AI & Automation in Biology

iNaturalist Computer Vision

• iNaturalist (a joint initiative of the California Academy of Sciences and National Geographic) uses deep convolutional neural networks (CNNs) trained on its community-contributed photo database. As of 2024, the model can suggest identifications for over 76,000 species across plants, animals, and fungi. The platform has surpassed 175 million observations globally.
• The computer vision model is retrained regularly as new verified observations accumulate, and it draws on the large-scale visual recognition challenge infrastructure. iNaturalist's API allows third-party apps to use the CV model.
• Key detail: The model achieves roughly 90%+ accuracy at the species level for well-photographed taxa (birds, butterflies, common plants), but accuracy drops significantly for underrepresented species, demonstrating that AI is a tool that still depends on human expert verification.

Merlin Bird ID (Cornell Lab of Ornithology)

• Merlin uses AI for both photo-based and sound-based bird identification. Its Sound ID feature, launched in 2021 and continually expanded, uses a deep neural network trained on recordings from the Macaulay Library (the world's largest archive of animal sounds, with 1.5+ million recordings). As of 2024–2025, Merlin can identify 1,300+ bird species by sound across multiple regions worldwide.
• The sound identification model processes spectrograms (visual representations of sound) using image classification techniques. It can identify multiple species simultaneously in a single recording, even in noisy environments.
• Research group: Cornell Lab of Ornithology's Center for Conservation Bioacoustics, led by researchers including Holger Klinck.

Automated Camera Traps and Wildlife Cameras

• Wildlife Insights (a Google-backed platform in partnership with WWF, the Smithsonian, and Wildlife Conservation Society): Uses AI to automatically classify species in millions of camera trap images. Their model can identify hundreds of species and filter out blank images (a huge time-saver, as 70–80% of camera trap photos are typically empty).
• LILA BC (Labeled Information Library of Alexandria: Biology and Conservation): A massive open repository of camera trap datasets used for training AI models, managed by Microsoft AI for Earth.
• MegaDetector (Microsoft AI for Earth): An open-source AI model specifically designed to detect animals, people, and vehicles in camera trap images. It doesn't identify species but sorts images into “animal/no animal,” dramatically reducing the manual review burden. Widely used by conservation groups worldwide.
• Zooniverse platforms like Snapshot Serengeti pioneered citizen science classification of camera trap images, and their datasets now train AI models.
• Specific companies: Conservation X Labs, Wildbook (using AI for individual animal identification from photos using spot/stripe patterns — used for whale sharks, giraffes, zebras), and SMART (Spatial Monitoring and Reporting Tool) for protected area management.

eDNA refers to genetic material shed by organisms into their environment (water, soil, air). AI and machine learning are now essential in analyzing the massive sequencing datasets that eDNA produces.

Key Tools and Platforms

• QIIME 2 (Quantitative Insights Into Microbial Ecology): Open-source bioinformatics platform that uses machine learning classifiers (including naive Bayes and scikit-learn-based models) for taxonomic assignment of DNA sequences.
• DADA2: Algorithm for high-resolution sample inference from amplicon sequencing data — uses statistical modeling to distinguish real biological sequences from sequencing errors.
• eDNAFlow and Anacapa Toolkit: Bioinformatics pipelines specifically designed for eDNA metabarcoding analysis.
• Jonah Ventures (Boulder, CO): Commercial eDNA analysis company using advanced bioinformatics pipelines for biodiversity assessment.
• NatureMetrics (UK-based): Commercial eDNA company providing biodiversity monitoring services using AI-powered bioinformatics to process water and soil samples.
• EnviroDNA (Australia): Similar commercial eDNA services.

Research Groups

• University of Washington's eDNA Collaborative, led by Ryan Kelly
• The Spygen lab in France (pioneers in eDNA methodology)
• NOAA's Atlantic Oceanographic and Meteorological Laboratory

AI Application

Machine learning models are increasingly used to predict species presence/absence from eDNA data, optimize sampling designs, and integrate eDNA data with other environmental variables for habitat modeling.

Satellite Platforms

• Planet Labs: Operates the largest commercial fleet of Earth-observing satellites (200+ satellites), providing daily imagery of the entire Earth's landmass at 3–5 meter resolution. AI/ML algorithms are applied to detect deforestation, coral bleaching, agricultural land conversion, and habitat fragmentation.
• Global Fishing Watch: Uses AI to analyze satellite AIS (Automatic Identification System) data and synthetic aperture radar (SAR) to detect and track fishing vessels globally, including identifying illegal fishing. A joint initiative of Google, Oceana, and SkyTruth. Published landmark research in Nature (2024) showing that 75% of industrial fishing vessels are not publicly tracked.
• Global Forest Watch (World Resources Institute): Uses Landsat and Sentinel satellite imagery processed with machine learning to track deforestation in near-real-time. Their GLAD (Global Land Analysis and Discovery) alerts system detects forest loss weekly.
• Allen Coral Atlas: Uses satellite imagery (from Planet Labs) combined with AI to map and monitor the world's shallow coral reefs. A collaboration between the Vulcan Foundation (Paul Allen), Arizona State University, and the University of Queensland.

