has moved with remarkable speed and an alarming lack of transparency. By wasting money on systems with no proven track record, this procurement not only entrenches the largest AI companies, but risks infringing the civil liberties of all people subject to these automated decisions.

These harms aren’t theoretical, we have already seen a move to adopt experimental AI tools in policing and national security, including immigration enforcement. Recent reports also indicate the Department of Government Efficiency (DOGE) intends to apply AI to evaluate federal workers, and use the results to make decisions about their continued employment.

Automating important decisions about people is reckless and dangerous. At best these new AI tools are ineffective nonsense machines which require more labor to correct inaccuracies, but at worst result in irrational and discriminatory outcomes obscured by the blackbox nature of the technology.

Instead, the adoption of such tools must be done with a robust public notice-and-comment practice as required by the Administrative Procedure Act. This process helps weed out wasteful spending on AI snake oil, and identifies when the use of such AI tools are inappropriate or harmful.

Additionally, the AI action plan should favor tools developed under the principles of free and open-source software. These principles are essential for evaluating the efficacy of these models, and ensure they uphold a more fair and scientific development process. Furthermore, more open development stokes innovation and ensures public spending ultimately benefits the public—not just the most established companies.

Don’t Enable Powerful Gatekeepers

Spurred by the general anxiety about Generative AI, lawmakers have drafted sweeping regulations based on speculation, and with little regard for the multiple public interests at stake. Though there are legitimate concerns, this reactionary approach to policy is exactly what we warned against back in 2023.

For example, bills like NO FAKES and NO AI Fraud expand copyright laws to favor corporate giants over everyone else’s expression. NO FAKES even includes a scheme for a DMCA-like notice takedown process, long bemoaned by creatives online for encouraging broader and automated online censorship. Other policymakers propose technical requirements like watermarking that are riddled with practical points of failure.

Among these dubious solutions is the growing prominence of AI licensing schemes which limit the potential of AI development to the highest bidders. This intrusion on fair use creates a paywall protecting only the biggest tech and media publishing companies—cutting out the actual creators these licenses nominally protect. It’s like helping a bullied kid by giving them more lunch money to give their bully.

This is the wrong approach. Looking for easy solutions like expanding copyright, hurts everyone. Particularly smaller artists, researchers, and businesses who cannot compete with the big gatekeepers of industry. AI has threatened the fair pay and treatment of creative labor, but sacrificing secondary use doesn’t remedy the underlying imbalance of power between labor and oligopolies.

People have a right to engage with culture and express themselves unburdened by private cartels. Policymakers should focus on narrowly crafted policies to preserve these rights, and keep rulemaking constrained to tested solutions addressing actual harms.

You can read our comments here.

EFF’s most important platform for welcoming everyone to join us in our fight for a better digital future is our website, eff.org. We thank Fastly for their generous in-kind contribution of services helping keep EFF’s website online.

Eff.org was first registered in 1990, just three months after the organization was founded, and long before the web was an essential part of daily life. Our website and the fight for digital rights grew rapidly alongside each other. However, along with rising threats to our freedoms online, threats to our site have also grown.

It takes a village to keep eff.org online in 2025. Every day our staff work tirelessly to protect the site from everything from DDoS attacks to automated hacking attempts, and everything in between. As AI has taken off, so have crawlers and bots that scrape content to train LLMs, sometimes without respecting rate limits we’ve asked them to observe. Newly donated security add-ons from Fastly help us automate DDoS prevention and rate limiting, preventing our servers from getting overloaded when misbehaving visitors abuse our sites. Fastly also caches the content from our site around the globe, meaning that visitors from all over the world can access eff.org and our other sites quickly and easily.

EFF is member-supported by people who share our vision for a better digital future. We thank Fastly for showing their support for our mission to ensure that technology supports freedom, justice, and innovation for all people of the world with an in-kind gift of their full suite of services.

More than 60,000 tons of plastic makes the journey down the Amazon River to the Atlantic Ocean every year. And that doesn’t include what finds its way to the river’s banks, or the microplastics ingested by the region’s abundant and diverse wildlife.

