In some ways, AI may turn out to be like the transistor economically—a big scientific discovery that scales well and that seeps into almost every corner of the economy. We don’t think much about transistors, or transistor companies, and the gains are very widely distributed. But we do expect our computers, TVs, cars, toys, and more to perform miracles.
Artificial intelligence has been undergoing an extraordinary development process for several years and is increasingly achieving capabilities that were long reserved exclusively for humans. Particularly in the area of research, we are currently experiencing remarkable progress: so-called “research agents”, specialized AI models that can independently take on complex research tasks, are rapidly gaining in importance. One prominent example is OpenAI's DeepResearch, which has already achieved outstanding results in various scientific benchmarks. Such AI-supported agents not only analyze large data sets, but also independently formulate research questions, test hypotheses, and even create scientific summaries of their results.
Alongside these technological breakthroughs, a new scientific practice is emerging that is no longer focused solely on direct research, but increasingly also on the management and coordination of these intelligent systems. In addition, there are so-called “computer-use agents” (CUA), which operate independently on the web and take over numerous processes autonomously. Another milestone is expected before the end of this year: the publication of the “SWE-Agent” (Software Engineering Agent), an AI system that independently optimizes and autonomously executes software development processes, which is also expected to lead to far-reaching changes in technical research fields.
These technological developments are raising fundamentally new questions about the future of academic research and, in particular, about the significance of traditional scientific degrees. In a world in which artificial intelligence can conduct research independently and in which laypersons may also be able to achieve high-quality scientific results with the help of these systems, does it still make sense to pursue a doctorate or other academic degrees? Will academic titles remain relevant in their current form or are we possibly on the verge of a fundamental change in the academic qualification structure?
What Is a PhD and What Purpose Does It Serve?
Some of the twists have been joyful; some have been hard. It’s been fun watching a steady stream of research miracles occur, and a lot of naysayers have become true believers.
The Ph.D. (“Doctor of Philosophy”) is the highest academic degree that can be obtained in the U.S. and many other countries. Despite the term “philosophy,” this title is by no means limited to philosophical subjects, but covers practically all scientific disciplines - from natural and engineering sciences to social sciences and humanities, economics and computer science.
Admission to a Ph.D. program in the U.S. usually requires at least a bachelor's degree. However, a master's degree is often also required or at least recommended. In addition to academic qualifications, proof of professional aptitude in the form of letters of recommendation, personal essays, research proposals and often standardized tests such as the GRE (Graduate Record Examination) are crucial. Furthermore, very good academic performance – measurable, for example, by the Grade Point Average (GPA) – is a prerequisite for being considered for renowned programs at all.
A Ph.D. program in the U.S. is usually very time-consuming and requires an intensive commitment over several years. On average, it takes about four to six years to complete a Ph.D., sometimes longer, especially in research-oriented subjects. This time is characterized by intensive independent research work, seminars, lectures, regular participation in scientific conferences, publications in scientific journals and often also teaching activities at the respective university.
The focus of the doctoral program is the independent development of an extensive research project, the so-called dissertation. This dissertation must contain new scientific findings, which are documented in the form of a comprehensive, independent research project. Doctoral students spend countless hours thoroughly analyzing their research topic, conducting experiments or empirical studies, sifting through extensive literature, and developing and validating theoretical and methodological concepts. At the end of the process, the doctoral student usually has to defend their dissertation orally before a panel of experts in order to present their research work in a credible and comprehensible way.
In short, a doctorate is not just another academic title, but a demanding and time-consuming process that requires a great deal of commitment, initiative and passion for a specific research topic. Those who choose this path invest many years in intensive research in order to make an important contribution to their field of expertise.
AI and Science
We are beginning to turn our aim beyond that, to superintelligence in the true sense of the word. We love our current products, but we are here for the glorious future. With superintelligence, we can do anything else. Superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own, and in turn massively increase abundance and prosperity.
Research itself is undergoing a profound transformation. More and more tasks traditionally performed by human scientists – especially doctoral students and postdocs – are being taken over by artificial intelligence. A vivid and well-known example is the AI model AlphaFold, which can predict protein structures with a level of accuracy and speed that would have been unthinkable just a few years ago. Previously, scientists, including many doctoral students, had to spend months or even years conducting meticulous experiments to elucidate protein structures. AlphaFold does the same work in a fraction of the time, although even this system cannot completely replace all facets of research.