Drone / UAV Applications

• SnotBot (Ocean Alliance): Drones that fly through whale blow (exhaled breath) to collect biological samples for health assessment.
• Wildlife Drones (Australia): Company developing radio-tracking drones that can locate tagged wildlife.
• AI-powered drone surveys for counting penguin colonies, seabird nests, and marine mammal haul-outs — replacing manual counts from boats or ground surveys.
• Duke University Marine Robotics and Remote Sensing Lab: Pioneering AI-analyzed drone surveys of marine megafauna.
• Conservation Drones (founded by Lian Pin Koh and Serge Wich): Non-profit providing low-cost drone solutions for tropical conservation.

• Saildrone: Autonomous surface vehicles (wind and solar powered) that collect ocean data including bathymetry, atmospheric conditions, and fisheries acoustics. NOAA has deployed Saildrone vehicles for Arctic surveys, fisheries stock assessments, and hurricane reconnaissance. In 2024–2025, Saildrone expanded to over 100 vehicles and conducted major surveys in the Bering Sea and Gulf of Alaska.
• MBARI (Monterey Bay Aquarium Research Institute): Operates multiple AUVs and ROVs, and is a leader in applying AI to deep-sea video annotation. Their FathomNet project is building an open-source image database for training AI to identify deep-sea organisms from underwater video — analogous to ImageNet for the ocean.
• Seabed 2030 (Nippon Foundation-GEBCO): Global initiative to map the entire ocean floor by 2030, using AI to process multibeam sonar data and interpolate between survey tracks.
• Ocean Infinity: Commercial company operating the world's largest fleet of AUVs for seafloor mapping and survey.
• Schmidt Ocean Institute: Operates the research vessel Falkor (and its successor Falkor (too)), deploying ROVs with AI-enabled species identification capabilities.
• Sofar Ocean (and its Bristlemouth initiative): Developing open-source, low-cost ocean sensors and the data infrastructure to process their output using AI.
• OceanX: Media and science exploration organization using advanced submersibles and AI for deep-ocean research.

AlphaFold (Google DeepMind)

• AlphaFold2, published in Nature in 2021, solved the protein structure prediction problem that had stumped biology for 50 years. It can predict 3D protein structures from amino acid sequences with accuracy rivaling experimental methods (X-ray crystallography, cryo-EM).
• AlphaFold Protein Structure Database: In partnership with EMBL-EBI, DeepMind released predicted structures for over 200 million proteins — essentially every known protein sequence. This is freely available and has been accessed by over 2 million researchers.
• AlphaFold3 (released 2024): Extends prediction to protein-DNA, protein-RNA, and protein-ligand interactions, dramatically expanding its utility for drug discovery and understanding molecular biology.
• Impact on marine/wildlife biology: Understanding venom proteins from cone snails and other marine organisms; predicting protein functions in non-model organisms; understanding evolutionary relationships through structural genomics.

Other Genomics AI Tools

• ESM (Evolutionary Scale Modeling) by Meta AI: Large language models for proteins that can predict structure and function. ESMFold is faster (though slightly less accurate) than AlphaFold.
• RoseTTAFold (Baker Lab, University of Washington): Alternative protein structure prediction tool, also used for protein design.
• CRISPR gene editing combined with AI: Companies like Inscripta and Synthego use AI to design optimal guide RNAs for CRISPR experiments.
• Oxford Nanopore Technologies: Real-time DNA/RNA sequencing technology increasingly used in field settings for rapid species identification and pathogen detection. AI basecalling algorithms convert raw electrical signals to DNA sequences.
• 10x Genomics and spatial transcriptomics: AI is essential for analyzing single-cell and spatial gene expression data.

• NOAA Climate and Ecosystem Forecasting: NOAA's Integrated Ecosystem Assessment (IEA) program uses ensemble models (including ML components) to forecast how marine ecosystems will respond to climate change.
• NEON (National Ecological Observatory Network): NSF-funded continental-scale observatory collecting standardized ecological data. Machine learning is used to process its massive data streams (phenocam images, flux tower data, biodiversity surveys).
• NASA's Earth System Digital Twins: AI-powered digital replicas of Earth systems for climate and ecological forecasting.
• GenCast (Google DeepMind, 2024): AI weather forecasting model that outperforms traditional numerical weather prediction for medium-range forecasts. Implications for ecological forecasting.
• ClimateBench and WeatherBench: ML benchmark datasets for climate and weather prediction research.
• Species Distribution Modeling (SDM): Tools like MaxEnt are being augmented or replaced by deep learning approaches. Research groups at Yale (Walter Jetz's Map of Life), University of Wisconsin, and others are pioneering this work.
• Map of Life (Yale / Walter Jetz): Global biodiversity platform integrating species distribution models with remote sensing and AI for conservation planning.