It’s easy to demonize plastic, but it has been crucial in developing the society we live in today. Creating materials that have the benefits of plastics while reducing the harms of traditional production methods is a goal of chemical engineering and materials science labs the world over, including that of Bradley Olsen, the Alexander and I. Michael Kasser (1960) Professor of Chemical Engineering at MIT.

Olsen, a Fulbright Amazonia scholar and the faculty lead of MIT-Brazil, works with communities to develop alternative plastics solutions that can be derived from resources within their own environments.

“The word that we use for this is co-design,” says Olsen. “The idea is, instead of engineers just designing something independently, they engage and jointly design the solution with the stakeholders.”

In this case, the stakeholders were small businesses around Manaus in the Brazilian state of Amazonas curious about the feasibility of bioplastics and other alternative packaging.

“Plastics are inherent to modern life and actually perform key functions and have a really beautiful chemistry that we want to be able to continue to leverage, but we want to do it in a way that is more earth-compatible,” says Desirée Plata, MIT associate professor of civil and environmental engineering.

That’s why Plata joined Olsen in creating the course 1.096/10.496 (Design of Sustainable Polymer Systems) in 2021. Now, as a Global Classroom offering under the umbrella of MISTI since 2023, the class brings MIT students to Manaus during the three weeks of Independent Activities Period (IAP).

“In my work running the Global Teaching Labs in Brazil since 2016, MIT students collaborate closely with Brazilian undergraduates,” says Rosabelli Coelho-Keyssar, managing director of MIT-Brazil and MIT-Amazonia, who also runs MIT’s Global Teaching Labs program in Brazil. “This peer-learning model was incorporated into the Global Classroom in Manaus, ensuring that MIT and Brazilian students worked together throughout the course.”

The class leadership worked with climate scientist and MIT alumnus Carlos Nobre PhD ’83, who facilitated introductions to faculty at the Universidade Estadual de Amazonas (UAE), the state university of Amazonas. The group then scouted businesses in the Amazonas region who would be interested in partnering with the students.

“In the first year, it was Comunidade Julião, a community of people living on the edge of the Tarumã Mirim River west of Manaus,” says Olsen. “This year, we worked with Comunidade Para Maravilha, a community living in the dry land forest east of Manaus.”

A tailored solution

Plastic, by definition, is made up of many small carbon-based molecules, called monomers, linked by strong bonds into larger molecules called polymers. Linking different monomers and polymers in different ways creates different plastics — from trash bags to a swimming pool float to the dashboard of a car. Plastics are traditionally made from petroleum byproducts that are easy to link together, stable, and plentiful.

But there are ways to reduce the use of petroleum-based plastics. Packaging can be made from materials found within the local ecosystem, as was the focus of the 2024 class. Or carbon-based monomers can be extracted from high-starch plant matter through a number of techniques, the goal of the 2025 cohort. But plants that grow well in one location might not in another. And bioplastic production facilities can be tricky to install if the necessary resources aren’t immediately available.

“We can design a whole bunch of new sustainable chemical processes, use brand new top-of-the-line catalysts, but if you can’t actually implement them sustainably inside an environment, it falls short on a lot of the overall goals,” says Brian Carrick, a PhD candidate in the Olsen lab and a teaching assistant for the 2025 course offering.

So, identifying local candidates and tailoring the process is key. The 2025 MIT cohort collaborated with students from throughout the Amazonas state to explore the local flora, study its starch content in the lab, and develop a new plastic-making process — all within the three weeks of IAP.

“It’s easy when you have projects like this to get really locked into the MIT vacuum of just doing what sounds really cool, which isn’t always effective or constructive for people actually living in that environment,” says Claire Underwood, a junior chemical-biological engineering major who took the class. “That’s what really drew me into the project, being able to work with people in Brazil.”

The 31 students visited a protected area of the Amazon rainforest on Day One. They also had chances throughout IAP to visit the Amazon River, where the potential impact of their work became clear as they saw plastic waste collecting on its banks.

“That was a really cool aspect to the class, for sure, being able to actually see what we were working towards protecting and what the goal was,” says Underwood.

