The expression real intelligence brings together two fundamental ideas: the nature of reality and the capacity to understand it. The first concerns what actually exists, what an object truly is and what can be established about its characteristics. The second concerns the ability to interpret evidence, distinguish significance from appearance and arrive at conclusions that are justified by what is known. Their relationship is particularly important wherever decisions depend upon the accurate understanding of things whose identity, condition, history or value cannot be established through superficial examination alone.
The word real entered English in the early fourteenth century, initially carrying meanings associated with actual existence and, by the fifteenth century, with things themselves, particularly in legal distinctions concerning property. It derives through Old French from the Late Latin realis, meaning actual or belonging to the thing itself, ultimately related to the Latin res, meaning a thing, matter, affair or property. Its modern meanings include that which actually exists, that which is genuine and that which corresponds to the facts rather than merely to appearance or representation.¹
The intellectual significance of the word lies in its distinction between a thing and the descriptions, beliefs or representations through which that thing is understood. An object may be described inaccurately. A record may contain an error. An apparent characteristic may prove misleading when examined more closely. Reality does not necessarily conform to the account made of it.
The word intelligence has a complementary history. It entered English in the late fourteenth century through Old French intelligence and Latin intelligentia, meaning understanding, knowledge and the power of discernment. The underlying Latin verb, intelligere, means to understand, comprehend or come to know. The concept consequently concerns more than possession of information. It involves the capacity to determine what information means, how it relates to other evidence and what conclusions can reasonably be drawn from it.
The combination real intelligence has no single, universally acknowledged inventor or definitive historical point of origin as a separate scientific discipline. Its meaning must be understood through the wider histories of language, philosophy, epistemology and the development of artificial intelligence. The phrase may refer to intelligence that is genuine rather than simulated, to an understanding grounded in actual circumstances, or to the practical distinction between a credible representation and the reality it purports to describe.
The development of artificial intelligence has given the first of these questions particular prominence. In his 1950 paper Computing Machinery and Intelligence, Alan Turing examined the question of whether machines could think and proposed replacing an abstract dispute over definitions with a more operational test based upon observable performance. His work established an influential starting point for the subsequent discussion of machine intelligence, while leaving open the wider philosophical question of what would constitute genuine understanding.
That question remains important. In Artificial Intelligence: A Very Short Introduction, published by Oxford University Press in 2018, Margaret A. Boden examined whether a future artificial general intelligence matching human performance would necessarily possess real intelligence, understanding or creativity. She considered the distinctions between outward performance, cognition, consciousness and the philosophical interpretation of intelligent behaviour. Her discussion illustrates that the expression real intelligence belongs to a continuing debate rather than a concept whose meaning has been conclusively settled.
The application developed by X from GENERAL INTELLIGENCE PLC to fine art and specie insurance providers has a distinct practical emphasis. It is concerned with the relationship between an object as it actually exists and the information through which an insurer understands, values and protects it. The question is not only whether an analytical system is genuinely intelligent, but whether the institution employing it can establish a sufficiently accurate understanding of the real property at the centre of its insurance decision.
From Representation to Reality
Insurance depends upon representations. An application describes the property to be insured. A schedule records its identity and declared value. A valuation expresses an opinion about its financial worth. Photographs and condition reports document its appearance. Ownership records, provenance documents and security assessments contribute further information. Each representation has a purpose, but no single one necessarily captures the complete reality of the object.
This distinction is especially important in fine art and specie insurance. A painting, sculpture, manuscript, rare watch, gemstone, precious-metal holding or other valuable object may possess characteristics that are incompletely described, difficult to verify or subject to specialist interpretation. Its financial significance may depend upon factors that cannot be established from an inventory entry alone.
A painting's value may be affected by attribution, provenance, condition, rarity, historical significance and the standing of its artist. A manuscript may derive value from its authorship, age, completeness and historical context. A gemstone may require specialist examination to establish its identity and characteristics. A watch may depend upon its reference, originality, components, service history and condition. Precious metals may be valued principally by weight, fineness and market price, while additional considerations may arise from their form, custody and commercial purpose.