• NOAA's Passive Acoustic Monitoring (PAM): Extensive networks of hydrophones monitored using AI algorithms to detect and classify whale calls, dolphin clicks, and other marine sounds. The NOAA Northeast Fisheries Science Center uses machine learning for North Atlantic right whale detection.
• Google/NOAA Humpback Whale AI: Partnership applying Google's AI to identify humpback whale songs in vast underwater audio datasets. Published pattern detection across ocean basins.
• Rainforest Connection (RFCx): Though originally for rainforests, their Arbimon platform uses AI to analyze audio recordings for species identification — applicable to marine environments. Uses recycled phones as acoustic sensors.
• Reef health monitoring: Researchers at the University of Exeter (Tim Lamont, Steve Simpson) have used AI to analyze coral reef soundscapes — healthy reefs have distinct acoustic signatures compared to degraded ones. Published in Nature Communications and other journals.
• WHOI (Woods Hole Oceanographic Institution): Runs the Digital Signal Processing Lab and applies AI to marine acoustics data from deep-ocean monitoring systems.
• Orcasound: Open-source project using AI to detect orca calls in real-time from hydrophone networks in the Pacific Northwest. Community-driven with contributions from citizen scientists and AI developers.

• MARK and RMark: Long-standing statistical software for capture-recapture population modeling, now being augmented with Bayesian methods and ML approaches.
• movebank.org: Global repository of animal movement data. AI is used to classify animal behaviors from GPS tracking data (feeding, migrating, resting) and to predict movement corridors.
• Wildbook (Wild Me, now part of Conservation X Labs): Uses AI for photo-identification of individual animals — whale sharks (spot patterns), manta rays, humpback whales (fluke patterns), giraffes, zebras. Combines crowd-sourced photos with computer vision.
• SMART (Spatial Monitoring and Reporting Tool): Used in over 1,000 protected areas across 70+ countries for adaptive management, increasingly integrating AI-analyzed data streams.
• Integrated Population Models (IPMs): Combining multiple data sources (surveys, mark-recapture, demographic data) with Bayesian statistical frameworks. Researchers at USGS, state wildlife agencies, and universities use these for species like sage-grouse, waterfowl, and marine mammals.
• AI for invasive species: Models predicting spread patterns, optimizing removal efforts, and early detection. USDA and USGS maintain AI-assisted monitoring programs.

The ocean is estimated to contain millions of undiscovered compounds with pharmaceutical potential. AI accelerates the identification and optimization of drug candidates.

• Scripps Institution of Oceanography (UC San Diego): Long history in marine natural products chemistry, now integrating AI for compound discovery.
• Specific examples:
  • Ziconotide (Prialt), derived from cone snail venom, was discovered through traditional methods but AI is now used to explore the vast chemical space of conotoxins.
  • Halichondrin B (from sea sponges) led to eribulin (Halaven), a cancer drug. AI is being used to find and optimize similar compounds.
• GNPS (Global Natural Products Social Molecular Networking): Uses machine learning to identify molecular families in mass spectrometry data from marine organisms.
• Companies: Biosortia Pharmaceuticals (marine microbiome drug discovery), Nautilus Biosciences (marine-derived antibiotics).

• Opentrons: Open-source liquid-handling robots used in biology labs for automated DNA extraction, PCR setup, eDNA library preparation. Cost-effective enough for academic labs ($10K–$25K range vs. $100K+ for traditional lab robots).
• Hamilton and Beckman Coulter: Industrial-scale lab automation platforms used in large genomics centers and pharmaceutical companies.
• Andrew Alliance (now part of Waters Corporation): Pipetting robots for reproducible lab work.
• AI-driven experimental design: Tools like Benchling (lab management software with AI features) and platforms from Emerald Cloud Lab that allow researchers to run experiments remotely on robotic platforms.
• Self-driving labs concept: The idea of fully autonomous laboratories where AI designs experiments, robots execute them, and AI analyzes results in a closed loop. Still mostly in chemistry/materials science, but the concept is entering biology.

High Automation Potential (likely within 5–10 years)

• Routine species identification from photos and audio: AI will increasingly handle first-pass identification, with human experts needed mainly for difficult or novel cases. Camera trap image sorting is already largely automated.
• Lab sample processing: DNA extraction, PCR, sequencing library preparation, and other bench protocols are rapidly being automated with liquid-handling robots and AI-controlled workflows.
• Data entry and management: Transcribing field notes, entering survey data, managing specimen databases.
• Basic statistical analyses and report generation: Routine analyses (t-tests, ANOVA, standard regression, basic species distribution models) can be largely automated. AI can generate standard reports from monitoring data.
• Literature review and synthesis: AI tools (Semantic Scholar, Elicit, Consensus) are already dramatically accelerating the process of finding and summarizing relevant research papers.
• Satellite/drone image classification: Mapping land cover types, counting animals in aerial photos, detecting habitat change from satellite imagery — these are ripe for near-full automation.
• Water quality monitoring and environmental compliance reporting: Routine collection and analysis of standard parameters.

