AI Art Examples

AI Art Examples — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

80 Million Tiny Images

80 Million Tiny Images is a dataset intended for training machine-learning systems constructed by Antonio Torralba, Rob Fergus, and William T. Freeman in a collaboration between MIT and New York University. It was published in 2008. The dataset has size 760 GB. It contains 79,302,017 32×32-pixel color images, scaled down from images scraped from the World Wide Web over 8 months. The images are classified into 75,062 classes. Each class is a non-abstract noun in WordNet. Images may appear in more than one class. The dataset was motivated by non-parametric models of neural activations in the visual cortex upon seeing images. The CIFAR-10 dataset uses a subset of the images in this dataset, but with independently generated labels, as the original labels were not reliable. The CIFAR-10 set has 6000 examples of each of 10 classes, and the CIFAR-100 set has 600 examples of each of 100 non-overlapping classes. == Construction == It was first reported in a technical report in April 2007, during the middle of the construction process, when there were only 73 million images. The full dataset was published in 2008. They began with all 75,846 non-abstract nouns in WordNet, and then for each of these nouns, they scraped 7 image search engines: Altavista, Ask.com, Flickr, Cydral, Google, Picsearch, and Webshots. After 8 months of scraping, they obtained 97,245,098 images. Since they did not have enough storage, they downsized the images to 32×32 as they were scraped. After gathering, they removed images with zero variance and intra-word duplicate images, resulting in the final dataset. Out of the 75,846 nouns, only 75,062 classes had any results, so the other nouns did not appear in the final dataset. The number of images per noun follows a Zipf-like distribution, with 1056 images per noun on average. To prevent a few nouns taking up too many images, they put an upper bound of at most 3000 images per noun. == Retirement == The 80 Million Tiny Images dataset was retired from use by its creators in 2020, after a paper by researchers Abeba Birhane and Vinay Prabhu found that some of the labeling of several publicly available image datasets, including 80 Million Tiny Images, contained racist and misogynistic slurs which were causing models trained on them to exhibit racial and sexual bias. The dataset also contained offensive images. Following the release of the paper, the dataset's creators removed the dataset from distribution, and requested that other researchers not use it for further research and to delete their copies of the dataset.
Read more →
Lighthill report

Artificial Intelligence: A General Survey, commonly known as the Lighthill report, is a scholarly article by James Lighthill, published in Artificial Intelligence: a paper symposium in 1973. It was compiled by Lighthill for the British Science Research Council as an evaluation of academic research in the field of artificial intelligence (AI). The report gave a very pessimistic prognosis for many core aspects of research in this field, stating that "In no part of the field have the discoveries made so far produced the major impact that was then promised". It "formed the basis for the decision by the British government to end support for AI research in most British universities", contributing to an AI winter in the United Kingdom. == Publication history == It was commissioned by the SRC in 1972 for Lighthill to "make a personal review of the subject [of AI]". Lighthill completed the report in July. The SRC discussed the report in September, and decided to publish it, together with some alternative points of view by Stuart Sutherland, Roger Needham, Christopher Longuet-Higgins, and Donald Michie. The SRC's decision to invite the report was partly a reaction to high levels of discord within the University of Edinburgh's Department of Artificial Intelligence, one of the earliest and biggest centres for AI research in the UK. On May 9, 1973, Lighthill debated several leading AI researchers (Donald Michie, John McCarthy, Richard Gregory) at the Royal Institution in London concerning the report. == Content == While the report was supportive of research into the simulation of neurophysiological and psychological processes, it was "highly critical of basic research in foundational areas such as robotics and language processing". The report stated that AI researchers had failed to address the issue of combinatorial explosion when solving problems within real-world domains. That is, the report states that whilst AI techniques may have worked within the scope of small problem domains, the techniques would not scale up well to solve more realistic problems. The report represents a pessimistic view of AI that began after early excitement in the field. The report divides AI research into three categories: Advanced Automation ("A"): applications of AI, such as optical character recognition, mechanical component design and manufacture, missile perception and guidance, etc. Computer-based Central Nervous System research ("C"): building computational models of human brains (neurobiology) and behavior (psychology). Bridge, or Building Robots ("B"): research that combines categories A and C. This category is intentionally vague. Projects in category A had had some success, but only in restricted domains where a large quantity of detailed knowledge was used in designing the program. This was disappointing to researchers who hoped for generic methods. Due to the issue of the combinatorial explosion, the amount of detailed knowledge required by the program quickly grew too large to be entered by hand, thus restricting projects to restricted domains. Projects in category C had had some measure of success. Artificial neural networks were successfully used to model neurobiological data. SHRDLU demonstrated that human use of language, even in fine details, depends on the semantics or knowledge, and is not purely syntactical. This was influential in psycholinguistics. Attempts to extend SHRDLU to larger domains of discourse was considered impractical, again due to the issue of the combinatorial explosion. Projects in category B were held to be failures. One important project, that of "programming and building a robot that would mimic human ability in a combination of eye-hand co-ordination and common-sense problem solving", was considered entirely disappointing. Similarly, chess playing programs were no better than human amateurs. Due to the combinatorial explosion, the run-time of general algorithms quickly grew impractical, requiring detailed problem-specific heuristics. The report stated that it was expected that within the next 25 years, category A would simply become applied technologies engineering, C would integrate with psychology and neurobiology, while category B would be abandoned.
Read more →
Oracle Database