They interviewed stakeholders, such as farmers who could provide the feedstock and plastics manufacturers who could incorporate new techniques. Then, they got into the classroom, where massive intellectual ground was covered in a crash course on the sustainable design process, the nitty gritty of plastic production, and the Brazilian cultural context on how building such an industry would affect the community. For the final project, they separated into teams to craft preliminary designs of process and plant using a simplified model of these systems.

Connecting across boundaries

Working in another country brought to the fore how interlinked policy, culture, and technical solutions are.

“I know nothing about economics, and especially not Brazilian economics and politics,” says Underwood. But one of the Brazilian students in her group was a management and finance major. “He was super helpful when we were trying to source things and account for inflation and things like that — knowing what was feasible, and not just academically feasible.”

Before they parted at the end of IAP, each team presented their proposals to a panel of company representatives and Brazilian MIT alumni who chose first-, second-, and third-place winners. While more research is needed before comfortably implementing the ideas, the experience seemed to generate legitimate interest in creating a local bioplastics production facility.

Understanding sustainable design concepts and how to do interdisciplinary work is an important skill to learn. Even if these students don’t wind up working on bioplastics in the heart of the Amazon, being able to work with people of different perspectives — be it a different discipline or a different culture — is valuable in virtually every field.

“The exchange of knowledge across different fields and cultures is essential for developing innovative and sustainable solutions to global challenges such as climate change, waste management, and the development of eco-friendly materials,” says Taisa Sampaio, a PhD candidate in materials chemistry at UEA and a co-instructor for the course. “Programs like this are crucial in preparing professionals who are more aware and better equipped to tackle future challenges.”

Right now, Olsen and Plata are focused on harnessing the deep well of connections and resources they have around Manaus, but they hope to develop that kind of network elsewhere to expand this sustainable design exploration to other regions of the world.

“A lot of sustainability solutions are hyperlocal,” says Plata. “Understanding that not all locales are exactly the same is really powerful and important when we’re thinking about sustainability challenges. And it’s probably where we've gone wrong with the one-size-fits-all or silver-bullet solution — seeking that we’ve been doing for the past many decades.”

Collaborations for the 2026 trip are still in development but, as Olsen says, “we hope this is an experience we can continue to offer long into the future, based on how positive it has been for our students and our Brazilian partners.”

Many companies invest heavily in hiring talent to create the high-performance library code that underpins modern artificial intelligence systems. NVIDIA, for instance, developed some of the most advanced high-performance computing (HPC) libraries, creating a competitive moat that has proven difficult for others to breach.

But what if a couple of students, within a few months, could compete with state-of-the-art HPC libraries with a few hundred lines of code, instead of tens or hundreds of thousands?

That’s what researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have shown with a new programming language called Exo 2.

Exo 2 belongs to a new category of programming languages that MIT Professor Jonathan Ragan-Kelley calls “user-schedulable languages” (USLs). Instead of hoping that an opaque compiler will auto-generate the fastest possible code, USLs put programmers in the driver's seat, allowing them to write “schedules” that explicitly control how the compiler generates code. This enables performance engineers to transform simple programs that specify what they want to compute into complex programs that do the same thing as the original specification, but much, much faster.

One of the limitations of existing USLs (like the original Exo) is their relatively fixed set of scheduling operations, which makes it difficult to reuse scheduling code across different “kernels” (the individual components in a high-performance library).

In contrast, Exo 2 enables users to define new scheduling operations externally to the compiler, facilitating the creation of reusable scheduling libraries. Lead author Yuka Ikarashi, an MIT PhD student in electrical engineering and computer science and CSAIL affiliate, says that Exo 2 can reduce total schedule code by a factor of 100 and deliver performance competitive with state-of-the-art implementations on multiple different platforms, including Basic Linear Algebra Subprograms (BLAS) that power many machine learning applications. This makes it an attractive option for engineers in HPC focused on optimizing kernels across different operations, data types, and target architectures.