Moderate Automation Potential (10–15 years)

• Acoustic monitoring analysis: AI will handle the bulk of species identification from audio, but interpreting unusual patterns and novel vocalizations will still need experts.
• Population modeling: Standard models will be increasingly automated, but novel modeling approaches and interpreting results for management decisions will remain human.
• Genetic analysis interpretation: Bioinformatics pipelines will become more automated, but biological interpretation will still require human expertise.
• Survey design optimization: AI will suggest optimal sampling strategies, but adapting to local conditions will need human judgment.

Roles That May Be Reduced or Transformed

• Field technicians doing pure data collection (e.g., point counts, transect surveys) may see reduced demand as automated sensors and drones take over some monitoring tasks. However, they won't disappear — sensors break, need deployment, and can't handle all terrain or conditions.
• Lab technicians performing routine procedures will likely see significant displacement. Genomics and molecular biology labs are already heavily automated at large centers.
• Taxonomists doing routine identification may find less demand for bread-and-butter work, but deep taxonomic expertise for undescribed species, difficult groups, and training AI models will remain valuable.
• GIS analysts doing standard mapping and spatial analysis may see their routine work automated, but complex spatial analysis and novel applications will still need humans.
• Data managers handling standard database operations may see roles shrink as AI-powered data pipelines automate ingestion and quality control.

What Research Literature Says About AI Displacement in STEM

• A widely cited 2023 study by Eloundou et al. (OpenAI / University of Pennsylvania) examined exposure of occupations to large language models. They found that higher-wage, higher-education occupations (including STEM) actually have more tasks exposed to AI assistance — but “exposed” means augmented, not necessarily replaced.
• The World Economic Forum's 2023 Future of Jobs Report projected that AI and automation would create a net positive for environmental-related jobs, including conservation scientists, environmental engineers, and related roles, due to growing demand for climate and biodiversity work.
• A 2024 analysis in Nature Ecology & Evolution (Tuia et al.) specifically examined AI in ecology and conservation, concluding that AI will transform workflows but the field's inherent complexity, variability, and need for place-based knowledge make full automation unlikely.
• The U.S. Bureau of Labor Statistics projects that zoologist and wildlife biologist employment will grow 5% from 2022–2032 (about as fast as average), while conservation scientist roles are projected to grow 6%. Notably, these projections have been revised upward as biodiversity crisis awareness increases policy demand.
• Key finding across studies: STEM fields with a strong fieldwork, place-based, and interpersonal component are among the least susceptible to outright displacement by AI, even though they may see the most augmentation.

How Fieldwork vs. Lab Work vs. Desk Work Are Affected Differently

Fieldwork (LEAST affected): The physical, unpredictable, unstructured nature of fieldwork makes it extremely difficult to automate. You cannot send a robot to do a coral reef survey in rough seas, navigate a dense tropical forest to set camera traps, or handle a live sea turtle for tagging. AI augments fieldwork (better tools, real-time ID assistance on tablets, drone support) but does not replace the human in the field. Field conditions (weather, terrain, equipment failures, unexpected events) require adaptability that current AI/robotics cannot match.

Lab work (MOST affected): Structured, repetitive, controlled environments are ideal for automation. High-throughput genomics labs are already heavily robotic. The trend is toward fewer lab technicians doing manual pipetting and more automated systems managed by fewer, more skilled operators. However, many marine/wildlife biology labs involve non-standard samples (varied species, field-collected specimens of varying quality) that are harder to automate than clinical or pharmaceutical lab work.

Desk work (HEAVILY augmented): Data analysis, writing, literature review, grant applications, and administrative tasks are heavily augmented by AI. AI can draft sections of papers, suggest statistical approaches, create visualizations, and summarize literature. But scientific judgment, novel hypothesis generation, interpreting ambiguous results, and communicating nuanced findings remain fundamentally human. Desk work will become much more productive per person (one biologist can do the analytical work that used to require a team), which may reduce total headcount for some roles but increase the scope and impact of individual researchers.

Fieldwork and Field Ecology Remain Fundamentally Human

• Physical presence in complex environments: Marine biology often requires diving, boating in variable conditions, working in remote locations, handling live organisms. Wildlife biology requires hiking rugged terrain, enduring extreme weather, navigating by instinct. These environments are unstructured and unpredictable in ways that current robotics cannot handle.
• Sensory integration: Experienced field biologists integrate smell, sound, “feel” of an ecosystem, subtle visual cues, and patterns of animal behavior in ways that no sensor array can replicate. A marine biologist can sense changes in reef health by the smell of the water, the color of the algae, the behavior of fish — a gestalt perception that comes only from extensive experience.
• Adaptive decision-making in real-time: When conditions change (weather shifts, equipment breaks, an unexpected species appears, an animal is injured), field biologists make rapid decisions drawing on deep contextual knowledge. AI cannot do this in unstructured natural environments.
• Relationship building: Fieldwork in conservation requires building trust with local communities, indigenous groups, fishers, ranchers, and park staff. This is fundamentally human.
• Ethical in-field decisions: When to handle vs. observe an animal, when to intervene in a suffering individual vs. let nature take its course, how to minimize disturbance — these require moral reasoning and compassion.