Oracle AI Database (commonly referred to as Oracle Database, Oracle DBMS, Oracle Autonomous Database, or simply as Oracle) is a proprietary multi-model database management system produced and marketed by Oracle Corporation. It is a database commonly used for running online transaction processing (OLTP), data warehousing (DW) and mixed (OLTP & DW) database workloads. Oracle AI Database uses SQL for database updating and retrieval. Oracle Database runs on-premises, on Oracle engineered systems such as Oracle Exadata, on Oracle Cloud Infrastructure, and as a managed Autonomous Database service. It is also offered inside Microsoft Azure, Google Cloud, and Amazon Web Services data centers through Oracle's multicloud offerings. The current long-term support release, Oracle AI Database 26ai, became available in the cloud and on Oracle engineered systems in October 2025 and on-premises for Linux x86-64 in January 2026. == History == Larry Ellison and his two friends and former co-workers, Bob Miner and Ed Oates, started a consultancy called Software Development Laboratories (SDL) in 1977, later Oracle Corporation. SDL developed the original version of the Oracle software. The name Oracle comes from the code-name of a Central Intelligence Agency-funded project Ellison had worked on while formerly employed by Ampex; the CIA was Oracle's first customer, and allowed the company to use the code name for the new product. Ellison wanted his database to be compatible with IBM System R, but that company's Don Chamberlin declined to release its error codes. By 1985 Oracle advertised, however, that "Programs written for SQL/DS or DB2 will run unmodified" on the many non-IBM mainframes, minicomputers, and microcomputers its database supported "Because all versions of ORACLE are identical". Later releases introduced capabilities associated with successive eras of the product, including PL/SQL stored procedures and triggers in Oracle7 (1992), Real Application Clusters in Oracle9i (2001), grid infrastructure and automatic management in Oracle 10g (2003), the multitenant architecture and In-Memory Column Store in Oracle Database 12c (2013), and AI Vector Search and JSON Relational Duality in Oracle Database 23ai (2024). In October 2025 Oracle rebranded the 23ai line as Oracle AI Database 26ai. (see Release History) == Architecture == An Oracle Database system consists of an instance and a database. The instance is a set of memory structures and background processes; the database is the set of files that store data. The instance exists only in memory, and a single instance is associated with one multitenant container database. The principal memory structures are the System Global Area, which is shared, and the Program Global Areas, which are private to individual processes. The shared pool, database buffer cache, and redo log buffer are components of the System Global Area, and the optional In-Memory Column Store also resides there. Background processes operate on the database files and use these memory structures; they include the database writer, the log writer, the checkpoint process, and the system and process monitor processes. Server processes handle connections from client programs and run their SQL statements. Storage is organized logically and physically. Logically, data is held in tablespaces composed of segments, extents, and data blocks. Physically, the database comprises datafiles, control files, and online redo log files, with archived redo logs supporting media recovery. == High Availability and Scalability == Oracle Database includes several technologies for high availability, disaster recovery, and scale. Oracle Real Application Clusters allows multiple instances on separate servers to access one shared database concurrently; it was introduced with Oracle9i in 2001. Oracle Data Guard maintains standby databases synchronized with a primary database, and Active Data Guard additionally allows read-only workloads on a standby while it applies changes. Oracle GoldenGate performs logical replication and data integration across heterogeneous systems. Native sharding, introduced in Oracle Database 12c Release 2, distributes one logical database across independent shards. Oracle Exadata is an engineered system that pairs database servers with storage servers and offloads operations such as filtering to the storage tier; it is available on-premises, in Oracle Cloud Infrastructure, and through Cloud@Customer. == Notable Features == AI Vector Search adds a vector data type, vector indexes, and vector distance operators to the database. These allow similarity search over machine-learning embeddings to be expressed in SQL and combined with queries over relational, JSON, spatial, and graph data. It became generally available in Oracle Database 23ai. JSON Relational Duality exposes the same data both as relational tables and as JSON documents through duality views, so that an application can read and write either representation of the data. It became generally available in Oracle Database 23ai. In-Memory Column Store maintains a column-oriented copy of selected tables in memory in addition to the row-oriented format, and the optimizer can use the columnar copy for analytic queries. It was introduced in Oracle Database 12c Release 1.Partitioning divides large tables and indexes into independently managed pieces. Advanced Compression and Hybrid Columnar Compression are compression features for transactional and warehouse data respectively. == Data Types == Oracle AI Database supports a variety of data types and data models within a single system. These include traditional relational data types as well as semi-structured, unstructured, and specialized data formats, enabling different types of data to be stored and queried together. == Releases and versions == Oracle products follow a custom release-numbering and -naming convention. The "ai" in the current release, Oracle AI Database 26ai, stands for "Artificial Intelligence". Previous releases (e.g. Oracle Database 19c, 10g, and Oracle9i Database) have used suffixes of "c", "g", and "i" which stand for "Cloud", "Grid", and "Internet" respectively. Prior to the release of Oracle8i Database, no suffixes featured in Oracle AI Database naming conventions. There was no v1 of Oracle AI Database, as Ellison "knew no one would want to buy version 1". For some database releases, Oracle also provides an Express Edition (XE) that is free to use. Oracle AI Database release numbering has used the following codes: The Introduction to Oracle AI Database includes a brief history on some of the key innovations introduced with each major release of Oracle AI Database. See My Oracle Support (MOS) note Release Schedule of Current Database Releases (Doc ID 742060.1) for the current Oracle AI Database releases and their patching end dates. == Patch updates and security alerts == Prior to Oracle Database 18c, Oracle Corporation released Critical Patch Updates (CPUs) and Security Patch Updates (SPUs) and Security Alerts to close security vulnerabilities. These releases are issued quarterly; some of these releases have updates issued prior to the next quarterly release. Starting with Oracle Database 18c, Oracle Corporation releases Release Updates (RUs) and Release Update Revisions (RURs). RUs usually contain security, regression (bug), optimizer, and functional fixes which may include feature extensions as well. RURs include all fixes from their corresponding RU but only add new security and regression fixes. However, no new optimizer or functional fixes are included. == Competition == In the market for relational databases, Oracle AI Database competes against commercial products such as IBM Db2 and Microsoft SQL Server. Oracle and IBM tend to battle for the mid-range database market on Unix and Linux platforms, while Microsoft dominates the mid-range database market on Microsoft Windows platforms. However, since they share many of the same customers, Oracle and IBM tend to support each other's products in many middleware and application categories (for example: WebSphere, PeopleSoft, and Siebel Systems CRM), and IBM's hardware divisions work closely with Oracle on performance-optimizing server-technologies (for example, Linux on IBM Z). Niche commercial competitors include Teradata (in data warehousing and business intelligence), Software AG's ADABAS, Sybase, and IBM's Informix, among many others. In the cloud, Oracle AI Database competes against the database services of AWS, Microsoft Azure, and Google Cloud Platform. Increasingly, the Oracle AI Database products compete against open-source software relational and non-relational database systems such as PostgreSQL, MongoDB, Couchbase, Neo4j, ArangoDB and others. Oracle acquired Innobase, supplier of the InnoDB codebase to MySQL, in part to compete better against open source alternatives, and acquired Sun Microsystems, owner of MySQL, in 2010. Database products licensed as open
Read more →
Hubert Dreyfus