These differences matter because an insurance representation can be internally consistent while remaining incomplete. A valuation may be accurately recorded but based upon an assumption about attribution that has not been adequately substantiated. A photograph may correctly show an object without establishing the originality of its components. An inventory may identify an item while failing to record a material change in condition or location.
REAL INTELLIGENCE® is founded upon the proposition that an insurer should seek to understand the reality behind these representations. Artificial intelligence can assist by connecting documentation, visual records, specialist assessments and other legitimate evidence, identifying discrepancies and directing attention towards the questions that matter.
The objective is not to assume that every uncertainty can be eliminated. It is to distinguish established facts from claims requiring verification, identify the limitations of available evidence and improve the quality of the judgements through which the insurer determines its exposure.
Fine Art and Specie as Distinct Forms of Value
Fine art and specie insurance share an emphasis upon valuable movable property, but the classes encompass different kinds of assets and different sources of value. Fine art insurance may concern paintings, sculptures, drawings, antiques, rare books, manuscripts and other cultural or collectible objects. Specie insurance may concern precious metals, jewellery, gemstones, rare coins, cash and other high-value physical commodities. The precise classification of a risk depends upon the property, the coverage sought and the applicable insurance arrangements.
Fine art may be unique in a material and historical sense. Its value can depend upon attribution, provenance, cultural significance, condition, rarity and the market in which it is exchanged. Even where comparable works exist, their histories, states of preservation and market positions may differ sufficiently to make direct comparison difficult.
Specie often presents a different valuation problem. Some forms of precious metal or other commodities have established market prices, allowing value to be determined substantially through quantity, purity and prevailing market conditions. Other items, including jewellery, rare coins and certain collectible watches, may possess additional qualities that make their value less directly comparable with a commodity benchmark.
These distinctions influence the intelligence required. A painting may require extensive research into provenance and attribution. A bullion holding may depend more directly upon verified quantity, fineness, custody and market price. A rare watch may combine technical identification with specialist knowledge of condition and collectability. The insurer cannot assume that the same valuation method, evidential threshold or analytical process will be appropriate to every asset.
REAL INTELLIGENCE® therefore treats the nature of the insured object as the starting point for determining what must be known. Its purpose is to connect the relevant evidence to the specific questions of identity, value, condition, custody and exposure that the particular class of property presents.
The Object as Intelligence
The physical object is the centre of the insurance relationship, but the information surrounding it is often distributed across different documents, systems and professional assessments. An item may have photographs, valuations, invoices, certificates, ownership records, condition reports, restoration records and movement documentation, each created at a different time and for a different purpose.
The challenge is to determine whether these sources collectively provide a sufficiently accurate representation of the object. The same item may be described differently by a dealer, valuer, conservator, auction house and insurer. Some differences may be entirely legitimate; others may indicate that information is missing, outdated or inconsistent.
Artificial intelligence can help compare records, extract identifying characteristics, organise documents and identify discrepancies that warrant examination. It can assist in connecting images with written descriptions, linking valuations to their supporting evidence and locating relevant historical information when an object is reviewed or a claim occurs.
The technology must not mistake a discrepancy for proof of wrongdoing. A description may have been amended legitimately, a photograph may depict an object under different lighting conditions, or an older record may omit details that subsequently became relevant. The purpose of analysis is to identify the need for review, not to impose an automatic conclusion.
The intelligent insurer consequently treats the object as a continuing subject of knowledge. Its records should be sufficiently coherent to establish what is insured, what is known about it and where significant uncertainties remain. The object is not reduced to a digital profile; rather, the profile is continually tested against the evidence of the object itself.
Authenticity, Attribution and Identity
The distinction between appearance and reality is particularly consequential where authenticity or attribution materially affects value. A work may be attributed to a particular artist, a manuscript to a particular author, a watch to a particular reference or a gemstone to a particular identity. Each claim requires evidence appropriate to the nature of the object.
In fine art, questions of authorship and attribution may involve documentary research, examination of materials and technique, comparison with other works, conservation analysis and specialist opinion. An attribution can remain uncertain even where extensive research has been undertaken. The insurer must therefore understand whether a description represents an established conclusion, a qualified opinion or an unresolved possibility.