“It’s a bottom-up approach to automation, rather than doing an ML/AI search over high-performance code,” says Ikarashi. “What that means is that performance engineers and hardware implementers can write their own scheduling library, which is a set of optimization techniques to apply on their hardware to reach the peak performance.”

One major advantage of Exo 2 is that it reduces the amount of coding effort needed at any one time by reusing the scheduling code across applications and hardware targets. The researchers implemented a scheduling library with roughly 2,000 lines of code in Exo 2, encapsulating reusable optimizations that are linear-algebra specific and target-specific (AVX512, AVX2, Neon, and Gemmini hardware accelerators). This library consolidates scheduling efforts across more than 80 high-performance kernels with up to a dozen lines of code each, delivering performance comparable to, or better than, MKL, OpenBLAS, BLIS, and Halide.

Exo 2 includes a novel mechanism called “Cursors” that provides what they call a “stable reference” for pointing at the object code throughout the scheduling process. Ikarashi says that a stable reference is essential for users to encapsulate schedules within a library function, as it renders the scheduling code independent of object-code transformations.

“We believe that USLs should be designed to be user-extensible, rather than having a fixed set of operations,” says Ikarashi. “In this way, a language can grow to support large projects through the implementation of libraries that accommodate diverse optimization requirements and application domains.”

Exo 2’s design allows performance engineers to focus on high-level optimization strategies while ensuring that the underlying object code remains functionally equivalent through the use of safe primitives. In the future, the team hopes to expand Exo 2’s support for different types of hardware accelerators, like GPUs. Several ongoing projects aim to improve the compiler analysis itself, in terms of correctness, compilation time, and expressivity.

Ikarashi and Ragan-Kelley co-authored the paper with graduate students Kevin Qian and Samir Droubi, Alex Reinking of Adobe, and former CSAIL postdoc Gilbert Bernstein, now a professor at the University of Washington. This research was funded, in part, by the U.S. Defense Advanced Research Projects Agency (DARPA) and the U.S. National Science Foundation, while the first author was also supported by Masason, Funai, and Quad Fellowships.

Please join EFF for the next segment of EFFecting Change, our livestream series covering digital privacy and free speech. 

EFFecting Change Livestream Series:
Is There Hope for Social Media?
Thursday, March 20th
12:00 PM - 1:00 PM Pacific - Check Local Time
This event is LIVE and FREE!

Users are frustrated with legacy social media companies. Is it possible to effectively build the kinds of communities we want online while avoiding the pitfalls that have driven people away?

Join our panel featuring EFF Civil Liberties Director David Greene, EFF Director for International Freedom of Expression Jillian York, Mastodon's Felix Hlatky, Bluesky's Emily Liu, and Spill's Kenya Parham as they explore the future of free expression online and why social media might still be worth saving.

We hope you and your friends can join us live! Be sure to spread the word, and share our past livestreams. Please note that all events will be recorded for later viewing on our YouTube page.

Want to make sure you don’t miss our next livestream? Here’s a link to sign up for updates about this series: eff.org/ECUpdates.

In January, Meta made targeted changes to its hateful conduct policy that would allow dehumanizing statements to be made about certain vulnerable groups. More specifically, Meta’s hateful conduct policy now contains the following text:

People sometimes use sex- or gender-exclusive language when discussing access to spaces often limited by sex or gender, such as access to bathrooms, specific schools, specific military, law enforcement, or teaching roles, and health or support groups. Other times, they call for exclusion or use insulting language in the context of discussing political or religious topics, such as when discussing transgender rights, immigration, or homosexuality. Finally, sometimes people curse at a gender in the context of a romantic break-up. Our policies are designed to allow room for these types of speech. 

The revision of this policy timed to Trump’s second election demonstrates that the company is focused on allowing more hateful speech against specific groups, with a noticeable and particular focus on enabling more speech challenging LGBTQ+ rights. For example, the revised policy removed previous prohibitions on comparing people to inanimate objects, feces, and filth based on their protected characteristics, such as sexual identity.