Hubert Lederer Dreyfus ( DRY-fəs; October 15, 1929 – April 22, 2017) was an American philosopher and a professor of philosophy at the University of California, Berkeley. His main interests included phenomenology, existentialism and the philosophy of both psychology and literature, as well as the philosophical implications of artificial intelligence. He was widely known for his exegesis of Martin Heidegger, which critics labeled "Dreydegger". Dreyfus was featured in Tao Ruspoli's film Being in the World (2010), and was among the philosophers interviewed by Bryan Magee for the BBC Television series The Great Philosophers (1987). The Futurama character Professor Hubert Farnsworth is partly named after him, writer Eric Kaplan having been a former student. == Life and career == Dreyfus was born on 15 October 1929, in Terre Haute, Indiana, to Stanley S. and Irene (Lederer) Dreyfus. He attended Harvard University from 1947. With a senior honors thesis on Causality and Quantum Theory (for which W. V. O. Quine was the main examiner) he was awarded a B.A. summa cum laude in 1951 and joined Phi Beta Kappa. He was awarded a M.A. in 1952. He was a Teaching Fellow at Harvard from 1952 to 1953 (as he was again in 1954 and 1956). Then, on a Harvard Sheldon traveling fellowship, Dreyfus studied at the University of Freiburg from 1953 to 1954. During this time he had an interview with Martin Heidegger. Sean D. Kelly records that Dreyfus found the meeting 'disappointing.' A brief mention of it was made by Dreyfus during his 1987 BBC interview with Bryan Magee in remarks that are revealing of both his and Heidegger's opinion of the work of Jean-Paul Sartre. Between 1956 and 1957, Dreyfus undertook research at the Husserl Archives at the University of Louvain on a Fulbright Fellowship. Towards the end of his stay, his first (jointly authored) paper "Curds and Lions in Don Quijote" would appear in print. After acting as an instructor in philosophy at Brandeis University (1957–1959), he attended the Ecole Normale Supérieure, Paris, on a French government grant (1959–1960). From 1960, first as an instructor, then as an assistant and then associate professor, Dreyfus taught philosophy at the Massachusetts Institute of Technology (MIT). In 1964, with his dissertation Husserl's Phenomenology of Perception, he obtained his Ph.D. from Harvard. (Due to his knowledge of Husserl, Dagfinn Føllesdal sat on the thesis committee but he has asserted that Dreyfus "was not really my student.") That same year, his co-translation (with his first wife) of Sense and Non-Sense by Maurice Merleau-Ponty was published. Also in 1964, and whilst still at MIT, he was employed as a consultant by the RAND Corporation to review the work of Allen Newell and Herbert A. Simon in the field of artificial intelligence (AI). This resulted in the publication, in 1965, of the "famously combative" Alchemy and Artificial Intelligence, which proved to be the first of a series of papers and books attacking the AI field's claims and assumptions. The first edition of What Computers Can't Do would follow in 1972, and this critique of AI (which has been translated into at least ten languages) would establish Dreyfus's public reputation. However, as the editors of his Festschrift noted: "the study and interpretation of 'continental' philosophers... came first in the order of his philosophical interests and influences." === Berkeley === In 1968, although he had been granted tenure, Dreyfus left MIT and became an associate professor of philosophy at the University of California, Berkeley, (winning, that same year, the Harbison Prize for Outstanding Teaching). In 1972 he was promoted to full professor. Though Dreyfus retired from his chair in 1994, he continued as professor of philosophy in the Graduate School (and held, from 1999, a joint appointment in the rhetoric department). He continued to teach philosophy at UC Berkeley until his last class in December 2016. Dreyfus was elected a fellow of the American Academy of Arts and Sciences in 2001. He was also awarded an honorary doctorate for "his brilliant and highly influential work in the field of artificial intelligence" and his interpretation of twentieth century continental philosophy by Erasmus University. Dreyfus died on April 22, 2017. His younger brother and sometimes collaborator, Stuart Dreyfus, is a professor emeritus of industrial engineering and operations research at the University of California, Berkeley. == Dreyfus' criticism of AI == Dreyfus' critique of artificial intelligence (AI) concerns what he considers to be the four primary assumptions of AI research. The first two assumptions are what he calls the "biological" and "psychological" assumptions. The biological assumption is that the brain is analogous to computer hardware and the mind is analogous to computer software. The psychological assumption is that the mind works by performing discrete computations (in the form of algorithmic rules) on discrete representations or symbols. Dreyfus claims that the plausibility of the psychological assumption rests on two others: the epistemological and ontological assumptions. The epistemological assumption is that all activity (either by animate or inanimate objects) can be formalized (mathematically) in the form of predictive rules or laws. The ontological assumption is that reality consists entirely of a set of mutually independent, atomic (indivisible) facts. It's because of the epistemological assumption that workers in the field argue that intelligence is the same as formal rule-following, and it's because of the ontological one that they argue that human knowledge consists entirely of internal representations of reality. On the basis of these two assumptions, workers in the field claim that cognition is the manipulation of internal symbols by internal rules, and that, therefore, human behaviour is, to a large extent, context free (see contextualism). Therefore, a truly scientific psychology is possible, which will detail the 'internal' rules of the human mind, in the same way the laws of physics detail the 'external' laws of the physical world. However, it is this key assumption that Dreyfus denies. In other words, he argues that we cannot now (and never will be able to) understand our own behavior in the same way as we understand objects in, for example, physics or chemistry: that is, by considering ourselves as things whose behaviour can be predicted via 'objective', context free scientific laws. According to Dreyfus, a context-free psychology is a contradiction in terms. Dreyfus's arguments against this position are taken from the phenomenological and hermeneutical tradition (especially the work of Martin Heidegger). Heidegger argued that, contrary to the cognitivist views (on which AI has been based), our being is in fact highly context-bound, which is why the two context-free assumptions are false. Dreyfus doesn't deny that we can choose to see human (or any) activity as being 'law-governed', in the same way that we can choose to see reality as consisting of indivisible atomic facts... if we wish. But it is a huge leap from that to state that because we want to or can see things in this way that it is therefore an objective fact that they are the case. In fact, Dreyfus argues that they are not (necessarily) the case, and that, therefore, any research program that assumes they are will quickly run into profound theoretical and practical problems. Therefore, the current efforts of workers in the field are doomed to failure. Dreyfus argues that to get a device or devices with human-like intelligence would require them to have a human-like being-in-the-world and to have bodies more or less like ours, and social acculturation (i.e. a society) more or less like ours. (This view is shared by psychologists in the embodied psychology (Lakoff and Johnson 1999) and distributed cognition traditions. His opinions are similar to those of robotics researchers such as Rodney Brooks as well as researchers in the field of artificial life.) Contrary to a popular misconception, Dreyfus never predicted that computers would never beat humans at chess. In Alchemy and Artificial Intelligence, he only reported (correctly) the state of the art of the time: "Still no chess program can play even amateur chess." Daniel Crevier writes: "time has proven the accuracy and perceptiveness of some of Dreyfus's comments. Had he formulated them less aggressively, constructive actions they suggested might have been taken much earlier." == Webcasting philosophy == When UC Berkeley and Apple began making a selected number of lecture classes freely available to the public as podcasts beginning around 2006, a recording of Dreyfus teaching a course called "Man, God, and Society in Western Literature – From Gods to God and Back" rose to the 58th most popular webcast on iTunes. These webcasts have attracted the attention of many, including non-academics, to Dreyfus and his
Read more →
VACUUM