Artificial intelligence can assist by organising comparative images, examining documentary records, identifying inconsistencies and retrieving relevant information. Computer-vision techniques may help identify visual features that deserve specialist consideration, while language-processing tools can extract assertions from catalogues, reports and correspondence.
These capabilities do not establish authorship or authenticity by themselves. Visual resemblance is not proof that a work was created by a particular artist and an apparently consistent historical record may repeat an earlier assertion without independently verifying it. The evidential basis for a conclusion must remain clear.
REAL INTELLIGENCE® treats identity as an aspect of risk that requires disciplined verification. Its objective is to improve the breadth and accessibility of evidence available to specialists while ensuring that the conclusions drawn from that evidence remain proportionate to what can actually be established.
Provenance as a Structure of Evidence
Provenance is more than a narrative about an object's past. It is a body of evidence concerning its creation, ownership, acquisition, location and history. Depending upon the object, provenance may influence attribution, legal status, market value, cultural significance and the circumstances in which it can legitimately be traded.
The reliability of a provenance account depends upon the quality and continuity of its supporting evidence. Invoices, catalogues, correspondence, estate records, exhibition histories, ownership documents and other materials may establish different parts of the history. Some links may be well documented; others may remain incomplete or dependent upon an earlier source.
Artificial intelligence can assist in organising these records, comparing dates and descriptions, connecting references and identifying contradictions that require further investigation. It may help establish whether several records plausibly relate to the same object and reveal gaps that would otherwise be difficult to identify across a large documentary collection.
A digital record does not automatically establish the truth of the information it contains. Repeated assertions may originate from the same source, a transcription may introduce an error and an apparently continuous ownership history may omit relevant circumstances. The system must preserve the distinction between a documented fact, an assertion made by a source and an inference drawn from several records.
For the insurer, provenance intelligence can inform the understanding of identity, value and risk. It can also assist when an object is transferred, exhibited, sold, restored or presented in support of a claim. The objective is to make the evidence surrounding its history more coherent, without pretending that every historical uncertainty can be resolved.
Condition and the Reality of Change
Condition is a defining element of the risk associated with fine art and other sensitive objects. Physical condition may change through age, environmental exposure, handling, transportation, restoration, accidental damage or other events. The significance of a change depends upon the object and the characteristics that determine its value and preservation.
A condition report provides a record against which subsequent observations can be compared. Photographs, annotations, descriptions and specialist assessments may establish the state of an object at a particular time, helping later examiners distinguish pre-existing deterioration from newly observed damage. The American Museum of Natural History's conservation guidance emphasises the importance of documenting objects before and after loans or exhibitions, including the use of annotated photographs to distinguish new changes from prior damage or deterioration.
Artificial intelligence can assist in comparing successive condition records, identifying potential visual differences and directing attention towards areas requiring specialist examination. It may help preserve a more continuous account of how the object has been documented over time, provided that the records are sufficiently comparable and the limits of image-based analysis are understood.
The presence of an apparent difference does not establish that new damage has occurred. Lighting, camera angle, image resolution, photographic processing and other factors can affect the result. Equally, a photographic comparison may fail to reveal structural or material changes that require physical inspection or specialist testing.
REAL INTELLIGENCE® therefore treats condition as a time-dependent aspect of the insured object. Its purpose is to improve the continuity and accessibility of evidence, enabling conservators, surveyors, underwriters and claims professionals to evaluate changes with greater clarity while retaining responsibility for the interpretation.
Valuation and the Limits of Quantification
Valuation is often the most visible expression of an object's insured significance, but the figure recorded in a policy schedule is the conclusion of an assessment, not the complete explanation of value.
For fine art, value may depend upon attribution, provenance, rarity, condition, the reputation of the artist, historical significance, comparable transactions and the circumstances of the market. The value of specie may depend more directly upon weight, fineness and market prices, although collectible objects can introduce additional dimensions. Jewellery and watches may combine material worth with workmanship, brand, rarity, originality and demand.
Different valuation purposes should not be conflated. A retail replacement figure, an auction result, an agreed value and a market estimate may represent different concepts. The appropriate basis depends upon the terms of the policy and the purpose for which the valuation is required. A past valuation may remain a useful reference while no longer adequately reflecting the current circumstances.