In response, LGBTQ+ rights organization AllOut gathered social justice groups and civil society organizations, including EFF, to demand that Meta immediately reverse the policy changes. By normalizing such speech, Meta risks increasing hate and discrimination against LGBTQ+ people on Facebook, Instagram and Threads. 

The campaign is supported by the following partners: All Out, Global Project Against Hate and Extremism (GPAHE), Electronic Frontier Foundation (EFF), EDRi - European Digital Rights, Bits of Freedom, SUPERRR Lab, Danes je nov dan, Corporación Caribe Afirmativo, Fundación Polari, Asociación Red Nacional de Consejeros, Consejeras y Consejeres de Paz LGBTIQ+, La Junta Marica, Asociación por las Infancias Transgénero, Coletivo LGBTQIAPN+ Somar, Coletivo Viveração, and ADT - Associação da Diversidade Tabuleirense, Casa Marielle Franco Brasil, Articulação Brasileira de Gays - ARTGAY, Centro de Defesa dos Direitos da Criança e do Adolescente Padre, Marcos Passerini-CDMP, Agência Ambiental Pick-upau, Núcleo Ypykuéra, Kurytiba Metropole, ITTC - Instituto Terra, Trabalho e Cidadania. 

Sign the AllOut petition (external link) and tell Meta: Stop hate speech against LGBT+ people!

If Meta truly values freedom of expression, we urge it to redirect its focus to empowering some of its most marginalized speakers, rather than empowering only their detractors and oppressive voices.

2006 AT&T whistleblower Mark Klein has died.

Converting one type of cell to another — for example, a skin cell to a neuron — can be done through a process that requires the skin cell to be induced into a “pluripotent” stem cell, then differentiated into a neuron. Researchers at MIT have now devised a simplified process that bypasses the stem cell stage, converting a skin cell directly into a neuron.

Working with mouse cells, the researchers developed a conversion method that is highly efficient and can produce more than 10 neurons from a single skin cell. If replicated in human cells, this approach could enable the generation of large quantities of motor neurons, which could potentially be used to treat patients with spinal cord injuries or diseases that impair mobility.

“We were able to get to yields where we could ask questions about whether these cells can be viable candidates for the cell replacement therapies, which we hope they could be. That’s where these types of reprogramming technologies can take us,” says Katie Galloway, the W. M. Keck Career Development Professor in Biomedical Engineering and Chemical Engineering.

As a first step toward developing these cells as a therapy, the researchers showed that they could generate motor neurons and engraft them into the brains of mice, where they integrated with host tissue.

Galloway is the senior author of two papers describing the new method, which appear today in Cell Systems. MIT graduate student Nathan Wang is the lead author of both papers.

From skin to neurons

Nearly 20 years ago, scientists in Japan showed that by delivering four transcription factors to skin cells, they could coax them to become induced pluripotent stem cells (iPSCs). Similar to embryonic stem cells, iPSCs can be differentiated into many other cell types. This technique works well, but it takes several weeks, and many of the cells don’t end up fully transitioning to mature cell types.

“Oftentimes, one of the challenges in reprogramming is that cells can get stuck in intermediate states,” Galloway says. “So, we’re using direct conversion, where instead of going through an iPSC intermediate, we’re going directly from a somatic cell to a motor neuron.”

Galloway’s research group and others have demonstrated this type of direct conversion before, but with very low yields — fewer than 1 percent. In Galloway’s previous work, she used a combination of six transcription factors plus two other proteins that stimulate cell proliferation. Each of those eight genes was delivered using a separate viral vector, making it difficult to ensure that each was expressed at the correct level in each cell.

In the first of the new Cell Systems papers, Galloway and her students reported a way to streamline the process so that skin cells can be converted to motor neurons using just three transcription factors, plus the two genes that drive cells into a highly proliferative state.

Using mouse cells, the researchers started with the original six transcription factors and experimented with dropping them out, one at a time, until they reached a combination of three — NGN2, ISL1, and LHX3 — that could successfully complete the conversion to neurons.

Once the number of genes was down to three, the researchers could use a single modified virus to deliver all three of them, allowing them to ensure that each cell expresses each gene at the correct levels.