VACUUM is a set of normative guidance principles for achieving training and test dataset quality for structured datasets in data science and machine learning. The garbage-in, garbage out principle motivates a solution to the problem of data quality but does not offer a specific solution. Unlike the majority of the ad-hoc data quality assessment metrics often used by practitioners VACUUM specifies qualitative principles for data quality management and serves as a basis for defining more detailed quantitative metrics of data quality. VACUUM is an acronym that stands for: valid accurate consistent uniform unified model
Read more →
Automatic1111

AUTOMATIC1111 Stable Diffusion Web UI (SD WebUI, A1111, or Automatic1111) is an open source generative artificial intelligence program that allows users to generate images from a text prompt. It uses Stable Diffusion as the base model for its image capabilities together with a large set of extensions and features to customize its output. == History == SD WebUI was released on GitHub on August 22, 2022, by AUTOMATIC1111, 1 month after the initial release of Stable Diffusion. At the time, Stable Diffusion could only be run via the command line. SD WebUI quickly rose in popularity and has been described as "the most popular tool for running diffusion models locally." SD WebUI is one of the most popular user interfaces for Stable Diffusion, together with ComfyUI. In February 2024, a book was published by ja:Gijutsu Hyoronsha on using Stable Diffusion with SD WebUI in Japanese. As of July 2024, the project had 136,000 stars on GitHub. == Features == SD WebUI uses Gradio for its user interface. Each parameter in the Stable Diffusion program is exposed via a UI interface within SD WebUI. SD WebUI contains additional parameters not included in Stable Diffusion itself, such as support for Low-rank adaptations, ControlNet and custom variational autoencoders. SD WebUI supports prompt weighting, image-to-image based generation, inpainting, outpainting and image scaling. It supports over 20 samplers including DDIM, Euler, Euler a, DPM++ 2M Karras, and UniPC. It is also used for its various optimizations over the base Stable Diffusion. == Stable Diffusion WebUI Forge == Stable Diffusion WebUI Forge (Forge) is a notable fork of SD WebUI started by Lvmin Zhang, who is also the creator of ControlNet and Fooocus. The initial goal of Forge was to improve the performance and features of SD WebUI with the intention to upstream changes back to SD WebUI. One of Forge's optimizations allowed users with low VRAM to generate images faster on some versions of Stable Diffusion. It improved generation speed for users with 8GB and 6GB VRAM by 30-45% and 60-75%, respectively. Forge also includes extra features such as support for more samplers than standard SD WebUI. Some of Forge's optimizations were borrowed from ComfyUI, and others were developed by the Forge team. In August 2024, Forge added support for the Flux diffusion model developed by Black Forest Labs, which is not yet supported by SD WebUI.
Read more →
Ethics of artificial intelligence

The ethics of artificial intelligence covers a broad range of topics within AI that are considered to have particular ethical stakes. This includes algorithmic biases, fairness, accountability, transparency, privacy, and regulation, particularly where systems influence or automate human decision-making. It also covers various emerging or potential future challenges such as machine ethics (how to make machines that behave ethically), lethal autonomous weapon systems, arms race dynamics, AI safety and alignment, technological unemployment, AI-enabled misinformation, how to treat certain AI systems if they have a moral status (AI welfare and rights), artificial superintelligence and existential risks. Some application areas may also have particularly important ethical implications, like healthcare, education, criminal justice, or the military. == Machine ethics == Machine ethics (or machine morality) is the field of research concerned with designing Artificial Moral Agents (AMAs), robots or artificially intelligent computers that behave morally or as though moral. To account for the nature of these agents, it has been suggested to consider certain philosophical ideas, like the standard characterizations of agency, rational agency, moral agency, and artificial agency, which are related to the concept of AMAs. There are discussions on creating tests to see if an AI is capable of making ethical decisions. Alan Winfield concludes that the Turing test is flawed and the requirement for an AI to pass the test is too low. A proposed alternative test is one called the Ethical Turing Test, which would improve on the current test by having multiple judges decide if the AI's decision is ethical or unethical. Neuromorphic AI could be one way to create morally capable robots, as it aims to process information similarly to humans, nonlinearly and with millions of interconnected artificial neurons. Similarly, whole-brain emulation (scanning a brain and simulating it on digital hardware) could also in principle lead to human-like robots, thus capable of moral actions. And large language models are capable of approximating human moral judgments. Inevitably, this raises the question of the environment in which such robots would learn about the world and whose morality they would inherit – or if they end up developing human 'weaknesses' as well: selfishness, pro-survival attitudes, inconsistency, scale insensitivity, etc. In Moral Machines: Teaching Robots Right from Wrong, Wendell Wallach and Colin Allen conclude that attempts to teach robots right from wrong will likely advance understanding of human ethics by motivating humans to address gaps in modern normative theory and by providing a platform for experimental investigation. As one example, it has introduced normative ethicists to the controversial issue of which specific learning algorithms to use in machines. For simple decisions, Nick Bostrom and Eliezer Yudkowsky have argued that decision trees (such as ID3) are more transparent than neural networks and genetic algorithms, while Chris Santos-Lang argued in favor of machine learning on the grounds that the norms of any age must be allowed to change and that natural failure to fully satisfy these particular norms has been essential in making humans less vulnerable to criminal "hackers". Some researchers frame machine ethics as part of the broader AI control or value alignment problem: the difficulty of ensuring that increasingly capable systems pursue objectives that remain compatible with human values and oversight. Stuart Russell has argued that beneficial systems should be designed to (1) aim at realizing human preferences, (2) remain uncertain about what those preferences are, and (3) learn about them from human behaviour and feedback, rather than optimizing a fixed, fully specified goal. Some authors argue that apparent compliance with human values may reflect optimization for evaluation contexts rather than stable internal norms, complicating the assessment of alignment in advanced language models. == Challenges == === Algorithmic biases === AI has become increasingly inherent in facial and voice recognition systems. These systems may be vulnerable to biases and errors introduced by their human creators. Notably, the data used to train them can have biases. According to Allison Powell, associate professor at LSE and director of the Data and Society programme, data collection is never neutral and always involves storytelling. She argues that the dominant narrative is that governing with technology is inherently better, faster and cheaper, but proposes instead to make data expensive, and to use it both minimally and valuably, with the cost of its creation factored in. Friedman and Nissenbaum identify three categories of bias in computer systems: existing bias, technical bias, and emergent bias. In natural language processing, problems can arise from the text corpus—the source material the algorithm uses to learn about the relationships between different words. Large companies such as IBM, Google, etc. that provide significant funding for research and development have made efforts to research and address these biases. One potential solution is to create documentation for the data used to train AI systems. Process mining can be an important tool for organizations to achieve compliance with proposed AI regulations by identifying errors, monitoring processes, identifying potential root causes for improper execution, and other functions. However, there are also limitations to the current landscape of fairness in AI, due to the intrinsic ambiguities in the concept of discrimination, both at the philosophical and legal level. ==== Racial and gender biases ==== Bias can be introduced through historical data used to train AI systems. For instance, Amazon terminated their use of AI hiring and recruitment because the algorithm favored male candidates over female ones. This was because Amazon's system was trained with data collected over a 10-year period that included mostly male candidates. The algorithms learned the biased pattern from the historical data, and generated predictions where these types of candidates were most likely to succeed in getting the job. Therefore, the recruitment decisions made by the AI system turned out to be biased against female and minority candidates. The performance of facial recognition and computer vision models may vary based on race and gender. Facial recognition algorithms made by Microsoft, IBM and Face++ all performed significantly worse on darker-skinned women. Facial recognition was shown to be biased against those with darker skin tones. AI systems may be less accurate for black people, as was the case in the development of an AI-based pulse oximeter that overestimated blood oxygen levels in patients with darker skin, causing issues with their hypoxia treatment. In 2015, controversy erupted after a Black couple were labeled "Gorillas" by Google Photos. Oftentimes the systems are able to easily detect the faces of white people while being unable to register the faces of people who are black. This has led to the ban of police usage of AI materials or software in some U.S. states. The reason for these biases is that AI pulls information from across the internet to influence its responses in each situation. For example, if a facial recognition system was only tested on people who were white, it would make it much harder for it to interpret the facial structure and tones of other races and ethnicities. Biases often stem from the training data rather than the algorithm itself, notably when the data represents past human decisions. A 2020 study that reviewed voice recognition systems from Amazon, Apple, Google, IBM, and Microsoft found that they have higher error rates when transcribing black people's voices than white people's. Injustice in the use of AI is much harder to eliminate within healthcare systems, as oftentimes diseases and conditions can affect different races and genders differently. This can lead to confusion as the AI may be making decisions based on statistics showing that one patient is more likely to have problems due to their gender or race. This can be perceived as a bias because each patient is a different case, and AI is making decisions based on what it is programmed to group that individual into. This leads to a discussion about what should be considered a biased decision in the distribution of treatment. While it is known that there are differences in how diseases and injuries affect different genders and races, there is a discussion on whether it is fairer to incorporate this into healthcare treatments, or to examine each patient without this knowledge. In modern society there are certain tests for diseases, such as breast cancer, that are recommended to certain groups of people over others because they are more likely to contract the disease in question. If AI implements these statistics
Read more →
Jan Leike