Scientific Intuition, Hypothesis Generation, and Experimental Design

• AI can find patterns in data, but asking the right questions — “Why are these corals surviving when their neighbors are bleaching?” “Could this unusual migration pattern be related to a new food source?” — requires scientific intuition born from deep domain expertise.
• Designing experiments that can actually answer causal questions (not just correlational patterns) requires understanding confounding variables, practical constraints, ethical limits, and the biology of the system.
• Many of the greatest discoveries in biology came from serendipitous observations by attentive researchers — noticing something unexpected and knowing it was important. AI processes data according to its training; it does not wander through a forest and notice something odd.
• Abductive reasoning (inference to the best explanation) in novel situations remains a distinctly human strength.

Ethical Decision-Making in Conservation

• Conservation biology is fundamentally a values-driven discipline. Decisions about which species to prioritize, how to balance human livelihoods with wildlife protection, whether to use lethal control of invasive species, and how to allocate limited conservation funds involve ethical reasoning that AI should inform but not make.
• Triage decisions: With limited resources, conservationists must decide which species or habitats to prioritize. This involves not just data but societal values, cultural significance, and ethical frameworks.
• Human-wildlife conflict: Managing coexistence between wildlife and human communities requires empathy, negotiation, cultural sensitivity, and political awareness.
• Policy implications of scientific findings: Deciding how to present uncertainty, what to recommend to policymakers, and how to balance scientific rigor with actionable advice.

Stakeholder Engagement, Policy Advocacy, and Science Communication

• Marine and wildlife biologists increasingly serve as translators between science and society. They present findings to government agencies, advocate for policy changes, write for public audiences, and educate communities.
• Testifying before legislatures, serving on advisory boards, participating in international negotiations (like CITES): These roles require persuasion, credibility, and human connection.
• Science communication: While AI can help draft content, the most effective science communication comes from passionate individuals who can tell stories, connect emotionally with audiences, and adapt their message to diverse stakeholders.
• Community-based conservation: Programs like community-based marine protected areas or co-management arrangements with indigenous communities require deep cultural competence and relationship building over years.

Cross-Disciplinary Thinking and Novel Research Questions

• The most impactful research often happens at the intersection of disciplines — marine biology + economics + social science for fisheries management, ecology + genomics + climate science for adaptation research, conservation biology + political science + indigenous knowledge for effective protection.
• AI excels within defined domains but struggles to make the creative leaps across disciplines that humans do naturally.
• Novel research directions often emerge from informal conversations, conference hallways, field observations, and the kind of serendipitous connections that happen when curious humans interact with the natural world.

Why Marine/Wildlife Biology May Be MORE Resilient to AI Than Other STEM Fields

• Unstructured environments: Unlike a chemistry lab or a software engineering environment, ecosystems are messy, variable, and unpredictable. This is the worst possible domain for pure automation.
• Growing demand: The biodiversity crisis and climate change are increasing demand for conservation expertise, not decreasing it. The Kunming-Montreal Global Biodiversity Framework (2022) commits nations to protecting 30% of Earth's land and ocean by 2030, creating enormous demand for biologists.
• Emotional and cultural significance: Society values connection with nature and wildlife in ways that create enduring demand for human experts who can serve as guides, educators, and advocates.
• Regulatory requirements: Environmental impact assessments, endangered species consultations, and wildlife management plans typically require certified biologists with specific credentials — regulatory frameworks evolve slowly and tend to require human accountability.
• The data problem: Unlike text or images on the internet, ecological data is sparse, noisy, and context-dependent. AI models for ecology face fundamental limitations in training data availability, which means human expertise remains essential for interpretation.

New Roles That Didn't Exist 5 Years Ago

• Computational Ecologist: Combines ecology expertise with programming and data science to build models, analyze large datasets, and develop AI tools for ecological applications.
• Conservation Technologist: Designs, deploys, and manages technology solutions for conservation — from camera trap networks to satellite monitoring systems to acoustic arrays.
• Conservation Data Scientist: Analyzes large-scale conservation data (satellite imagery, sensor networks, citizen science databases) to inform management decisions.
• eDNA Program Manager: Designs and oversees environmental DNA monitoring programs, combining molecular biology with ecological expertise and project management.
• Wildlife AI/ML Engineer: Develops and trains machine learning models specifically for wildlife applications (species ID, behavior classification, population estimation).
• Marine Robotics Operator/Technician: Operates and maintains AUVs, ROVs, and autonomous surface vehicles for marine research.
• Bioacoustics Data Analyst: Specializes in processing and interpreting large acoustic datasets from passive acoustic monitoring programs.
• Biodiversity Informatics Specialist: Manages large biodiversity databases (GBIF, OBIS, eBird) and develops tools for data integration and analysis.
• Conservation Finance Analyst: Uses quantitative methods to evaluate conservation investments, design biodiversity credits, and analyze ecosystem service valuations — a role that increasingly requires AI/data science skills.