Using a separate virus, the researchers also delivered genes encoding p53DD and a mutated version of HRAS. These genes drive the skin cells to divide many times before they start converting to neurons, allowing for a much higher yield of neurons, about 1,100 percent.

“If you were to express the transcription factors at really high levels in nonproliferative cells, the reprogramming rates would be really low, but hyperproliferative cells are more receptive. It’s like they’ve been potentiated for conversion, and then they become much more receptive to the levels of the transcription factors,” Galloway says.

The researchers also developed a slightly different combination of transcription factors that allowed them to perform the same direct conversion using human cells, but with a lower efficiency rate — between 10 and 30 percent, the researchers estimate. This process takes about five weeks, which is slightly faster than converting the cells to iPSCs first and then turning them into neurons.

Implanting cells

Once the researchers identified the optimal combination of genes to deliver, they began working on the best ways to deliver them, which was the focus of the second Cell Systems paper.

They tried out three different delivery viruses and found that a retrovirus achieved the most efficient rate of conversion. Reducing the density of cells grown in the dish also helped to improve the overall yield of motor neurons. This optimized process, which takes about two weeks in mouse cells, achieved a yield of more than 1,000 percent.

Working with colleagues at Boston University, the researchers then tested whether these motor neurons could be successfully engrafted into mice. They delivered the cells to a part of the brain known as the striatum, which is involved in motor control and other functions.

After two weeks, the researchers found that many of the neurons had survived and seemed to be forming connections with other brain cells. When grown in a dish, these cells showed measurable electrical activity and calcium signaling, suggesting the ability to communicate with other neurons. The researchers now hope to explore the possibility of implanting these neurons into the spinal cord.

The MIT team also hopes to increase the efficiency of this process for human cell conversion, which could allow for the generation of large quantities of neurons that could be used to treat spinal cord injuries or diseases that affect motor control, such as ALS. Clinical trials using neurons derived from iPSCs to treat ALS are now underway, but expanding the number of cells available for such treatments could make it easier to test and develop them for more widespread use in humans, Galloway says.

The research was funded by the National Institute of General Medical Sciences and the National Science Foundation Graduate Research Fellowship Program.

EPA Administrator Lee Zeldin outlined an aggressive plan to dismantle the "holy grail" of climate regulation. It's a high-risk gamble.

The proposed list of words to avoid also includes terms that only tangentially relate to global warming, such as “water conservation.”

Challengers are already planning their appeal against the project, which President Donald Trump championed in his first address to Congress.

Judge Tanya Chutkan offered significant criticism of EPA’s action but stopped short of saying she will side with the Climate United Fund.

The U.S. Export-Import Bank is expected to approve the loan for a TotalEnergies facility that's located in a conflict zone.

California Air Resources Board Chair Liane Randolph is speaking out about the threats to the state’s electric vehicle goals.

State aid rules underpinning the Clean Industrial Deal aim to boost renewable power, decarbonization and clean tech.

Wall Street’s biggest banks have quit the Net-Zero Banking Alliance in rapid succession, with Canada’s largest banks quickly following.

Scientists have long predicted that as climate change makes the air warmer, it holds more moisture, which means bigger, nastier atmospheric rivers in the future.

The son of Australia’s first billionaire, Simon Holmes a Court used his family’s wealth in 2022 to back professional women running for office with a pro-climate message.

Sports analytics is fueled by fans, and funded by teams. The 19th annual MIT Sloan Sports Analytics Conference (SSAC), held last Friday and Saturday, showed more clearly than ever how both groups can join forces.

After all, for decades, the industry’s main energy source has been fans weary of bad strategies: too much bunting in baseball, too much punting in football, and more. The most enduring analytics icon, Bill James, was a teacher and night watchman until his annual “Baseball Abstract” books began to upend a century of conventional wisdom, in the 1980s. After that, sports analytics became a profession.

Meanwhile, franchise valuations keep rising, women’s sports are booming, and U.S. college sports are professionalizing. All of it should create more analytics jobs, as “Moneyball” author Michael Lewis noted during a Friday panel.

