Educators & Knowledge Creators

Five major commercial publishers -- Elsevier, Springer Nature, Wiley, Taylor & Francis, and SAGE -- dominate scientific publishing, controlling 56% of articles published and capturing 75% of European spending on scientific journals. Elsevier's parent company Relx reported £3.2 billion in adjusted operating profit for 2024 with a margin of 38.4% -- exceeding Google's 34.3%. Article processing charges for open access publications nearly tripled between 2019 and 2023, with APC spending rising from $910.3 million to $2.5 billion. Top-tier journals like Nature charge upward of $10,000 per article, while Elsevier's APCs range from $200 to $11,400, creating a system where publicly funded researchers must pay to publish their own taxpayer-funded work.

Nature's landmark survey of 1,576 researchers found that more than 70% have tried and failed to reproduce another scientist's experiments, and more than half have failed to reproduce their own. Failure rates vary by discipline: 87% of chemists, 77% of biologists, 69% of physicists and engineers, and 67% of medical researchers reported inability to replicate published results. A follow-up 2024 survey of 1,630 biomedical researchers by Fierce Biotech found that 72% agreed their field faces a reproducibility crisis, with 62% blaming publish-or-perish culture and only 16% saying their institution had procedures to improve reproducibility. The crisis undermines the foundational integrity of the scientific record and wastes billions in research funding pursuing unreliable findings.

NIH awards dropped 29% and NSF awards fell 50% in 2025 compared to recent years. Grant success rates for established investigators plunged from about 27% to 20%, while at-risk investigators saw rates fall from 24% to 17%. At the National Cancer Institute, success rates collapsed from one in 10 to one in 25 applicants. Researchers now spend an average of 40% of their time on grant-related activities rather than actual research, with 60% of principal investigators submitting more grants to compensate for lower success rates. One researcher estimated losing $1.1 million of his $3.3 million in NIH funding, forcing him to lay off up to half of his 21-member team.

Science magazine reported that the NIH raised the minimum postdoc salary to $61,008 in 2024 -- an 8% increase that fell far short of the $70,000 recommended by an NIH advisory group. Acting NIH director Lawrence Tabak acknowledged 'the current system is no longer sustainable,' but called improved pay 'a zero-sum game' within flat budgets. Faculty expressed concern that higher postdoc salaries would force smaller labs, with some warning they would simply hire fewer postdocs. Meanwhile, 87% of survey respondents expressed concerns about postdoc salaries, and postdocs routinely work 50-60 hour weeks (some reporting up to 80 hours), resulting in effective hourly wages that can fall below minimum wage. MIT and Princeton allocated temporary supplementary funding to cover the raises, while many other institutions provided no additional support -- illustrating the structural impossibility of paying researchers fairly within a grant-funded system where budgets have been flat for over a decade.

More than 14,000 retraction notices were issued in 2023, setting a record, with the annual number rising from approximately 1,000 to over 10,000 between 2015 and 2025 -- a 900% increase in less than a decade. Conservative estimates suggest up to 400,000 fraudulent articles have infiltrated scientific literature over 20 years. The phenomenon of "paper mills" -- commercial operations mass-producing fake studies -- has transformed the nature of misconduct, with output estimated to double every 1.5 years. Two-thirds of retractions involve misconduct, and the countries with the highest retraction counts between 2015-2025 are China (22,574; 45.8%), India (3,309), and the United States (2,655).

A survey of over 700 international postdoctoral researchers at Harvard Medical School revealed that more than 40% spent over a month in their home countries waiting to renew US visas, with 5% spending more than six months -- effectively unable to work. Scholars from Asia experienced longer delays and higher costs than Europeans. In September 2025, the H-1B visa fee was raised to $100,000, while new requirements included mandatory disclosure of social media accounts and prolonged vetting. A Nature analysis found US-based scientists submitted 32% more applications to positions abroad in early 2025 compared to 2024, while applications to US jobs from European and Chinese scientists plummeted, signaling a brain drain from the world's largest research enterprise.

The publish-or-perish culture is driving a mental health crisis across academia. Early-career researchers, particularly PhD students, experience markedly higher levels of anxiety, depression, and emotional stress, with 20-50% of graduate students reporting symptoms of depression or anxiety -- rates up to six times greater than the general population. A study of biomedical doctoral students found anxiety disorders in 31.9%, mood disorders in 14.5%, and personality disorders in 11.6%. The pressure contributes to talent attrition, with promising researchers leaving the field citing burnout, toxic environments, or ethical concerns. Scholars from developing countries and non-English-speaking backgrounds face compounded challenges meeting global standards set by Western institutions without equivalent resources.

