liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Development of AI-tools for making sense of future complex intelligent systems
Linköping University, Department of Management and Engineering, Project Innovations and Entrepreneurship. Linköping University, Faculty of Science & Engineering.
2024 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Artificial intelligence (AI) is increasingly introduced into many systems that modern society rely on and is often portrayed as a savior that can contribute to finding solutions to societal challenges e.g., social, and ecological sustainability. Many of these systems can be classified as complex systems, with interdependencies, emergent behaviors and a diversity of actors involved. As AI is increasingly introduced into these systems, we witness a transformation from complex systems into complex intelligent systems. At the same time caution is invoked toward the risks of AI regarding e.g., biases and loss of control as more tasks are transferred to AI. Hence, the introduction of AI into complex systems is associated with uncertainties around management of AI initiatives and their influence on future systems. Challenges like this can affect many different functions and professions and thus need to be understood collectively.

The aim of this thesis is to examine how the actors’ prospective collective sensemaking processes in developing complex systems are affected by AI introduction. Previous research within complex systems literature shows important aspects regarding sensemaking of the system and situations within operations of complex systems. However, sensemaking in the development process of complex systems has been less studied. By examining the introduction of AI in complex systems development this thesis explores collective prospective sensemaking processes in the development of complex intelligent systems.

To study an emerging phenomenon like AI introduction in complex systems, an explorative case study was found suitable. The case chosen for the study was a cross-organizational development project of an AI-tool based on Machine Learning for planning of energy systems to be used in the urban planning process of new city districts. This setting revealed plenty prospective and collective sensemaking occasions around AI introduction and exhibited continuous engagement in prospective collective sensemaking relating to the development of the AI tool and the imagined use of the AI tool in the system, which have been reported in the two appended papers. The first paper showed misalignment between actors’ sensemaking processes that alternated between seeking and disengaging behaviors. It also identified the use of boundary objects to retain disengaged actors and raised considerations around the level of detail of the boundary objects in relation to the sensemaking behaviors. The second paper identified dependencies between near- and distant-sensemaking loops that highlight challenges to connect retrospective insights with prospective imaginations by action in the present. In the creation of complex intelligent systems, human involvement seems inevitable, and the second paper exposes how AI can augment human cognition and organizational capabilities for creative imagination around possible and desirable distant-future scenarios.

This thesis extends previous research on prospective and collective sensemaking in the development process of complex intelligent systems by presenting a framework of near and distant future sensemaking and internal and external complexity. This provided new insights of how knowledge flows over system levels and how to use boundary objects throughout such projects. Insight that can be useful for management of purposeful AI introduction in complex systems and society. Moreover, it contributes with an empirical case of AI introduction in complex systems to the innovation management literature.

Abstract [sv]

Artificiell intelligens (AI) integreras alltmer i de system som vårt moderna samhälle vilar på och framställs ofta som en viktig faktor för att lösa de utmaningar som samhället står inför, till exempel social och ekologisk hållbarhet. Många av dessa system kan definieras som komplexa system, med ömsesidiga beroenden, oförutsedda beteenden och en mångfald av inblandade aktörer. När AI introduceras i sådana system kan vi skönja en transformation från komplexa system till komplexa intelligenta system. Samtidigt påtalas ofta riskerna med AI avseende till exempel partiskhet och minskad kontroll när uppgifter tas över av AI. Införandet av AI i samhällskritiska system förknippas därmed med osäkerheter kring hanteringen av AI-initiativ och dess påverkan på framtiden. Detta leder till utmaningar som berör många olika funktioner och professioner och därmed behöver förstås kollektivt.

Syftet med avhandlingen är att undersöka aktörernas framtidsorienterade, gemensamma, meningsskapande processer under utveckling av komplexa system och hur de påverkas av introduktionen av AI. Tidigare forskning inom komplexa system har visat på viktiga aspekter gällande människors förståelse av systemet och situationer framförallt inom driften av komplexa system. Meningsskapande i utvecklingsprocessen av komplexa system har dock hittills inte uppmärksammats i samma utsträckning. Genom att undersöka introduktionen av AI i utvecklingen av komplexa system utforskar denna avhandling kollektiva, framåtblickande, meningsskapande processer inom utvecklingen av komplexa intelligenta system.

För studier av ett framväxande fenomen som AI-introduktion i komplexa system ansågs en explorativ fallstudie lämplig. Det valda fallet var ett utvecklingsprojekt av ett AI-verktyg baserat på maskininlärning med syfte att användas i planeringen av energisystem inom stadsplaneringsprocessen av nya stadsdelar och hade flera medverkande organisationer. Fallet visade på flera framtidsorienterade och gemensamma meningsskapande situationer kring AI-introduktion. Därmed synliggjordes de medverkandes kontinuerliga deltagande i framåtblickande gemensamt meningsskapande relaterat till utvecklingen av AI-verktyget och dess tänkta användningen i systemet, vilket rapporterats i de två bilagda artiklarna. Den första artikeln visade att de meningsskapande processerna var ur fas mellan aktörerna, vilka växlade mellan sökande och oengagerade beteenden, och att användningen av gränsöverskridande objekt med rätt detaljnivå kan involvera oengagerade aktörer. Den andra artikeln identifierade beroenden mellan olika meningsskapande cykler, närliggande respektive avlägsen framtid, vilket belyser utmaningar med att koppla tillbakablickande insikter till framåtblickande föreställningar om systemet genom handling i nuet. Vid skapande av komplexa intelligenta system framstår mänsklig inblandning som oundviklig, och den andra artikeln belyser även hur AI kan förstärka mänsklig kognition och organisatoriska förmågor för att främja kreativ föreställningsförmåga kring möjliga och önskvärda scenarier av en avlägsen framtid.