It has long since ceased to be just about individual AI models that perform isolated tasks. It is already possible to combine different AI systems in what are known as agent systems. One agent, for example, takes on the research, collecting and filtering relevant scientific publications and data sources, while another agent then structures and interprets the results and even writes scientific texts. Such networked AI systems are already taking on large parts of the routine work that used to be done exclusively by human researchers.
https://www.linkedin.com/pulse/scientists-use-alphafold-ai-design-novel-drug-liver-cancer-colangelo
The most important factor in this development, however, is the speed with which AI models are improving. In an astonishingly short period of time, the systems are becoming more and more powerful, complex and precise. Already today, AI agents are emerging that not only support individual research processes, but can also carry out entire research projects independently – from the research question to data collection, evaluation and publication. It seems only a matter of time before advanced research agents can independently carry out both scientific experiments and their complete evaluation.
However, it must also be recognized that the use of AI in research has certain limitations. Certain processes, such as long-term biological studies, clinical research or social and cultural research fields, cannot necessarily be accelerated by pure AI support.
Intelligent agents need to operate interactively in the world in order to accomplish things and also to learn. But the world only moves so fast. Cells and animals run at a fixed speed so experiments on them take a certain amount of time which may be irreducible. The same is true of hardware, materials science, anything involving communicating with people, and even our existing software infrastructure.
Furthermore, in science many experiments are often needed in sequence, each learning from or building on the last. All of this means that the speed at which a major project—for example developing a cancer cure—can be completed may have an irreducible minimum that cannot be decreased further even as intelligence continues to increase.
These tasks are often closely linked to natural processes that cannot simply be scaled up or shortened by increased computing power. Despite impressive progress, some research work thus remains, at least for the time being, out of reach for artificial intelligence.
Conclusion
We expect the impact of AGI to be uneven. Although some industries will change very little, scientific progress will likely be much faster than it is today; this impact of AGI may surpass everything else.
At this point in time, it is difficult to make a clear prediction about how quickly the influence of artificial intelligence on academic research will develop. However, it is clear that AI does not yet work completely autonomously and it will therefore probably take a few more years before scientific research can be completely taken over by artificial intelligence.
However, the trend is clear: even today, AI models independently formulate important hypotheses and are able to handle large parts of the research process without direct human guidance. The long-term goal is for AI-based research agents to not only provide support, but to carry out entire research processes independently. In the future, these scientific agents could independently formulate research questions, conduct extensive literature analyses, plan and evaluate experiments independently, and finally provide valid new insights that were previously reserved for doctoral students and other researchers.
Traditionally, the original purpose of a doctorate was to write a dissertation that generated new scientific knowledge. To achieve this, the doctoral candidate had to invest countless hours of intensive research work: detailed and in-depth analyses of their research topic, extensive empirical studies or experiments, extensive literature research, and the independent development and validation of methodological and theoretical concepts. However, all these tasks could be taken over completely or largely autonomously by AI models in the near future.
Even more optimistically, it is possible that AI-enabled biological science will reduce the need for iteration in clinical trials by developing better animal and cell experimental models (or even simulations) that are more accurate in predicting what will happen in humans. This will be particularly important in developing drugs against the aging process, which plays out over decades and where we need a faster iteration loop.
In this context, a profound change is emerging in the course of which scientific degrees such as the Ph.D. may lose their traditional significance or at least need to be fundamentally redefined. It could be just a matter of a few years before scientific staff actually conduct independent research and thus take on the core of what has been the essential task and justification of a doctorate. Whether and how academic degrees will adapt to this change remains one of the crucial questions of the coming years.
It is no wonder, then, that Dario Amodei, CEO of Anthropic, is very optimistic about the future. AI will conduct research and generate knowledge independently in a matter of days that used to take decades.
To summarize the above, my basic prediction is that AI-enabled biology and medicine will allow us to compress the progress that human biologists would have achieved over the next 50-100 years into 5-10 years. I’ll refer to this as the “compressed 21st century”: the idea that after powerful AI is developed, we will in a few years make all the progress in biology and medicine that we would have made in the whole 21st century.
In this context, it is credible when the renowned doctor and scientist Prof. Derya Unutmaz says that he would no longer waste his time with doctorates (Ph.D.-programs) because they would soon be obsolete anyway. There is nothing to add to that.
—
Ready for more content from Kim Isenberg? Subscribe to FF Daily for free!
