Glossary 2025

An interface that allows software programs to communicate with each other and share information or services without revealing how they work internally.

The process by which goods are ascribed market value. This includes immaterial goods, such as models and methods, that become a tradeable asset.

A virus that infects and destroys bacteria, used in research and sometimes as a therapeutic option for bacterial infections that are resistant to conventional treatments.

The “letters” in DNA or RNA that connect to carry genetic information.

A measurable biological indicator, such as a metabolite or genetic feature, that signals a particular health condition or disease.

AI systems whose internal reasoning is hard to interpret, even by their developers.

The process by which powerful technologies become widely available across countries, groups, or individuals.

Molecules that exist in two forms that are mirror images of each other, like left and right hands, but cannot be perfectly aligned.

A technique that uses very cold temperatures and electron beams to visualize the detailed structures of biological molecules.

Technology with the potential to be used for both legitimate civilian and military purposes, or with the potential to be misused for malicious purposes.

When a piece of DNA or RNA contains the instructions for making a specific protein.

Large, general-purpose AI systems trained on broad datasets that can be adapted to many tasks (e.g., ChatGPT, Claude).

The graphic processing unit (GPU) is a specialized computing chip that was originally developed to accelerate computer graphics and image processing in PCs, smartphones, and games consoles. Due to their ability to calculate algorithms in massive parallel processing, GPUs are also suitable for non-graphical applications such as training neural networks and mining cryptocurrencies. In contrast to the central processing unit (CPU), which can be understood as the control center of a computer, GPUs are generally optimized for continuous operation under full load and generate the commensurate electrical power of several hundred watts, with the corresponding power and cooling requirements. Due to the massive boom, GPUs are also increasingly being optimized specifically for AI applications and produced as complete device units that can be interconnected in hundreds or thousands in specialized data centers. Such device units can achieve continuous power consumption of more than 1,000 watts, which has a significant impact on the power supply and cooling requirements of data centers.

Two mirror-image forms of the chemical building blocks that make up DNA and RNA; natural life uses only L-forms, while D-forms are synthetic mirror versions.

Depending on the type of training data used and the form of possible user interaction of the AI algorithm, a distinction is made between different types. In large language models (LLM), text data are used for training and output. Users therefore chat with the AI algorithm. In large multimodal models (LMM), image, video, and audio data are also used in training and for interaction between AI and the user. In this case, AI is therefore able to process and produce different media. However, given the speed of technological progress, these boundaries are fluid, depending on the requirements of the field of application. The next big step that tech companies are working on is the vision of what has been dubbed artificial general intelligence (AGI), which is no longer optimized to solve a specific problem, but should be a highly flexible artificial system that is equivalent or superior to human cognitive abilities in all areas. An AGI model should be able to adapt to problems and develop solutions without having been specifically trained for them.

Increasing an AI model’s size or training data to improve performance.

A method of analyzing genetic material that allows for the rapid sequencing of large amounts of DNA or RNA. Compared to traditional sequencing techniques (e.g., Sanger sequencing), NGS can simultaneously sequence millions of small fragments of DNA.

A nuclear reactor design utilizing reactor technology that has not been used in existing commercial reactors. These designs foresee the use of new types or forms of fuel and/or different coolants, such as molten salt, lead, or inert gases, unlike commercial reactors, which are mostly water-cooled.

A technique that uses magnetic fields to study the structure of molecules in solution, providing information about their shape and dynamics.

Technical measures and inspections to prevent the clandestine diversion and use of nuclear material for military purposes. World-wide, safeguards inspections are conducted by the IAEA and are a treaty obligation under the Non-Proliferation Treaty for non-nuclear-weapons states (NNWS). Some facilities in nuclear-weapons states are also under safeguards as part of Voluntary Offer Agreements.

Molecules that carry the instructions for how living things grow and function.

An enzyme that helps build DNA or RNA strands by connecting building blocks called nucleotides in the right order.

A nuclear reactor design utilizing light water as coolant and neutron moderator. A pressure vessel surrounding the reactor keeps the water in a liquid state throughout operation. PWRs are the most common design for commercial nuclear power plants, which are usually operated with fuel with 3-5% uranium enrichment.

A structure in or on a cell that binds specific molecules, triggering a response or allowing the cell to recognize certain substances.

A molecular machine inside cells that reads genetic instructions and assembles proteins from amino acids.

A nuclear reactor producing no more than 300 megawatts of electrical power. These reactors are planned to be produced as modules in an off-site factory and then be transported to a power plant. Many concepts foresee the combination of several modules on the same site.

Spatially concentrated networks of governments, universities, companies set up to create synergies between science and economic markets. A well-known example of a technopole is Silicon Valley in the U.S.

The process of separating the isotopes of natural uranium to raise the fraction of 235U within the uranium. Most traditional nuclear reactors require 3-5% enrichment to operate, while nuclear weapons typically require enrichments above 80%.

A method that uses X-rays to determine how atoms are arranged in a crystallized molecule, helping scientists see its 3D shape.

Explainable artificial intelligence approaches are intended to counteract the “black box” tendency of machine learning, i.e., the fact that it is not clear why an AI algorithm has reached a certain decision. Although it is technically possible to monitor the internal processing of a query within the model of an AI, no deterministic conclusions can be drawn about the actual reasoning process. It is therefore impossible to explain AI decisions in terms of “for an input (a), the result (b) was generated on the basis of the learned facts (X) and (Y)”, as it is common in human communication. Explainable artificial intelligence approaches are intended to make these chains of reasoning visible as an extension of an AI model.