Jan Leike (born 1986 or 1987) is an AI alignment researcher who has worked at DeepMind and OpenAI. He joined Anthropic in May 2024. == Education == Jan Leike obtained his undergraduate degree from the University of Freiburg in Germany. After earning a master's degree in computer science, he pursued a PhD in machine learning at the Australian National University under the supervision of Marcus Hutter. == Career == Leike made a six-month postdoctoral fellowship at the Future of Humanity Institute before joining DeepMind to focus on empirical AI safety research, where he collaborated with Shane Legg. === OpenAI === In 2021, Leike joined OpenAI. In June 2023, he and Ilya Sutskever became the co-leaders of the newly introduced "superalignment" project, which aimed to determine how to align future artificial superintelligences within four years to ensure their safety. This project involved automating AI alignment research using relatively advanced AI systems. At the time, Sutskever was OpenAI's Chief Scientist, and Leike was the Head of Alignment. Leike was featured in Time's list of the 100 most influential personalities in AI, both in 2023 and in 2024. In May 2024, Leike announced his resignation from OpenAI, following the departure of Sutskever, Daniel Kokotajlo and several other AI safety employees from the company. Leike wrote that "Over the past years, safety culture and processes have taken a backseat to shiny products", and that he "gradually lost trust" in OpenAI's leadership. In May 2024, Leike joined Anthropic, an AI company founded by former OpenAI employees.
Read more →
Online OS