Artificial intelligence can assist in examining supporting evidence, comparing relevant transactions, identifying changes in market information and flagging valuations that warrant review. It may also help identify where two ostensibly comparable objects differ materially in condition, attribution or provenance.
These methods cannot eliminate the need for professional valuation. Market information may be incomplete, transactions may be atypical and unique works may have few meaningful comparators. An apparently precise model can conceal the uncertainty inherent in the assumptions used to produce it.
REAL INTELLIGENCE® therefore treats valuation intelligence as an examination of the relationship between the object, the evidence and the stated value. The question is not merely whether the number appears plausible, but whether the basis of the valuation is understood, whether it corresponds to the relevant insurance purpose and whether the available evidence supports the conclusion.
The Collection as a Connected System
A major art collection or portfolio of valuable movable property is more than a list of separate items. Relationships between objects can influence their cultural, historical and financial significance. Works may form a coherent collection through their artist, period, provenance, subject or curatorial purpose. A collection may also create concentrations of value, shared security requirements or common exposure to a single event.
Artificial intelligence can help connect information across the collection, examining relationships between objects, ownership, location, value, condition, provenance and movement. It may assist in identifying whether several items share an important characteristic or whether the concentration of property in a particular location creates a material accumulation of risk.
The distinction between individual value and collective exposure is important. Each object may be insured separately, yet a single event could affect several works held together. Conversely, the importance of an individual work may derive partly from its relationship to the collection as a whole, even if the practical insurance treatment remains governed by its own terms.
The insurer must also distinguish the value of the collection to its owner from the sum of the values assigned to its constituent objects. Cultural, historical or personal significance may not translate directly into an agreed financial amount, although it can influence the owner's requirements and the specialist handling of the property.
REAL INTELLIGENCE® treats the collection as a system whose relationships are relevant to underwriting, accumulation management and loss prevention. Its objective is to connect information at object level with a broader understanding of how the insured property is assembled, situated and exposed.
Movement, Custody and Location
Fine art and specie may pass through a complex sequence of locations and custodians. Objects may be held at private residences, galleries, museums, dealers, auction houses, secure vaults, storage facilities, conservation studios or exhibition venues. Depending upon the policy and the arrangements made, they may also be transported between locations or internationally.
Movement changes the circumstances in which property is held and may affect its vulnerability to loss or damage. Custody arrangements, packaging, handling, environmental control, transport methods and security procedures can all be relevant. The precise coverage provided during movement, storage or exhibition depends upon the policy wording and the insurance arrangements in place. Lloyd's guidance on fine art, specie and exhibitions recognises that the nature of the cover, the location of the property and the circumstances of the insured are important when determining the applicable risk location.
Artificial intelligence can assist in organising records of location, custody and scheduled movement, identifying inconsistencies and maintaining a more coherent account of where objects are expected to be. It may help connect inventory data with transport documentation, exhibition records, handover information and other legitimate sources.
The objective is not unrestricted tracking or the assumption that every movement presents a problem. It is to ensure that material information about the object's circumstances is accessible to those responsible for the relevant insurance, custody and risk-management decisions.
REAL INTELLIGENCE® treats movement as part of the object's history of exposure. A sound understanding of the insured property must include not only what it is, but where it has been, the conditions under which it was held and the arrangements governing its movement.
Security as an Intelligence Discipline
The security of fine art and specie depends upon a system of measures rather than a single device. Premises, vaults, alarms, access controls, surveillance, personnel, transportation, custody arrangements and operating procedures can interact to determine how valuable property is protected.
The effectiveness of the overall arrangement depends upon its suitability for the objects and circumstances involved. A system designed for secure storage may not provide the same protection during an exhibition or transport. A change in personnel, premises or operating arrangements may alter the relevance of an existing assessment.
Artificial intelligence can assist in organising security assessments, reviewing incident records, comparing information across locations and identifying changes that warrant specialist consideration. It may help connect documented security arrangements with the particular objects and movements to which they apply.