The Growing Demand for Biologists Who Can Code and Use AI

• A 2023 study in BioScience found that job postings for ecological positions increasingly list programming skills (Python, R) as required or preferred qualifications — up from roughly 20% of postings in 2010 to over 50% in 2023.
• The “unicorn” biologist: The most sought-after candidates combine deep biological expertise with computational skills. These individuals command higher salaries and have their pick of positions because they can bridge the gap between domain science and technology.
• NOAA, USGS, USFWS, and state agencies are all increasingly seeking quantitative skills in their biologist hires.
• Academic job postings for ecology and marine biology faculty increasingly require computational/quantitative expertise.
• The flip side: Biology PhDs who also have strong data science skills are also attractive to industry (tech companies, consulting firms, biotech), providing career flexibility.

Bioinformatics as a Booming Subfield

• Bioinformatics sits at the intersection of biology, computer science, and statistics. The global bioinformatics market was valued at approximately $14 billion in 2023 and is projected to grow at 15–20% annually through 2030.
• While often associated with biomedical research, bioinformatics is increasingly central to conservation genomics, population genetics, metagenomics, and eDNA analysis.
• Training programs: Dedicated bioinformatics programs exist at universities including UC San Diego, Johns Hopkins, University of Michigan, Boston University, and many others. Many are available as master's programs designed for career changers.
• Career paths: Bioinformaticians work in academia, government agencies (NCBI, NOAA, USDA), pharmaceutical companies, biotech startups, and conservation organizations.

Conservation Technology as an Emerging Industry

• WILDLABS.NET: Global community of conservation technology practitioners with 6,000+ members, connecting technologists with conservation organizations.
• Conservation X Labs: Incubator and accelerator for conservation technology startups.

Companies at the intersection of AI and marine/wildlife biology:

• Saildrone (Alameda, CA): Autonomous ocean vehicles — raised over $300 million in funding.
• Planet Labs (San Francisco): Earth-observation satellites — publicly traded (PL on NYSE).
• Global Fishing Watch (Washington, DC): AI-powered illegal fishing detection.
• Whale Safe (Benioff Ocean Science Laboratory, UC Santa Barbara): AI system combining acoustic monitoring, satellite data, and whale habitat models to reduce ship-whale collisions.
• Rainforest Connection (San Francisco): Acoustic monitoring for forest protection and biodiversity.
• Conservation Metrics (Santa Cruz, CA): AI-powered analysis of ecological monitoring data.
• OrcaHello and Orcasound: Volunteer-driven projects using AI for killer whale conservation.
• Blue Robotics (Torrance, CA): Affordable underwater robotics platforms.
• Sofar Ocean (San Francisco): Ocean sensing and data platforms.
• NatureMetrics (UK): eDNA-based biodiversity monitoring.
• Upstream Tech (acquired by The Nature Conservancy in 2022): AI-powered land conservation monitoring platform (Lens product uses satellite imagery + ML for conservation easement monitoring).
• Clearbot (Hong Kong): AI-powered autonomous boats for ocean plastic cleanup.
• Coral Vita (Bahamas): Coral restoration company using technology to accelerate coral growth.
• Basecamp Research (UK): Building the world's largest biodiversity genomics network using AI for bioprospecting.

How AI Amplifies Individual Researchers

A single marine biologist equipped with AI tools can now accomplish what would have taken a team of 10 just a decade ago:

• Process thousands of camera trap images in hours instead of months.
• Analyze acoustic data from months of hydrophone recordings that would have taken years manually.
• Run species distribution models across entire continents using cloud computing.
• Access and integrate data from global databases (GBIF, OBIS, GenBank) with a few lines of code.
• Generate draft manuscripts, literature reviews, and visualizations with AI assistance.

This is fundamentally empowering: It means early-career researchers can have outsized impact, small conservation organizations can punch above their weight, and research in underfunded regions can leverage free or low-cost AI tools.

Technical Skills to Develop (in order of priority)

1. Programming in Python and R: These are the two essential languages. Python for general-purpose programming, machine learning, and working with AI tools. R for statistics, ecological modeling, and bioinformatics. Learn both.
2. Statistics and data science: Go beyond intro statistics. Take courses in Bayesian statistics, multivariate analysis, machine learning, and experimental design.
3. GIS and remote sensing: Proficiency in QGIS or ArcGIS, satellite imagery analysis, and spatial statistics.
4. Bioinformatics basics: Understanding DNA sequencing data, alignment tools, and genomic analysis pipelines.
5. Command-line/Unix skills: Essential for working with high-performance computing, bioinformatics tools, and remote servers.
6. Database management: SQL basics for working with large datasets.
7. Version control (Git/GitHub): Standard practice for reproducible science and collaborative coding.

Scientific Skills

1. Strong foundation in core biology: Ecology, evolution, genetics, physiology, marine biology. AI is a tool — you need deep domain knowledge to use it effectively.
2. Field skills: Diving (AAUS scientific diving certification), boat handling, animal handling, field identification, natural history knowledge. These are your competitive advantage over pure data scientists.
3. Experimental design and scientific writing: These remain essential and are harder to develop than technical skills.