The NIH awarded transition grants to 172 fewer postdoctoral researchers in the first nine months of 2025 compared to the previous year -- a 10% reduction. Overall, 896 fewer new early-career grants were funded for undergraduates, PhD students, and postdocs, the lowest number since 2016. NIH cut approximately 2,100 grants worth around $9.5 billion, and indirect cost rates were slashed to 15% per grant, devastating university laboratory infrastructure. A breast cancer researcher at Harvard lost seven of her 18 lab employees after funding was frozen, while approximately 10,000 HHS termination notices were issued in March 2025, including 1,000 at the NIH -- 20% of which were later identified as errors.

In a steady state, only 12.8% of PhD graduates can attain permanent academic positions in the US. In engineering, a professor graduates an average of 7.8 new PhDs during their career, but only one can replace that professor's position. Between 2005 and 2009, 100,000 doctorates were awarded against only 16,000 open professorships. Less than 17% of new science, engineering, and health PhDs find tenure-track positions within three years of graduation. With university expansion halted, the system is structurally saturated far beyond capacity to absorb new PhDs, creating a permanent underclass of highly trained researchers trapped in temporary positions with no path to stability.

In 2025, 68% of all US faculty hold contingent appointments, with 49% working part-time -- up from 47% contingent in 1987. Adjunct faculty earn between $2,500 and $5,000 per course (some as low as $1,500), while tenured professors earn $85,000-$150,000+ annually. The vast majority of part-time faculty qualify for SNAP benefits (food stamps). In Florida, all eight adjunct faculty unions representing more than 8,000 professors were dissolved in 2024 under state law requiring 60% dues-paying membership. Many "part-time" faculty actually teach full-time course loads without benefits, healthcare, or job security, while the gig academy model shifts economic risk entirely onto individual workers.

Despite decades of open access advocacy, 64% of scholarly publications remain behind paywalls, according to the European Commission. UK universities spent an average of £4 million on journal subscriptions, while SUNY faced an annual $9 million bill for approximately 2,200 Elsevier titles. The Gates Foundation announced that starting January 2025 it would no longer cover publishing costs, creating anxiety about how the open-access model can survive without funder support. Researchers at under-resourced institutions -- particularly in the Global South -- are systematically excluded from both reading and publishing in top journals, creating a two-tier system where access to scientific knowledge depends on institutional wealth rather than scientific merit.

State-of-the-art research equipment requires staggering capital investment. MRI machines cost between $150,000 and $3 million depending on configuration, with advanced 7-Tesla systems used in neuroscience research reaching $3-4 million. Installation adds up to $1.6 million, and annual maintenance contracts run $70,000-$150,000 (8-12% of purchase price). The NSF Major Research Instrumentation program caps proposals at $4 million, often insufficient for cutting-edge equipment. With NIH indirect cost rates slashed to 15% in 2025, universities' ability to maintain existing laboratory infrastructure has been severely compromised, forcing researchers to compete for shrinking equipment budgets or abandon entire lines of inquiry that require expensive instrumentation.

A GigaScience study analyzed 27,271 Jupyter notebooks from 2,660 GitHub repositories associated with 3,467 biomedical publications. Of 10,388 notebooks with successfully installed dependencies, only 1,203 ran without errors -- and just 879 (approximately 5.9%) produced results identical to the originals. The study found that journals including Nature and Nucleic Acids Research had exception rates well above 50%. Dependency management failures, outdated Python versions, and undeclared dependencies represent systemic barriers to computational reproducibility, threatening the integrity of data-driven scientific research.

A Proof News investigation revealed that Nvidia scraped YouTube videos to train its Cosmos AI model, with VP of Research Ming-Yu Liu writing in an internal email about building "a video data factory that can yield a human lifetime visual experience worth of training data per day." Nvidia used dozens of virtual machines with rotating IP addresses to evade YouTube's detection systems. Companies including Meta, Microsoft, and Nvidia extracted over 15.8 million videos from more than 2 million YouTube channels without creator consent, prompting class-action lawsuits alleging unjust enrichment and unfair competition.

OpenAI CEO Sam Altman confirmed that GPT-4's training cost exceeded $100 million, while Google's Gemini Ultra reached an estimated $191 million -- representing a 287,000x increase from the cost of training a Transformer model in 2017 ($670). These costs concentrate frontier AI development in a handful of well-resourced corporations. Academic researchers, as noted by cognitive scientist Sean Trott, find it "hard to run (and even harder to train) state-of-the-art LLMs on an academic budget, which limits the kinds of questions we can ask," creating a structural divide where publicly funded researchers cannot compete with or verify the claims of private AI labs.