The technology should not be mistaken for a security measure in itself. A system that identifies an apparent weakness does not establish how serious it is and an absence of recorded incidents does not prove that the arrangements are adequate. Physical protection continues to depend upon appropriate professional assessment, tested procedures and responsible operation.
The objective of REAL INTELLIGENCE® is to help insurers understand the relationship between the property, its location, its custody and its security arrangements. Intelligence is valuable when it informs an appropriate assessment of exposure and helps ensure that the protection corresponds with the actual circumstances.
Claims as Forensic Intelligence
A fine art or specie claim can raise questions that extend well beyond the immediate financial amount. The insurer may need to establish the identity of the object, the ownership or other relevant interest, its condition before the event, the circumstances of loss or damage, the value at the relevant time and the provisions of the policy that determine the response.
These questions may require a combination of documentary evidence, photographs, specialist reports, transaction records, valuations, condition assessments and information about location or custody. A weakness in one part of the evidence can affect the interpretation of another. An asserted value may depend upon attribution; an allegation of new damage may require comparison with earlier condition records; a missing object may require a more precise reconstruction of its identity and movements.
Artificial intelligence can assist by organising documentation, extracting relevant information, comparing records and identifying discrepancies that require human investigation. It may help establish a clearer chronology, connect evidence across different sources and make relevant prior assessments easier to retrieve.
The limitations remain important. An apparent inconsistency may have an innocent explanation, a photograph may not provide adequate evidence of condition and a model cannot determine the legal effect of the policy or the credibility of every assertion. These matters require appropriate professional assessment.
REAL INTELLIGENCE® treats the claim as an opportunity to strengthen the institution's understanding of the object and the circumstances of loss. The objective is not to automate forensic judgement, but to make the investigative process more coherent, preserve the distinction between evidence and conclusion and ensure that the lessons of the claim can inform subsequent decisions.
Portfolio Intelligence and Accumulation
For an insurer, individual objects form part of a wider portfolio of property, collections, locations and commercial relationships. Exposures may be concentrated by artist, collection, geographical area, exhibition, storage location, transport arrangement, object category or other shared characteristics.
An apparently diversified portfolio may consequently contain common sources of loss. Several collections may be held in the same building; multiple insureds may rely upon the same transport or storage arrangements; or a single event may affect different categories of property in a shared location. These relationships are relevant to accumulation even where the objects themselves have different values and characteristics.
Artificial intelligence can assist in analysing these relationships, comparing insured locations and object categories and identifying patterns that warrant further examination. It may help management consider how the portfolio would be affected by events involving multiple objects or shared facilities.
Such analysis should distinguish a possible concentration from an established one. The significance of a shared location, transport arrangement or other dependency depends upon the actual circumstances and the policy terms involved. Data may be incomplete and a simple classification may fail to reflect the protections or separation arrangements in place.
The objective is to understand how risk is distributed and connected across the portfolio. REAL INTELLIGENCE® therefore extends the analysis beyond the declared value of individual objects towards a more complete understanding of the circumstances in which multiple losses could arise.
The Market as an Intelligence Environment
The markets for fine art and specie are varied and dynamic. Auction results, private transactions, exhibitions, collecting patterns, institutional recognition, movements in wealth, scarcity and changing preferences may influence the value and desirability of particular objects.
These forces do not operate uniformly. A market movement affecting one artist, period or category of watch may have little significance for another. A single exceptional transaction may not provide a reliable indication of the wider market, while a change in attribution or provenance may affect an individual object in ways that cannot be inferred from general price trends.
Artificial intelligence can assist in organising large collections of market information, comparing relevant transactions and identifying developments that may warrant expert attention. It can also help distinguish market evidence from asking prices, promotional descriptions and other information that does not establish a completed transaction.
The purpose is not to turn insurance into market speculation or to assume that every change should result in an immediate adjustment to an insured value. It is to improve the insurer's understanding of the economic environment in which the object's value is assessed.
REAL INTELLIGENCE® recognises that the physical object and the market in which it is valued are related but different realities. The insurer must understand the object itself, the basis of its valuation and the limits of the evidence available concerning market value.