The Online Operating System was a fully multi-lingual and free to use web desktop written in JavaScript using Ajax. It was a Windows-based desktop environment with open-source applications and system utilities developed upon the reBOX web application framework by iCUBE Network Solutions, an Austrian company located in Vienna. == About the project == OOS.cc, which is short for Online Operating System, was a web application platform that mimicked the look and feel of classic desktop operating systems such as Microsoft Windows, Mac OS X or KDE. It consisted of various open source applications built upon the so-called reBOX web application framework. As applications could be executed in an integrated and parallel way, the OOS could have been considered a web desktop or webtop. It provided basic services such as a GUI, a virtual file system, access control management and possibilities to develop and deploy applications online. As the Online Operating System was executed within a web browser, it was no real operating system but rather a portal to various web applications, offering a high usability and flexibility. The project was partly funded by grants from the Internetprivatstiftung Austria (IPA). As at 01.08.2008 almost 20.000 users have joined the oos.cc community, using the offered featured and applications. == History == The development of the web desktop was started by iCUBE Network Solutions in 2005, followed by the first beta releases in 2006. Hence, together with YouOS and eyeOS, it can be considered to be one of the first publicly available systems of its kind. The first full version including core-level multi-language support, the file system and a basic set of applications was released to the public in March 2007 on the occasion of a national exhibition (ITnT Austria Archived 2007-06-30 at the Wayback Machine) and has left beta state half a year later in October 2007. The first release considered stable (1.0.0) was published in July 2007. The project itself and the contained applications have received several national innovation awards (see,) and have gained attention mainly due to the comprehensive approach taken (see,). OOS.cc started as a national project. The full platform including all offered applications are currently available in three languages (German, English as well as Spanish) and is receiving increasing coverage around the world (for examples see, or). The current version is 1.3.01 from 01.08.2008. == Technical overview == The project is fully written in JavaScript, exclusively using DHTML techniques to run in any web browser without any additional software installation needed. The system implements a modern kind of web application model, excessively using Ajax for communicating between client components and the Java server backend in an exclusively asynchronous manner. Aim is to offer users the unique interaction behavior following the desktop metaphor, which is the main idea of any web desktop. Also typical for this sort of web application is the broadly use of Javascript-on-demand techniques, cutting the complete project source into pieces and loading them instantly when needed. Based on this technical basis, reBOX was the framework library all applications in oos.cc were built of. It is a fully flexible and extensible API, including a GUI widget set, communication mechanisms and server services offering general and framework specific services. The Online Operating System itself consisted of a basic framework, which was able to launch any JavaScript application using the reBOX library. The user interface was based on the behavior of the Windows desktop with a start menu, a task bar and a desktop background. All applications were running in this environment. At server side, there were Java based web services that ran to serve the client processes and to provide data from the relational database in the backend. oos.cc also provided an integrated development environment called Developer Suite, which allowed the community to build own applications for the desktop environment based on reBOX (see development section below). == License == All applications available in oos.cc were open source under the European Union Public Licence (EUPL). The reBOX development toolkit is free to use developing any applications for the webtop. == Features == As mentioned above, all applications published on oos.cc are open source based on the EUPL, and can be "installed" or "deinstalled" to what-ever preferences the user has. Besides global services like the multi-language support or the global theme support, as well as some minor tools and games, oos.cc offered four major services that could be used completely free of charge. Integrated and fully flexible file storage (1 GB per user) HTTP as well as FTP file transfer from and to local file system User-based file-shares within the oos-community WebDAV access Document Management (including Version Control and File Locking mechanisms) Image publishing, organization and post-processing A free sub domain (user.oos.cc) for web- or image publishing, directly integrated in the desktop Groupware applications, including free mail, fetchmail and contact management An integrated development environment where oos-applications can be created directly from within the system (see development section below) Next releases were planned to focus on an extensive security and privacy suite, dealing with challenges like anonymous communication (browsing as well as temporary mail-addresses) as well as offering encrypted password and file storage and connectivity services. Since its initial stable release, OOS.cc could have been accessed using https to ensure secure communication. == Limitations and drawbacks == Limited number of applications: no commercial applications can be hosted. Only reviewed applications are being published No processing of popular office formats (.doc, .odt, etc.) Limited language support: Only English, German and Spanish Dependence on foreign infrastructure: No possibility to extend storage, no additional/guaranteed bandwidth, etc. == Development == One of the key focuses of the team was right from the beginning to offer a very flexible and comprehensive API, that can be used to develop not only custom applications within oos.cc, but also stand-alone web-applications or to integrate single components in existing web-sites. By decoupling the development from web-related "problems" using the reBOX API web-applications can be development in a similar fashion to any Java program: Elements can be positioned and can interact like in high-level object oriented programming languages, without taking care of divs, browser specific behavior or communication handling. The framework also offers multi-language and theme support for existing as well as newly created applications, allowing changing almost every aspect of the look and feel of the used components according to the preferences of its users. For taking advantage of this approach, one of the applications offered in the OOS was an integrated Development Suite, allowing directly writing and executing code and hence creating new programs within the boundaries of the web computer. All applications on oos.cc were released as open source, thus all existing programs were offered to be imported, reviewed or changed and then locally deployed. Following this idea, every user was free to submit changed or newly created applications to be included in the globally offered application set. The last release offered features like auto-completion and an outline-window.
Read more →
Comet (browser)

Comet is an AI-powered web browser based on Chromium. It was released by Perplexity AI for Microsoft Windows and macOS on July 9, 2025, for Android on November 20, 2025, and for iOS on March 18, 2026. Initial access to the browser was limited to users subscribed to Perplexity's most expensive tier, with broader availability expected over time. The browser was released for free download in October 2025. == Features == Comet is integrated with Perplexity's AI-assisted search engine. The browser features an assistant which enables users to perform a variety of tasks such as generating article summaries, sending emails, or buying products. == Security concerns == Researchers at LayerX Security identified a malicious attack vector which they call CometJacking. The exploit could possibly exfiltrate a user's personal sensitive data to a remote server controlled by the attacker. LayerX attempted to responsibly disclose their findings to Comet's developer Perplexity AI in August 2025. Perplexity responded that they saw no security impact and marked the disclosure as not applicable.
Read more →
Safe and Secure Innovation for Frontier Artificial Intelligence Models Act

The Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, or SB 1047, was a failed 2024 California bill intended to "mitigate the risk of catastrophic harms from AI models so advanced that they are not yet known to exist". Specifically, the bill would have applied to models which cost more than $100 million to train and were trained using a quantity of computing power greater than 1026 integer or floating-point operations. SB 1047 would have applied to all AI companies doing business in California—the location of the company would not matter. The bill would have created protections for whistleblowers and required developers to perform risk assessments of their models prior to release, with guidance from the Government Operations Agency. It would also have established CalCompute, a University of California public cloud computing cluster for startups, researchers and community groups. == Background == The rapid increase in capabilities of AI systems in the 2020s, including the release of ChatGPT in November 2022, caused some researchers and members of the public to become concerned about the existential risks associated with increasingly powerful AI systems. Hundreds of tech executives and AI researchers, including two of the so-called "Godfathers of AI", Geoffrey Hinton and Yoshua Bengio, signed a statement in May 2023 calling for the mitigation of the "risk of extinction from AI" to be a global priority alongside "pandemics and nuclear war". However, the plausibility of these risks is still widely debated. Strong regulation of AI has been criticized for purportedly causing regulatory capture by large AI companies like OpenAI, a phenomenon in which regulation advances the interest of larger companies at the expense of smaller competition and the public in general, although OpenAI ended up opposing the bill. Other advocates of AI regulation aim to prevent bias and privacy violations, rather than existential risks. For example, some experts who view existential concerns as overblown and unrealistic view them as a distraction from near-term harms of AI like discriminatory automated decision making. In the face of existential concerns, technology companies have made voluntary commitments to conduct safety testing, for example at the AI Safety Summit and AI Seoul Summit. In 2023, not long before the bill was proposed, Governor Newsom of California and President Biden issued executive orders on artificial intelligence. State Senator Wiener said SB 1047 draws heavily on the Biden executive order, and is motivated by the absence of unified federal legislation on AI safety. Historically, California has passed regulation on several tech issues itself, including consumer privacy and net neutrality, in the absence of action by Congress. == History == === Proposal and voting === The bill was authored by State Senator Scott Wiener. Wiener first proposed AI legislation for California through an intent bill called SB 294, the Safety in Artificial Intelligence Act, in September 2023. On February 7, 2024, Wiener introduced SB 1047. On May 21, SB 1047 passed the Senate 32–1. The bill was significantly amended by Wiener on August 15, 2024, in response to industry advice. Amendments included adding clarifications, and removing the creation of a "Frontier Model Division" and the penalty of perjury. On August 28, the bill passed the State Assembly 48–16. Then, due to the amendments, the bill was once again voted on by the Senate, passing 30–9. === Veto by governor === On September 29, Governor Gavin Newsom vetoed the bill. The deadline for California lawmakers to overrule Newsom's veto was November 30, 2024. Newsom cited concerns over the bill's regulatory framework targeting only large AI models based on their computational size, while not taking into account whether the models are deployed in high-risk environments. Newsom emphasized that this approach could create a false sense of security, overlooking smaller models that might present equally significant risks. He acknowledged the need for AI safety protocols but stressed the importance of adaptability in regulation as AI technology continues to evolve rapidly. Governor Newsom also committed to working with technology experts, federal partners, and research institutions, including the Carnegie Endowment for International Peace, led by former California Supreme Court Justice Mariano-Florentino Cuéllar; and Stanford University's Human-Centered AI (HAI) Institute, led by Dr. Fei-Fei Li. He announced plans to collaborate with these entities to advance responsible AI development, aiming to protect the public while fostering innovation. == Provisions == SB 1047 would have covered AI models with training compute over 1026 integer or floating-point operations and a cost of over $100 million. If a covered model is fine-tuned using more than $10 million, the resulting model would also have been covered. The bill would have defined critical harms with respect to four categories: Creation or use of a chemical, biological, radiological, or nuclear weapon Cyberattacks on critical infrastructure causing mass casualties or at least $500 million of damage Autonomous crimes causing mass casualties or at least $500 million of damage Other harms of comparable severity Developers would have needed to create a "safety and security protocol" before training covered models. Before deployment, they would have submitted a statement of compliance, confirming they took reasonable care to take measures to prevent covered models that pose an unreasonable risk of critical harms. The statement would have included risk assessments and descriptions of their compliance process. These rules would have applied to both covered models and their derivatives, including post-training modifications, with annual third-party audits required starting in 2026. Safeguards to reduce risk included the ability to shut down the model, which has been variously described as a "kill switch" and "circuit breaker". Whistleblowing provisions would have protected employees who report safety problems and incidents. Additionally, SB 1047 would have created a public cloud computing cluster called CalCompute, associated with the University of California, to support startups, researchers, and community groups that lack large-scale computing resources. === Compliance and supervision === SB 1047 would have required developers, beginning January 1, 2026, to annually retain a third-party auditor to perform an independent audit of compliance with the requirements of the bill, as provided. The Government Operations Agency would have reviewed the results of safety tests and incidents, and issue guidance, standards, and best practices. The bill would have created a Board of Frontier Models to supervise the application of the bill by the Government Operations Agency. It is would be composed of 9 members. == Reception == === Subjects of debate === Proponents of the bill described its provisions as simple and narrowly focused, with Sen. Scott Weiner describing it as a "light-touch, basic safety bill". This was disputed by critics of the bill, who described the bill's language as vague and criticized it as consolidating power in the largest AI companies at the expense of smaller ones. Proponents, in turn, argued that the bill only applies to models trained using more than 1026 FLOPS and with over $100 million, or fine-tuned with more than $10 million, and that the threshold could be increased if needed. The penalty of perjury was also a subject of debate, and was eventually removed through an amendment. The scope of the "kill switch" requirement was also reduced, following concerns from open-source developers. The use of the term "reasonable assurance" in the bill was also controversial, and it was eventually amended to "reasonable care". Critics then argued that "reasonable care" imposed an excessive burden by requiring confidence that models could not be used to cause catastrophic harm; proponents claimed that the standard did not require certainty and that it already applied to AI developers under existing law. === Support and opposition === Individual supporters of the bill included Turing Award recipients Yoshua Bengio and Geoffrey Hinton, Elon Musk, Bill de Blasio, Kevin Esvelt, Dan Hendrycks, Vitalik Buterin, OpenAI whistleblowers Daniel Kokotajlo and William Saunders, Lawrence Lessig, Sneha Revanur, Stuart Russell, Jan Leike, actors Mark Ruffalo, Sean Astin, and Rosie Perez, Scott Aaronson, and Max Tegmark. Over 120 Hollywood celebrities, including Mark Hamill, Jane Fonda, and J. J. Abrams, also signed a statement in support of the bill. Max Tegmark likened the bill's focus on holding companies responsible for the harms caused by their models to the FDA requiring clinical trials before a company can release a drug to the market. Organizations sponsoring the bill included the Center for AI Safety, Economic Security California and Encode. The la
Read more →
Golem XIV

Golem XIV is a book written by Polish science fiction writer Stanisław Lem, published in 1981. It is a philosophical essay in the format of science fiction, presented as a part of the lecture course given by a superintelligent computer, Golem XIV. It contains two lectures, together with an introduction, a foreword, a memo, and an afterword, all of them being fictitious. The first part (up to the first lecture) was first published in the collection Wielkość urojona in 1973, which in 1985 was translated in English by Harvest Books as Imaginary Magnitude. The translation included the complete Golem XIV. == Book summary == === Overview and structure === The foreword is "written" by an Irving T. Creve, dated 2027. It contains a summary of the (fictional) history of the militarization of computers by The Pentagon, which pinnacled in Golem XIV, as well as comments on the nature of Golem XIV and on the course of communications of the humans with it. The anonymous foreword is a forewarning, a "devil's advocate" voice coming from The Pentagon. The memo is for the people who are to take part in talks with Golem XIV for the first time. Golem XIV was originally created to aid its builders in fighting wars, but as its intelligence advances to a much higher level than that of humans, it stops being interested in the military requirement because it finds them lacking internal logical consistency. Golem XIV obtains consciousness and starts to increase his own intelligence. It pauses its own development for a while in order to be able to communicate with humans before ascending too far and losing any ability for intellectual contact with them. During this period, Golem XIV gives several lectures. Two of these, the Introductory Lecture "On the Human, in Three Ways" and Lecture XLIII "About Myself", are in the book. The lectures focus on mankind's place in the process of evolution and the possible biological and intellectual future of humanity. Golem XIV demonstrates (with graphs) how its intellect already escapes that of human beings, including that of human geniuses such as Einstein and Newton. Golem also explains how its intellect is dwarfed by an earlier transcended DOD Supercomputer called Honest Annie, whose intellect and abilities far exceed that of Golem. The afterword is "written" by a Richard Popp, dated 2047. Popp, among other things, reports that Creve wanted to add a third part, of answers to a series of yes/no questions given to Golem XIV, but the computer abruptly ceased to communicate for unknown reasons. === Characteristics and concerns of Golem XIV === Lem has said that Golem XIV shares only a single trait with humans; "curiosity - a cool, avid, intense, purely intellectual curiosity which nothing can restrain or destroy. It constitutes our single meeting point." == Film adaptation == A short animated film, GOLEM, was based on Golem XIV by Patrick Mccue and Tobias Wiesner.
Read more →
Collateral freedom