A comprehensive survey found that 90-95% of researchers in the US and UK rely on research software, and more than 63% reported they could not continue their work if such software stopped functioning. Yet Research Software Engineers (RSEs) who build and maintain this critical infrastructure remain largely invisible in academia -- unable to earn authorship credit on papers, excluded from traditional promotion criteria, and lacking formal career paths. The US-RSE community has grown to 2,800 members advocating for recognition, but as a Princeton University study documented, "lack of a clear career path" remains RSEs' top concern, with 194 different job titles fragmenting the profession.

A 365 Data Science study of 1,001 data scientist LinkedIn profiles across the US (35%), UK (25%), EU (25%), and India (15%) found that data scientists stay with their current employer for an average of just 1.7 years. A related survey of 600 data engineers found that 97% experienced burnout in their day-to-day work, with 79% considering leaving the industry entirely. The primary causes stem from organizational misunderstanding of the role, unrealistic analytics expectations, and the "data janitor" problem -- where professionals hired for advanced analysis spend up to 80% of their time on data cleaning and pipeline maintenance.

Despite requiring advanced degrees (often PhDs) and expertise spanning biology, statistics, and computer science, bioinformatics professionals face severe pay disparity. The bottom 10% of biological scientists earn just $54,500 annually, while average bioinformatics scientist salaries range from $85,012 to $116,054 depending on the source -- substantially below the $141,000-$250,000 range for ML engineers with comparable technical skills but no domain science requirement. Many bioinformaticians "know they're underpaid but don't know where to start" negotiating, and geographic concentration means those outside San Francisco, Boston, and San Diego earn 20-30% less than national averages.

Princeton University's RSE Group documented the systemic absence of career structures for research software engineers in academia. When polled, RSE Group members identified "lack of a clear career path" as their top concern. Rapid expansion of RSE programs combined with retention challenges and limited promotion paths were "amplified by growing demand for RSEs in the private sector, which added risk of turnover." Traditional academic evaluation metrics -- publications, citations, grants -- fail to capture RSE contributions, and universities struggle to align RSE work with existing HR frameworks, forcing these professionals into ill-fitting job classifications that undervalue their expertise.

The ninth annual State of Open Data report from Digital Science, Figshare, and Springer Nature found that while open data is "on the edge of becoming a recognized global standard," critical equity gaps persist. Average repository sharing rates hover around 25% in wealthy nations (US, UK, Germany, France) but remain "significantly below a quarter" in Brazil, Ethiopia, and India. Researchers face unfunded mandates to share data openly while corporations commercialize those same datasets. The report warns of "a potential divide where open science becomes the preserve of better-resourced research environments, potentially marginalizing researchers in low- and middle-income countries."

MIT Technology Review documented how AI companies have systematically scraped training data without consent or compensation, prompting a wave of lawsuits and licensing deals. The New York Times alleges millions of copyrighted articles were used to train AI models without consent. LinkedIn faces a class-action lawsuit for allegedly harvesting private messages for AI training. Reddit sued Perplexity AI for obtaining data "using false identities, proxies and other antisecurity techniques." Meanwhile, data scientists and researchers whose public datasets, code, and analyses were scraped for training have no mechanism for attribution, opt-out, or compensation -- their computational work treated as raw material for corporate AI products.

A comprehensive review in AI Magazine identified unique obstacles to reproducibility in machine learning research: sensitivity to training conditions, sources of randomness, inherent nondeterminism, and prohibitive computational costs. The reproducibility crisis is quantifiably severe, with one estimate citing "an annual $200 billion global drain on scientific computing resources" from irreproducible computational work. Unlike traditional science where methods can be independently replicated, ML experiments often cannot be verified because the compute costs are too high, the training data is proprietary, or the random seeds and hyperparameters are undocumented -- eroding the scientific foundation of data-driven research.

Despite headlines about AI growth, the data science job market has bifurcated dramatically. Tech companies laid off approximately 237,000 workers across 1,107 companies in 2024, with data and ML roles hit alongside broader engineering cuts. Companies are "no longer hiring ML engineers to 'figure out our AI strategy'" -- they want specialists who can ship production systems, creating a two-tier market where experienced practitioners command $185,000-$285,000 while entry-level data scientists face a saturated, contracting market. The paradox: organizations simultaneously claim they cannot find enough AI talent while eliminating data science positions in favor of automated ML pipelines and AI-as-a-service solutions.

A white paper submitted to the Heliophysics Decadal Survey documented that Research Software Engineers "receive unequal treatment compared to their science counterparts, including lack of credit for their contributions and insufficient training." Despite being essential to 63% of US researchers who cannot continue their work without software, RSEs are systematically excluded from authorship, grant PI eligibility, and academic recognition systems. The paper advocates for RSEs to receive "equality of contribution" -- equivalent credit, career advancement opportunities, and institutional support as domain scientists -- arguing that the current system exploits technical labor while reserving prestige and funding for traditional research roles.