The Specialist and the Machine
Fine art and specie insurance remains profoundly dependent upon specialist knowledge. Art historians, conservators, valuers, gemologists, watch specialists, security professionals, underwriters, claims experts and brokers each contribute a distinct perspective. Their knowledge may be developed over decades and can depend upon experience that is difficult to represent fully in a dataset.
Artificial intelligence should therefore be understood as an augmentation of specialist intelligence. It can retrieve information, compare documents, identify visual differences, organise histories and connect evidence across sources at a scale that would otherwise require substantial manual effort. These capabilities can give experts a wider field of examination and allow them to concentrate more of their attention upon the questions requiring judgement.
Yet the system must not presume to replace the expertise that gives the evidence its meaning. A visual similarity does not establish attribution; a market comparison does not establish an object's value; and a difference between condition images does not necessarily establish new damage. These conclusions depend upon the object, the evidence and the context in which the assessment is undertaken.
The relationship must therefore remain open to verification and challenge. Specialists should be able to understand the evidence underlying material findings, assess the limitations of a system and determine whether further examination is required.
REAL INTELLIGENCE® seeks to strengthen this relationship between evidence and expertise. The objective is not to create an artificial substitute for human knowledge, but to enable specialists to understand more, examine more systematically and make better-informed decisions.
Institutional Memory
A specialist insurer accumulates an important body of knowledge through underwriting decisions, valuations, claims, conservation issues, security assessments and the experience of handling unusual or exceptionally valuable objects. This experience can be difficult to preserve because it may remain distributed across files, reports, systems and individual professionals.
Artificial intelligence can help make appropriately recorded knowledge more accessible and easier to compare. Historical claims, previous valuations, condition reports, security findings and documented decisions can become part of an institutional intelligence resource. The insurer can retrieve relevant experience, examine how earlier assumptions performed and identify lessons that may apply to a new circumstance.
Institutional memory must not become an unquestioned repository of precedent. An earlier valuation may no longer reflect market conditions; a previous claim may have involved materially different facts; and a historical interpretation may have been revised by subsequent research. The value of memory is that it enables the organisation to examine experience rather than repeatedly rediscover it.
The objective is to make the insurer more capable through accumulated understanding. Knowledge should survive changes in personnel, remain open to correction and support decisions without obscuring the limitations of historical evidence.
An intelligent institution should become better equipped to understand each new object and each new claim because the lessons of its previous experience remain available.
Governance and the Integrity of Evidence
The use of artificial intelligence in fine art and specie insurance requires appropriate governance. The information involved can be sensitive, commercially significant or dependent upon specialist opinion. Incorrect conclusions about identity, provenance, condition or value may have substantial financial consequences.
The insurer must understand the quality and origin of its data, the limitations of analytical systems and the basis on which material conclusions are reached. Important records should remain traceable to their sources and AI-generated findings should be distinguished from verified facts and professional opinions.
The appropriate controls depend upon the use. A system that organises documents presents different risks from one used to support a consequential valuation or claims decision. Where a system contributes materially to a decision, its outputs should be capable of review and challenge by appropriately qualified professionals. Confidentiality, information security and lawful data use must remain integral to its operation.
Governance is not an administrative requirement separate from the concept of real intelligence. It establishes the conditions under which information can be trusted and conclusions can be justified. A system that produces persuasive answers without sufficient evidence cannot provide a reliable basis for understanding the property insured.
REAL INTELLIGENCE® therefore treats evidential integrity as a central institutional responsibility. The purpose is to make the insurer more capable of distinguishing reality from its representation and of acting responsibly when the available evidence remains incomplete.
The Research Imperative
The future of fine art and specie insurance will increasingly depend upon the relationship between machine analysis and specialist knowledge. The volume of documentation, imaging, valuation information, market evidence and historical records surrounding valuable property continues to create opportunities for more systematic analysis.
Relevant research includes machine learning, computer vision, document analysis, knowledge representation, provenance systems, image comparison, valuation methods and the assessment of uncertainty in analytical systems. It must also encompass the specialist fields that give the evidence its meaning: art history, conservation, gemology, horology, metallurgy, valuation, security and the legal and commercial circumstances surrounding ownership and movement.