Professional Skills

1. Science communication: Writing for public audiences, giving engaging presentations, social media literacy.
2. Grant writing: Increasingly requires articulating how computational methods enhance research.
3. Project management: Conservation projects involve multiple stakeholders, timelines, and budgets.
4. Collaboration across cultures and disciplines: Conservation is global and interdisciplinary.

Programming Languages That Matter

• Python: The dominant language for machine learning, AI, and general-purpose scientific computing. Libraries like TensorFlow, PyTorch, scikit-learn, pandas, and NumPy are essential for AI work. Also widely used in bioinformatics (Biopython) and remote sensing (rasterio, Google Earth Engine Python API).
• R: The dominant language for ecological statistics and modeling. Packages like vegan (community ecology), unmarked (occupancy modeling), lme4 (mixed effects models), and ggplot2 (visualization) are standard in ecology. R Markdown and Shiny for reproducible reports and interactive visualizations.
• Julia: Emerging language for scientific computing — faster than Python/R for numerical work but with a smaller ecosystem. Worth learning later in career but not a priority now.
• SQL: For database queries — used in virtually every data-related job.
• JavaScript: Useful for web-based visualization (D3.js) and tools like Google Earth Engine. Lower priority but valuable for conservation tech roles.
• Bash/shell scripting: Essential for bioinformatics and working on remote servers.

Recommendation: Start with Python and R. Add others as needed for specific projects.

Should You Double Major or Minor in CS/Data Science?

Many graduate programs (and the field generally) value a strong biology foundation + demonstrated computational skills over a CS degree + surface-level biology knowledge. Taking specific courses (machine learning, data structures, databases, statistics) is more strategic than completing an entire CS curriculum.

Alternative path: Some universities now offer specific programs like “Quantitative Biology,” “Computational Ecology,” or “Data Science for the Life Sciences.” Examples:

• UC Berkeley's College of Computing, Data Science, and Society has relevant interdisciplinary options.
• University of Florida's interdisciplinary ecology program integrates computational training.
• Duke University offers marine science + data science combinations.

How Top Biology Programs Are Adapting

• University of Washington (SAFS): Has integrated quantitative/computational training into its core curriculum and is a leader in eDNA and quantitative ecology.
• Scripps Institution of Oceanography (UC San Diego): Increasing emphasis on data science in marine biology training; proximity to San Diego's biotech cluster provides interdisciplinary exposure.
• Oregon State University (CEOAS): Strong in marine ecology with growing computational emphasis, including AI for fisheries.
• MIT (EAPS): Has launched programs integrating AI and climate science.
• Stanford (Hopkins Marine Station): Increasingly emphasizing computational methods in ecology and conservation.
• James Cook University (Australia): Leader in coral reef science with growing AI integration (CoralNet, reef monitoring technology).
• Many programs now require R or Python: Increasingly, graduate programs in ecology and marine biology have quantitative requirements that include programming. This was rare a decade ago and is now becoming standard.
• Data Carpentry and Software Carpentry workshops: Many biology departments now host these short-course training programs in computational skills for biologists.

What a “Future-Proof” Marine/Wildlife Biology Career Looks Like

The ideal career profile for maximum resilience and impact combines:

1. Deep biological expertise in a specific system or taxon — becoming a genuine expert on a group of organisms, an ecosystem, or a biological process. This is the foundation that cannot be automated.
2. Strong quantitative and computational skills — not necessarily at the level of a computer scientist, but enough to use AI tools effectively, write code for data analysis, and understand the strengths and limitations of AI outputs.
3. Field experience and practical skills — diving certifications, boat handling, animal handling permits, field identification skills. These differentiate you from desk-only researchers and pure data scientists.
4. Communication and leadership abilities — the capacity to explain complex findings to non-scientists, write compelling grant proposals, mentor students, and lead collaborative projects.
5. Adaptability and continuous learning — the willingness to adopt new tools, learn new methods, and pivot as the field evolves. The specific AI tools of 2026 will be obsolete by 2036, but the skill of learning and adapting to new tools is permanent.

Career trajectories that look strong for the next 20+ years:

• Conservation scientist with computational skills working for government agencies (NOAA, USFWS, EPA), NGOs (TNC, WWF, WCS), or international organizations (IUCN, UNEP).
• University professor combining field ecology with computational approaches — the most competitive applicants for faculty positions will have this combination.
• Conservation technology specialist developing and deploying AI tools for monitoring and management.
• Marine/wildlife policy advisor bringing scientific and technical expertise to decision-making.
• Environmental consultant with expertise in AI-powered monitoring and assessment methods.