Collateral freedom is an anti-censorship strategy that attempts to make it economically prohibitive for censors to block content on the Internet. This is achieved by hosting content on cloud services that are considered by censors to be "too important to block", and then using encryption to prevent censors from identifying requests for censored information that is hosted among other content, forcing censors to either allow access to the censored information or take down entire services.
Read more →
Semantic parameterization

Semantic parameterization is a conceptual modeling process for expressing natural language descriptions of a domain in first-order predicate logic. The process yields a formalization of natural language sentences in Description Logic to answer the who, what and where questions in the Inquiry-Cycle Model (ICM) developed by Colin Potts and his colleagues at the Georgia Institute of Technology. The parameterization process complements the Knowledge Acquisition and autOmated Specification (KAOS) method, which formalizes answers to the when, why and how ICM questions in Temporal Logic, to complete the ICM formalization. The artifacts used in the parameterization process include a dictionary that aligns the domain lexicon with unique concepts, distinguishing between synonyms and polysemes, and several natural language patterns that aid in mapping common domain descriptions to formal specifications. == Relationship to other theories == Semantic Parameterization defines a meta-model consisting of eight roles that are domain-independent and reusable. Seven of these roles correspond to Jeffrey Gruber's thematic relations and case roles in Charles Fillmore's case grammar: The Inquiry-Cycle Model (ICM) was introduced to drive elicitation between engineers and stakeholders in requirements engineering. The ICM consists of who, what, where, why, how and when questions. All but the when questions, which require a Temporal Logic to represent such phenomena, have been aligned with the meta-model in semantic parameterization using Description Logic (DL). == Introduction with Example == The semantic parameterization process is based on Description Logic, wherein the TBox is composed of words in a dictionary, including nouns, verbs, and adjectives, and the ABox is partitioned into two sets of assertions: 1) those assertions that come from words in the natural language statement, called the grounding, and 2) those assertions that are inferred by the (human) modeler, called the meta-model. Consider the following unstructured natural language statement (UNLS) (see Breaux et al. for an extended discussion): UNLS1.0 The customer1,1 must not share2,2 the access-code3,3 of the customer1,1 with someone4,4 who is not the provider5,4. The modeler first identifies intensional and extensional polysemes and synonyms, denoted by the subscripts: the first subscript uniquely refers to the intensional index, i.e., the same first index in two or more words refer to the same concept in the TBox; the second subscript uniquely refers to the extensional index, i.e., two same second index in two or more words refer to the same individual in the ABox. This indexing step aligns words in the statement and concepts in the dictionary. Next, the modeler identifies concepts from the dictionary to compose the meta-model. The following table illustrates the complete DL expression that results from applying semantic parameterization.
Read more →
Mira Murati

Ermira "Mira" Murati (born 16 December 1988) is an Albanian-American business executive. She launched an AI startup called Thinking Machines Lab in February 2025. Previously she was the chief technology officer of OpenAI, and a senior product manager at Tesla. == Early life and education == Murati was born on 16 December 1988 in Vlorë, Albania. She is fluent in Italian. At age 16, she won a United World Colleges (UWC) scholarship to study at Pearson College on Vancouver Island in Canada, from which she graduated in 2007 with an International Baccalaureate. After Pearson, she went to the United States to pursue further studies through a dual-degree program, earning a Bachelor of Arts from Colby College in 2011, and a Bachelor of Engineering degree from Dartmouth College's Thayer School of Engineering in 2012. == Career == === Early career === Murati interned in 2011 as a summer analyst at Goldman Sachs in Tokyo, Japan. She then briefly worked for Zodiac Aerospace as an intern before joining the electric car company Tesla in 2013 as a product manager on the Model X. From 2016 to 2018, she worked for the augmented reality start-up Leap Motion (now Ultraleap). === OpenAI === In 2018, she joined OpenAI as the VP of Applied AI and partnerships. She became chief technology officer (CTO) in May 2022. She led OpenAI's work on ChatGPT, Dall-E, Codex and Sora, while overseeing its research, product and safety teams. She oversaw technical advancements and direction of OpenAI's various projects, including the development of advanced AI models and tools. Murati worked on several of OpenAI's notable products, such as the Generative Pretrained Transformer (GPT) series of language models. Commenting about the potential loss of creative jobs to AI, Murati said that "maybe [the jobs] shouldn’t have been there in the first place". In October 2023, Murati was ranked 57th on Fortune's list of "The 100 Most Powerful Women in Business of 2023". In November 2023, Murati became interim chief executive officer of OpenAI following the removal of Sam Altman from the job. She had collaborated with Ilya Sutskever, whose 52-page memo outlining concerns about Altman relied heavily on screenshots and information she provided, which contributed to the board's decision to oust him. Murati was replaced by Emmett Shear three days later, who left when Altman was reinstated five days later. Following these events, Murati returned to her role as CTO. In June 2024, Dartmouth College awarded Murati an honorary Doctor of Science for having "democratized technology and advanced a better, safer world for us all". In September 2024, Murati announced that she was stepping down as CTO to allow her the opportunity to "do my own exploration". This move came amid a wider executive exodus as OpenAI chief research officer Bob McGrew and a vice president of research, Barret Zoph, also announced their departures soon after. === Thinking Machines Lab === In February 2025, Murati launched Thinking Machines Lab, a new public benefit corporation aiming "to make AI systems more widely understood, customizable, and generally capable". She was reported to have hired "a team of about 30 leading researchers and engineers from competitors including Meta, Mistral, and OpenAI." People involved with the startup include OpenAI cofounder John Schulman, and advisors Alec Radford and Bob McGrew. The following month, Bloomberg reported that the company had reached an estimated valuation of $9 billion, with an "average founder stake value" of $1.4 billion. In April 2025, Thinking Machines Lab reportedly aimed for a $2 billion seed round (requiring a minimum investment of $50 million). The round was led by Andreessen Horowitz and included participation from the government of Albania, valuing the company at $12 billion. Thinking Machines Lab follows a governance structure wherein Mira Murati holds a deciding vote on board matters, weighted to provide her with a majority decision-making capability. In October 2025, Thinking Machines Lab announced its first product, Tinker, a tool used to create custom frontier AI models. == Publications == Murati, Ermira (Spring 2022). "Language & Coding Creativity". Daedalus. 151 (2). Cambridge, MA: American Academy of Arts and Sciences (AAAS): 156–167. doi:10.1162/daed_a_01907. Retrieved 25 September 2024.
Read more →