The objective is not to assume that every specialist task can be automated. It is to determine where artificial intelligence improves the reliability, accessibility or interpretation of evidence and where physical examination, expert judgement or additional documentation remain necessary.
Research must distinguish between a system that performs well on familiar examples and one that remains reliable when confronted with unusual or historically significant objects. It must consider whether conclusions can be explained, whether the available data are representative and whether the resulting output supports the decision for which it is intended.
The wider research programme of X from GENERAL INTELLIGENCE PLC encompasses human intelligence, artificial intelligence, machine intelligence, collective intelligence and institutional intelligence. REAL INTELLIGENCE® applies this enquiry to fine art and specie insurance, where an accurate understanding of the actual object is the foundation upon which sound insurance judgement must rest.
The REAL INTELLIGENCE® Test
A fine art and specie insurer should ultimately be able to ask a demanding series of questions.
- Does it understand the actual object rather than merely its description?
- Can it distinguish an established fact from an attribution, assertion or inference?
- Does it understand what its documents prove and what remains uncertain?
- Can it connect the object's identity with its provenance, condition, valuation, location, custody and security arrangements?
- Does it understand the basis of the valuation and whether it remains appropriate to the purpose of the insurance?
- Can it distinguish the financial value of an individual object from the exposures created by the wider collection or portfolio?
- Does it recognise how movement, storage and changes in condition affect the risk? Can it investigate claims through a coherent body of evidence?
- Can artificial intelligence improve the accessibility and consistency of information without concealing uncertainty?
- Are specialist conclusions open to verification?
- Does the institution preserve valuable knowledge as institutional memory?
- Can it learn from losses and use that experience to improve future assessment?
- And, above all, does the insurer understand the real property it has undertaken to protect, the evidence supporting its description and value and the circumstances in which its loss or damage may occur?
If the answers are increasingly affirmative, the institution has begun to develop the capability that REAL INTELLIGENCE® is intended to represent: a disciplined understanding of the actual object, grounded in evidence and strengthened by the complementary capabilities of artificial intelligence and specialist human judgement.
The Objective
The intellectual history of real concerns the distinction between what actually exists and the representations through which it is described. The history of intelligence concerns the capacity to understand, discern and interpret. Their combination provides a useful organising principle for fine art and specie insurance, where the quality of an insurance decision depends upon the relationship between a physical object and the information used to identify, value and protect it.
REAL INTELLIGENCE® is an artificial intelligence consultancy framework provided by X from GENERAL INTELLIGENCE PLC to fine art and specie insurance providers. Its purpose is to examine how artificial intelligence can strengthen the institution's understanding of the property it insures: its identity, authenticity, provenance, condition, value, location, custody, movement and security.
The framework does not assume that reality can be represented completely by a database or that machine intelligence can resolve every question of attribution, condition or valuation. Some judgements require physical examination, specialist knowledge, additional evidence or an acknowledgement that uncertainty remains. Its ambition is to make those distinctions clearer, connect the relevant evidence and improve the quality of the professional decisions made in light of it.
The most capable insurer will not necessarily be the organisation with the largest collection of records or the most sophisticated analytical systems. It will be the institution that understands what its records establish, recognises what they do not establish, preserves specialist expertise and uses artificial intelligence to bring evidence into a more coherent relationship with the object itself.
REAL INTELLIGENCE® therefore represents a transition from the administration of descriptions to the disciplined understanding of reality: from inventories to evidence; from isolated valuations to informed judgements about value; from disconnected histories to coherent provenance; from periodic records to a more continuous understanding of condition and movement; from individual claims to institutional learning; and from artificial intelligence as an information-processing instrument to artificial intelligence as a means of strengthening the insurer's understanding of the actual property it protects.
The objective is not to replace the expert, remove uncertainty or reduce exceptional objects to standardised data. It is to create an insurance institution whose decisions are more closely grounded in the physical, historical, financial and operational realities of the risks it undertakes.
Artificial intelligence is the means. REAL INTELLIGENCE® is the capability. Evidential integrity is the discipline. The accurate understanding and responsible protection of exceptional value is the objective.