How to Position Yourself as Someone Who USES AI Rather Than Being Replaced BY It

Mindset: Think of AI as the most powerful microscope or telescope ever invented — it's a tool that extends your capabilities, not a replacement for your brain and your fieldwork. The biologists who thrive will be the ones who:

1. Learn to use AI tools fluently: Don't fear them. Start using them now — play with iNaturalist, run AI analyses on your own data, take online courses in Python and machine learning.
2. Develop skills AI cannot replicate: Spend time in the field. Build natural history knowledge. Learn to observe carefully. Develop relationships with communities. Practice science communication. These are your moat.
3. Become a bridge: The most valuable person in any conservation organization will be the one who can translate between the AI engineers and the field biologists — someone who understands both the algorithms and the ecosystems.
4. Stay current but don't chase every trend: The AI tools will change rapidly. Focus on understanding principles (how ML works, what it can and cannot do) rather than mastering specific tools that may be obsolete in five years.
5. Publish and share your work: Using AI tools to accelerate your research output means you can build a publication record faster. Use this to establish yourself as an expert in your biological domain.
6. Network across disciplines: Go to both ecology conferences and tech conferences. Join WILDLABS.NET. Attend hackathons for conservation. Build relationships with data scientists and engineers.

Key Publications and Expert Voices

• Tuia et al. (2022), “Perspectives in machine learning for wildlife conservation,” Nature Communications: Comprehensive review by a consortium of ecologists and computer scientists outlining how ML is being applied across conservation and what's needed for responsible implementation. Emphasizes that ML should complement, not replace, ecological expertise.
• Christin, Hervet, Lecomte (2019), “Applications for deep learning in ecology,” Methods in Ecology and Evolution: Influential review that mapped the landscape of deep learning applications in ecology and identified key opportunities and challenges.
• Besson et al. (2022), “Towards the fully automated monitoring of ecological communities,” Ecology Letters: Argued that autonomous monitoring systems combining sensors, AI, and robotics could fundamentally change how ecologists study communities — but emphasized the ongoing need for ecological expertise to design systems and interpret outputs.
• Jane Lubchenco (former NOAA Administrator and special advisor): Has spoken extensively about how technology and data science must be integrated into marine conservation while keeping human decision-making central to policy.
• Enric Sala (National Geographic Explorer-in-Residence): His Pristine Seas program is one of the most technology-forward conservation initiatives, using AI-analyzed satellite data, drones, and genomics to make the case for marine protection.
• David Gruber (City University of New York): Pioneer in combining marine biology with technology, including biofluorescence research, underwater robotics, and AI-powered marine exploration.

Institutional Reports

NOAA AI Strategy:

• NOAA released its Artificial Intelligence Strategy in 2021 (updated through 2024–2025), identifying AI as a top priority across all mission areas: weather prediction, ocean monitoring, fisheries management, and climate science.
• Key NOAA AI initiatives include: AI-powered fisheries stock assessment, automated satellite imagery analysis for harmful algal blooms, machine learning for weather/climate forecasting, and AI for coral reef monitoring.
• NOAA's Center for Artificial Intelligence (NCAI) coordinates AI research and application across the agency.

NSF (National Science Foundation):

• NSF's “Harnessing the Data Revolution” (HDR) initiative has funded numerous projects at the intersection of AI and ecology/biology.
• The NSF BIO Directorate has increasingly funded projects that integrate computational and biological approaches, including AI for biodiversity research.
• NSF has also funded the development of ecological data infrastructure (NEON, LTER, iDigBio) that provides the training data necessary for ecological AI.

Professional Societies:

• The Ecological Society of America (ESA) has featured sessions on AI and machine learning at its annual meetings with increasing frequency, and published several perspective pieces in its journals (Ecology, Ecological Applications, Frontiers in Ecology and the Environment).
• The Society for Conservation Biology (SCB) has hosted symposia on conservation technology and AI.
• The American Fisheries Society has addressed AI in fisheries management in its journals and conferences.
• The Marine Technology Society focuses specifically on ocean technology including autonomous systems and AI.

Workforce Projections

• U.S. Bureau of Labor Statistics (BLS): Projects 5% growth for zoologists and wildlife biologists (2022–2032) and 6% growth for conservation scientists. Median salary for wildlife biologists was approximately $67,000 in 2023, with federal positions (NOAA, USGS, USFWS) typically paying higher.
• The Nature Conservancy's 2023 workforce analysis noted increasing difficulty hiring candidates with combined conservation + data science skills, suggesting strong demand.
• LinkedIn's 2024 Jobs on the Rise reports have consistently highlighted sustainability, environmental science, and data science roles as growing categories. The intersection of these fields is particularly hot.
• Conservation biology workforce diversity: Multiple reports have noted that the field remains less diverse than the general population, and that efforts to recruit diverse candidates are creating new opportunities for underrepresented groups.
• International demand: The Kunming-Montreal Global Biodiversity Framework (30x30 target) is creating enormous demand for conservation professionals worldwide, particularly those with monitoring and technology skills needed to track progress toward targets.
• Private sector growth: The market for biodiversity credits and nature-based solutions is growing rapidly (estimated market of $2+ billion by 2030), creating new career paths for biologists in conservation finance, environmental consulting, and ecosystem service assessment — all of which increasingly require computational skills.