Avhandlingen bidrar genom att bredda tidigare forskning om framtidsorienterat och kollektivt meningsskapande i utvecklingsprocessen av komplexa intelligenta system genom att presentera ett ramverk av närliggande och avlägset framåtblickande och intern och extern komplexitet. Det visar på nya insikter om hur kunskap flödar över systemnivåer och gränsöverskridande objekt kan användas under sådana utvecklingsprojekt. Sådana insikter kan vara praktiskt användbara för en välgrundad AI-introduktion i komplexa system och i samhället i stort. Dessutom bidrar den med en empirisk fallstudie av resan mot AI-introduktion i komplexa system till litteraturen inom innovationsledning.

Place, publisher, year, edition, pages
Linköping: Linköping University Electronic Press, 2024. , p. 49
Series
Linköping Studies in Science and Technology. Licentiate Thesis, ISSN 0280-7971 ; 1992
National Category
Peace and Conflict Studies Other Social Sciences not elsewhere specified
Identifiers
URN: urn:nbn:se:liu:diva-202589DOI: 10.3384/9789180756228ISBN: 9789180756211 (print)ISBN: 9789180756228 (electronic)OAI: oai:DiVA.org:liu-202589DiVA, id: diva2:1852216
Presentation
2024-05-03, ACAS, A Building, Campus Valla, Linköping, 13:00 (Swedish)
Opponent
Supervisors
Note

Funding agency: Wallenberg AI, Autonomous Systems and Software Program – Humanity and Society (WASP-HS), funded by the Marianne and Marcus Wallenberg Foundation under Grant 2019-0126

Available from: 2024-04-17 Created: 2024-04-17 Last updated: 2025-02-20Bibliographically approved
List of papers
1. Prospective Sensemaking in the Front End of Innovation of AI Projects
Open this publication in new window or tab >>Prospective Sensemaking in the Front End of Innovation of AI Projects
2024 (English)In: Research technology management, ISSN 0895-6308, E-ISSN 1930-0166, Vol. 67, no 4, p. 72-83Article in journal (Refereed) Published
Abstract [en]

Overview: Using artificial intelligence (AI) to help develop new complex systems poses challenges for less tech-savvy organizations and may prolong the front end of innovation phase. Complications arise from diverging understandings of AI functionality and requirements among involved actors and the difficulties of determining the usefulness of AI in such a complex setting. This article explores a cross-industry project that entailed developing a functional prototype of an AI tool for planning (complex) energy systems in new city districts, engaging both system (domain) actors and AI developers. By analyzing prospective collective sensemaking processes in two episodes from the project, we discovered misaligned sensemaking processes between system actors and AI developers. During the project these actors alternated between "seeking" and "disengaging" sensemaking behavior. We highlight how various prototypes supported alignment in sensemaking processes concerning AI and progress in the project. Practitioners can use the managerial implications to better understand sensemaking dynamics in AI projects and implement suitable measures, like education or support at various stages of the project duration, to mitigate the problems that can arise due to misaligned sensemaking processes. PRACTITIONER TAKEAWAYS: Recognize that the diverse actors in AI projects have different sensemaking needs and that gaps in knowledge may exist. Actively monitor for signs of seeking and disengaging among actors, and support behaviors that facilitate prospective collective sensemaking. Use boundary objects to facilitate prospective collective sensemaking, keeping in mind that the level of detail can support actors in different ways.

Place, publisher, year, edition, pages
TAYLOR & FRANCIS INC, 2024
Keywords
Complex systems; Artificial intelligence; Prospective collective sensemaking; Multi-actor collaboration; Front end of innovation
National Category
Peace and Conflict Studies Other Social Sciences not elsewhere specified
Identifiers
urn:nbn:se:liu:diva-206421 (URN)10.1080/08956308.2024.2350407 (DOI)001266334500007 ()
Note

Funding Agencies|Wallenberg AI, Autonomous Systems and Software Program - Humanity and Society (WASP-HS) - Marianne and Marcus Wallenberg Foundation [2019-0126]

Available from: 2024-08-19 Created: 2024-08-19 Last updated: 2026-04-30

Open Access in DiVA

fulltext(6916 kB)879 downloads
File information
File name FULLTEXT01.pdfFile size 6916 kBChecksum SHA-512
730f7775011963fff8445214e352a5502f8c99b19ea43ae384df30dbacadf1d5c2c92c80688f1bb060f8c237a4635b1220669cd1a811927cbcc06a9d6e9ccfc2
Type fulltextMimetype application/pdf
Order online >>

Other links

Publisher's full text

Authority records

Särner, Elinor

Search in DiVA

By author/editor
Särner, Elinor
By organisation
Project Innovations and EntrepreneurshipFaculty of Science & Engineering
Peace and Conflict StudiesOther Social Sciences not elsewhere specified

Search outside of DiVA

GoogleGoogle Scholar
Total: 885 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
isbn
urn-nbn

Altmetric score

doi
isbn
urn-nbn
Total: 1485 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf