What Is Computer Science? Research Areas, Funding, and Career Paths

Computer science is the study of computation — what can be computed, how to compute it efficiently and correctly, and how to design the hardware, software, and systems that carry computation out in practice. It sits at the intersection of mathematics, engineering, and (increasingly) empirical and behavioral science: some of the field asks purely abstract questions (what problems are solvable at all, and how fast can a solution run in the worst case), while other parts of it are experimental and applied, building real systems and measuring how they perform, fail, or get used. That range — from formal proof to systems engineering to studying how people interact with the systems built — is part of what makes computer science unusually broad among the sciences, and it is also why the field funds and organizes itself differently than most of the disciplines CASRAI covers elsewhere in this series.

What Computer Science Actually Studies

Computer science asks a recurring set of questions across nearly every one of its subfields: What is this problem, formally? Is it solvable at all, and if so, how efficiently? What is the right way to represent the data and structure the process? How do you build a system that does this reliably, at scale, and verify that it actually works? And, for the parts of the field closest to people, how do humans actually use, understand, or get harmed by the systems built to answer the questions above?

• Computability and complexity — which problems can be solved by any computer at all (computability theory), and, among the solvable ones, how the resources required (time, memory) grow as the problem gets larger (complexity theory) — the P vs. NP question is the field’s most famous open problem in this area.
• Algorithms and data structures — systematic, provably correct procedures for solving a class of problems, and the ways of organizing data that make those procedures efficient.
• Systems — how computation is actually carried out at scale: operating systems, networks, distributed systems, and the computer architecture underneath them.
• Software correctness and design — how to specify, build, test, and formally verify that a program does what it is supposed to do, and how to manage that process across large codebases and teams.
• Information and representation — how data, knowledge, and meaning are encoded, stored, retrieved, and reasoned over, from a single database record to a large trained model.
• Interaction — how people perceive, understand, and use computing systems, and how to design systems that serve human needs rather than just technical ones.

Unlike most of the natural sciences in this series, computer science studies an artifact of human design — computation itself, and the machines and programs that carry it out — rather than a naturally occurring phenomenon. That gives it a dual character: parts of the field (complexity theory, computability) are as mathematically rigorous and proof-driven as pure mathematics, while other parts (systems research, human–computer interaction, empirical software engineering) are experimental sciences in the ordinary sense, building something and measuring how it actually behaves under real conditions.

A Brief History of Computer Science

The field’s theoretical foundations predate the electronic computer by decades. Charles Babbage designed (but never completed) the Analytical Engine in the 1830s–1840s, a mechanical general-purpose computer; Ada Lovelace’s notes on it are widely credited as containing the first published algorithm intended for a machine. The modern theory of computation dates to 1936, when Alan Turing’s paper “On Computable Numbers” introduced what is now called the Turing machine — a formal model that still defines what it means for a problem to be computable at all. Around the same period, Claude Shannon’s 1937 master’s thesis showed that Boolean algebra could describe electrical switching circuits, the conceptual link between logic and hardware that digital computing is built on; his 1948 paper “A Mathematical Theory of Communication” separately founded information theory.

The first general-purpose electronic computers followed in the 1940s (ENIAC, built at the University of Pennsylvania and completed in 1945), alongside John von Neumann’s stored-program architecture — the design, describing a machine that stores both instructions and data in the same memory, that still underlies most computers built today. Academic computer science emerged as its own discipline starting in the 1960s, as universities split it out from mathematics and electrical engineering departments; the term “artificial intelligence” itself dates to the 1956 Dartmouth workshop organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon. ARPANET, the US Department of Defense–funded packet-switching network that began operating in 1969, became the direct technical precursor to the modern internet, and its funding agency — ARPA, later DARPA — remains a major funder of computing research today (see the funding section below). Software engineering as a named discipline dates to a 1968 NATO conference convened specifically to address what attendees called the “software crisis”: the growing gap between what programmers could reliably build and what increasingly complex systems demanded.

The field has gone through repeated cycles of AI optimism and retrenchment (the “AI winters” of the 1970s and late 1980s–1990s), followed by the deep-learning resurgence that accelerated sharply after 2012, when a convolutional neural network’s performance on the ImageNet visual-recognition benchmark demonstrated that large trained models could substantially outperform prior approaches on real tasks — a result widely credited with launching the current era of large-scale machine learning research.

How Computer Science Relates to Neighboring Disciplines

Computer science grew directly out of mathematics and electrical engineering and still overlaps heavily with both: discrete mathematics, logic, and probability theory underpin algorithms, complexity theory, and machine learning, while computer architecture and digital-circuit design sit on the electrical-engineering side of the field. Statistics and applied mathematics converge with computer science in data science and machine learning specifically — a working statistician and a machine-learning researcher increasingly draw on the same underlying theory, applied through different tooling and different disciplinary conventions.

Several of computer science’s subfields are themselves genuinely interdisciplinary. Computational linguistics and natural language processing sit between computer science and linguistics — see CASRAI’s what is linguistics guide for the language side of that boundary. Cognitive science and human–computer interaction draw on psychology, drawing methods and theory from how people perceive, learn, and make decisions — see CASRAI’s companion what is psychology guide. Bioinformatics and computational biology apply computer-science methods (algorithms, databases, machine learning) to biological data, overlapping with the disciplines covered in CASRAI’s what is genetics and what is molecular biology guides. Computational neuroscience and neuroscience-adjacent AI research overlap with the field covered in CASRAI’s what is neuroscience guide, and scientific computing and high-performance computing are essential infrastructure for computationally intensive fields such as the one covered in CASRAI’s what is astrophysics guide. Computational social science, applying computer-science methods to social data, connects the field to sociology and economics as well.

Major Branches and Subfields of Computer Science

Computer science is broad enough that most departments organize themselves into a set of recurring subfields, though the exact boundaries and names vary by institution:

• Theory of computation and algorithms — computability, complexity theory, algorithm design and analysis, cryptography’s theoretical foundations. The field’s most mathematically rigorous branch, closest to pure mathematics in method.
• Artificial intelligence and machine learning — systems that learn from data or reason toward a goal, including the major applied subfields of natural language processing, computer vision, and robotics. Currently the field’s fastest-growing and most heavily funded area, both federally and by industry.
• Systems, networking, and computer architecture — operating systems, distributed systems, computer networks, and the hardware architecture underneath them; concerned with how computation scales, stays available, and performs under real load.
• Databases and data management — how large volumes of structured and unstructured data are stored, indexed, queried, and kept consistent; overlaps heavily with data science as an applied discipline.
• Software engineering — the design, construction, testing, verification, and maintenance of software systems, including formal methods for proving a program meets its specification.
• Human–computer interaction (HCI) — how people perceive, learn, and interact with computing systems, and how to design interfaces and systems around human capability and behavior rather than purely technical constraints. Frequently involves human-subjects research and its associated IRB/ethics oversight, a genuine point of overlap with CASRAI’s broader research-compliance coverage.
• Computer security and cryptography — protecting systems and data against adversarial actors, and the mathematical foundations (cryptography) that much of that protection depends on.
• Programming languages and compilers — the design of languages used to express computation, and the tools that translate that expression into something a machine can run efficiently and correctly.
• Computer graphics and visualization — rendering, modeling, and displaying visual information, from scientific visualization to real-time rendering.
• Computational science and high-performance computing — building and using large-scale computing infrastructure (supercomputers, GPU clusters) to simulate and analyze problems from other sciences, effectively the shared infrastructure layer for much of computationally intensive research across disciplines.

Many active research programs deliberately sit across more than one of these branches — a systems researcher studying machine-learning infrastructure, or a security researcher analyzing machine-learning models, is doing legitimate work in two subfields at once.

Who Funds Computer Science Research

In the United States, the National Science Foundation (NSF) is the primary federal funder of academic, non-biomedical computer science research, through its Directorate for Computer and Information Science and Engineering (CISE). NSF has periodically reorganized CISE’s internal structure; as currently organized, CISE groups its work around thematic science areas — including Computing and AI Foundations, Software and Systems Foundations, Cyber, Physical and Intelligent Systems, Human-Centered Computing and Education, Center-Scale and Testbeds, and the Office of Advanced Cyberinfrastructure (OAC), which funds shared computing infrastructure used across many disciplines — rather than the narrower divisional structure (researchers may still see older references to divisions like CCF or IIS) the directorate used previously. See CASRAI’s dictionary entry on NSF CISE for more on the directorate’s role.

The Defense Advanced Research Projects Agency (DARPA), part of the US Department of Defense, has funded foundational computing research since the field’s earliest days (it funded ARPANET) and remains a major, if more mission-directed, funder of computer science research, particularly in AI, security, and networking. The Department of Energy’s Office of Advanced Scientific Computing Research (ASCR) funds high-performance and scientific computing research, much of it conducted at DOE national laboratories, aimed at the computational infrastructure other sciences depend on for large-scale simulation. On the biomedical side, the National Institutes of Health funds computer science research that intersects with health and biology mainly through the National Library of Medicine (NLM), whose mission explicitly includes conducting and supporting research, development, and training in biomedical informatics and health information technology, both intramurally and through extramural grants; other NIH institutes fund computational methods relevant to their own biomedical focus areas as well.

Computer science also has a funding source most other academic disciplines in this series do not have at comparable scale: substantial direct industry funding, in the form of sponsored research agreements, unrestricted research gifts, PhD fellowships, and, increasingly, entire corporate research labs (e.g., Microsoft Research, Google DeepMind, and IBM Research, among others) that employ researchers who also publish, review, and collaborate within the same academic venues as university faculty. That industry presence shapes the field’s funding landscape and research culture in ways distinct from a discipline funded almost entirely through federal grants and private foundations. Among private foundations, the Simons Foundation is a genuinely notable funder of theoretical computer science specifically, having established the Simons Institute for the Theory of Computing at UC Berkeley; the Alfred P. Sloan Foundation has a long history of funding computing and, more recently, technology-and-society research; and newer entrants such as Schmidt Sciences fund large-scale AI research initiatives. Researchers should verify current program details directly with each funder, since specific programs and eligibility change over time.

Typical Research Methods, Tools, and Practices

Computer science research methods split roughly along the same theoretical/empirical line the field itself does. Theoretical work proceeds through mathematical proof and formal analysis — establishing that an algorithm is correct, bounding its worst-case running time or memory use, or proving a security property holds under a defined adversarial model. Empirical and systems work looks more like an experimental science: building a system or model, then evaluating it against benchmarks, baselines, and real workloads, with results reported as measured performance rather than proof. Human-subjects research, common in HCI and some AI-evaluation work, follows the same institutional review board (IRB) and human-subjects protection requirements that apply across CASRAI’s broader research-compliance content, since studies involving human participants’ behavior, data, or interaction with a system fall under the same regulatory framework regardless of the researcher’s home department.

Reproducibility and shared infrastructure are central to how the field verifies its own results. Standard datasets and benchmarks let different research groups compare methods on a common basis; open-source code release has become a near-default expectation at most top venues, and the ACM Artifact Review and Badging program formalizes that expectation into a peer-reviewed evaluation of whether a paper’s code and data actually reproduce its claimed results. Preprint culture in computer science is heavily concentrated on arXiv’s computer-science categories (governed by the same arXiv endorsement system used across other fields), supplemented by field-specific outlets like TechRxiv, IEEE’s preprint server. Simulation is widely used both as a research method in its own right (modeling network or systems behavior at scales too large to instrument directly) and as infrastructure for other computational sciences.

Career and Training Pathways

A research career in computer science typically runs through a PhD, structured similarly to other research-intensive fields: coursework and qualifying exams in the first one to two years, followed by original dissertation research conducted under a faculty advisor, culminating in a written dissertation and defense. Compared to many of the sciences covered elsewhere in this series, computer science PhDs move into industry research roles — not only academic or national-laboratory positions — unusually often, and a postdoctoral position, while common in some subfields (particularly theory and some AI research areas), is less universal a step than it is in, say, the biomedical sciences; many PhD graduates move directly into faculty positions or industry research labs.

The field’s major professional societies are the Association for Computing Machinery (ACM), whose numerous special interest groups (SIGs) organize the discipline’s conference-centered publication culture — conference papers, not journal articles, are the primary venue for reporting new results in most CS subfields, unlike most other sciences in this series — and the IEEE Computer Society, the computing arm of the Institute of Electrical and Electronics Engineers. USENIX is a well-known systems-focused professional association and conference organizer. Outside academia and national labs, computer scientists work across software and hardware industries broadly, government and defense research, and increasingly in policy and standards-setting roles as computing systems (particularly AI systems) draw more regulatory attention.

Frequently Asked Questions

What is computer science, in one sentence?
Computer science is the study of computation: what problems can be solved algorithmically, how to solve them efficiently and correctly, and how to design the software and hardware systems that carry that computation out in practice.

Is computer science the same as software engineering?
No, though they overlap heavily and many people use the terms loosely. Computer science is the broader academic discipline, including highly theoretical branches (complexity theory, computability) that have no direct engineering application; software engineering is the specific subfield concerned with how software is designed, built, tested, and maintained, generally applied rather than theoretical in orientation.

Is computer science a STEM field, and is it a “real” science?
Yes to both, though it has an unusual character among the sciences: it studies an artifact of human design (computation) rather than a naturally occurring phenomenon, which is why parts of it (theory of computation) look like pure mathematics while other parts (systems research, HCI) function as ordinary experimental science, building something and measuring how it behaves.

What can you do with a computer science degree?
Career paths span software and hardware engineering across virtually every industry, corporate and government research labs, cybersecurity, data science and machine learning roles, academia, and increasingly policy and standards work; the specific path depends heavily on specialization and degree level, from bachelor’s-level software engineering roles to PhD-level research positions.

How is computer science research typically funded in the United States?
Mainly through the National Science Foundation’s CISE directorate, with DARPA, the Department of Energy’s ASCR office, and (for biomedically framed work) NIH’s National Library of Medicine also playing major roles — alongside substantial direct industry funding that is unusually significant compared to most other academic research fields. See the funding section above for how these sources divide the field’s territory.

What is the difference between computer science and data science?
Data science is generally treated as an applied, interdisciplinary field that draws on computer science (algorithms, databases, machine learning), statistics, and domain expertise to extract insight from data; computer science is the broader discipline that data science draws several of its core methods from, alongside data science’s equally significant statistical and domain-specific components.

Where Computer Science Fits Among the Sciences

For a broader map of how computer science relates to the full set of major scientific disciplines — from physics and mathematics through to biology and the social sciences — see CASRAI’s overview guide to the branches of science, which this page is part of a companion series alongside. Related guides in that series include what is physics, what is biostatistics, and what is sociology, each of which shares a genuine methodological or infrastructural border with computer science.