“This whole analytics movement is a byproduct of the decisions becoming really expensive decisions,” Lewis said. “It didn’t matter if you got it wrong if you were paying someone $50,000 a year. But if you’re going to pay them $50 million, you better get it right. So, all of a sudden, someone who can give you a little bit more of an edge in that decision-making has more value.”

Would you like to be a valued sports analytics professional? Here are five ideas, gleaned from MIT’s industry-leading event, about how to gain traction in the field.

1. You can jump into this industry.

Bill James, as it happens, was the first speaker on the opening Friday-morning panel at SSAC, held at the Hynes Convention Center in Boston. His theme: the value of everyone’s work, since today’s amateurs become tomorrow’s professionals.

“Time will reveal that the people doing really important work here are not the people sitting on the stages, but the people in the audience,” James said.

This year, that audience had 2,500 attendees, from 44 U.S. states, 42 countries, and over 220 academic institutions, along with dozens of panels, a research paper competition, and thousands of hallway conversations among networking attendees. SSAC was co-founded in 2007 by Daryl Morey SM ’00, president of basketball operations for the Philadelphia 76ers, and Jessica Gelman, CEO of KAGR, the Kraft Analytics Group. The first three conferences were held in MIT classrooms.

But even now, sports analytics remains largely a grassroots thing. Why? Because fans can observe sports intensively, without being bound to its conventions, then study it quantitatively.

“The driving thing for a lot of people is they want to take this [analytical] way of thinking and apply it to sports,” soccer journalist Ryan O’Hanlon of ESPN said to MIT News, in one of those hallway conversations.

O’Hanlon’s 2022 book, “Net Gains,” chronicles the work of several people who held non-sports jobs, made useful advances in soccer analytics, then jumped into the industry. Soon, the sport may have more landing spots, between the growth of Major League Soccer in the U.S. and women’s soccer everywhere. Also, in O’Hanlon’s estimation, only three of the 20 clubs in England’s Premier League are deeply invested in analytics: Brentford, Brighton, and (league-leading) Liverpool. That could change.

In any case, most of the people who leap from fandom to professional status are willing to examine issues that others take for granted.

“I think it’s not being afraid to question the way everyone is doing things,” O’Hanlon added. “Whether that’s how a game is played, how we acquire players, how we think about anything. Pretty much anyone who gets to a high level and has impact [in analytics] has asked those questions and found a way to answer some.”

2. Make friends with the video team.

Suppose you love a sport, start analyzing it, produce good work that gets some attention, and — jackpot! — get hired by a pro team to do analytics.

Well, as former NBA player Shane Battier pointed out during a basketball panel at SSAC, you still won’t spend any time talking to players about your beloved data. That just isn’t how professional teams work, not even stat-savvy ones.

But there is good news: Analysts can still reach coaches and athletes through skilled use of video clips. Most European soccer managers ignore data, but will pay attention to the team’s video analysts. Basketball coaches love video. In American football, film study is essential. And technology has made it easier than ever to link data to video clips.

So analysts should become buddies with the video group. Importantly, analytics professionals now grasp this better than ever, something evident at SSAC across sports.

“Video in football [soccer] is the best way to communicate and get on the same page,” said Sarah Rudd, co-founder and CTO of src | ftbl, and a former analyst for Arsenal, at Friday’s panel on soccer analytics.

3. Seek opportunities in women’s sports analytics.

Have we mentioned that women’s sports is booming? The WNBA is expanding, the size of the U.S. transfer market in women’s soccer has doubled for three straight years, and you can now find women’s college volleyball in a basic cable package.

That growth is starting to fund greater data collection, in the WNBA and elsewhere, a frequent conversation topic at SSAC.

As Jennifer Rizzotti, president of the WNBA’s Connecticut Sun, noted of her own playing days in the 1990s: “We didn’t have statistics, we didn’t have [opponents’] tendencies that were being explained to us. So, when I think of what players have access to now and how far we’ve come, it’s really impressive.” And yet, she added, the amount of data in men’s basketball remains well ahead of the women’s game: “It gives you an awareness of how far we have to go.”

Some women’s sports still lack the cash needed for basic analytics infrastructure. One Friday panelist, LPGA golfer Stacy Lewis, a 13-time winner on tour, noted that the popular ball-tracking analytics system used in men’s golf costs $1 million per week, beyond budget for the women’s game.

And at a Saturday panel, Gelman said that full data parity between men’s and women’s sports was not imminent. “Sadly, I think we’re years away because we just need more investment into it,” she said.

But there is movement. At one Saturday talk, data developer Charlotte Eisenberg detailed how the website Sports Reference — a key resource of free public data —has been adding play-by-play data for WNBA games. That can help for evaluating individual players, particularly over long time periods, and has long been available for NBA games.

In short, as women’s sports grow, their analytics opportunities will, too.

4. Don’t be daunted by someone’s blurry “eye test.”

A subtle trip-wire in sports analytics, even at SSAC, is the idea that analytics should match the so-called “eye test,” or seemingly intuitive sports observations.

Here’s the problem: There is no one “eye test” in any sport, because people’s intuitions differ. For some basketball coaches, an unselfish role player stands out. To others, a flashy off-the-dribble shooter passes the eye test, even without a high shooting percentage. That tension would exist even if statistics did not.

Enter analytics, which confirms the high value of efficient shooting (as well as old-school virtues like defense, rebounding, and avoiding turnovers). But in a twist, the definition of a good shot in basketball has famously changed. In 1979-80, the NBA introduced the three-point line; in 1985, teams were taking 3.1 three-pointers per game; now in 2024-25, teams are averaging 37.5 three-pointers per game, with great efficiency. What happened?

“People didn’t use [the three-point shot] well at the beginning,” Morey said on a Saturday panel, quipping that “they were too dumb to know that three is greater than two.”

Granted, players weren’t used to shooting threes in 1980. But it also took a long time to change intuitions in the sport. Today, analytics shows that a contested three-pointer is a higher-value shot that an open 18-foot two-pointer. That might still run counter to someone’s “eye test.”

Incidentally, always following analytically informed coaching might also lead to a more standardized, less interesting game, as Morey and basketball legend Sue Bird suggested at the same panel.

“There’s a little bit of instinct that is now removed from the game,” Bird said. Shooting threes makes sense, she concurred, but “You’re only focused on the three-point line, and it takes away all the other things.”

5. Think about absolute truths, but solve for current tactics. 

Bill James set the bar high for sports analytics: His breakthrough equation, “runs created,” described how baseball works with almost Newtonian simplicity. Team runs are the product of on-base percentage and slugging percentage, divided by plate appearances. This applies to individual players, too.

But it’s almost impossible to replicate that kind of fundamental formula in other sports.

“I think in soccer there’s still a ton to learn about how the game works,” O’Hanlon told MIT News. Should a team patiently build possession, play long balls, or press up high? And how do we value players with wildly varying roles?

That sometimes leads to situations where, O’Hanlon notes, “No one really knows the right questions that the data should be asking, because no one really knows the right way to play soccer.”

Happily, the search for underlying truths can also produce some tactical insights. Consider one of the three finalists in the conference’s research paper competition, “A Machine Learning Approach to Player Value and Decision Making in Professional Ultimate Frisbee,” by Braden Eberhard, Jacob Miller, and Nathan Sandholtz.

In it, the authors examine playing patterns in ultimate, seeing if teams score more by using a longer string of higher-percentage short-range passes, or by trying longer, high-risk throws. They found that players tend to try higher-percentage passes, although there is some variation, including among star players. That suggests tactical flexibility matters. If the defense is trying to take away short passes, throw long sometimes.

It is a classic sports issue: The right way to play often depends on how your opponent is playing. In the search for ultimate truths, analysts can reveal the usefulness of short-term tactics. That helps team win, which helps analytics types stay employed. But none of this would come to light if analysts weren’t digging into the sports they love, searching for answers and trying to let the world know what they find.

“There is nothing happening here that will change your life if you don’t follow through on it,” James said. “But there are many things happening here that will change your life if you do.”