Timo Honkela: Ihminen+ -esitelmä, Mikkeli, 22.9.2016Timo Honkela
Timo Honkela discusses "Wise Machines" that help people to acquire knowledge and to understand and solve problems related to economy/business, emotions and health.
Timo Honkela: Tekoälyn ja koneoppimisen uhat ja mahdollisuudet, Turku, 27.10....Timo Honkela
Professor Timo Honkela's Studia Generalia presentation for the Society of Futures Studies at Turku University, Finland, on Thursday 27th of October, 2016. The title of the talk is "Threats and opportunities related to artificial intelligence and machine learning". The topics include 1) an introduction to AI and ML, 2) information on why AI and ML are societally relevant just in this moment of history, 3) natural language processing based on ML, 4) presentation of meta-analysis in humanities (cf. Helsinki Studia Generalia presentation a week earlier), 5) AI and intuition, and 6) discussion on the positive and negative scenarios related to AI and ML.
Esitelmä Eduskunnan tulevaisuusvaliokunnalle 4.5.2016 "Tieteen uuden valtatiet: Tiede 2.0 ja digitaaliset ihmistieteet".
A talk to the Committee for the Future of Finnish Parliament with the topic "New highways of science: Science 2.0 and Digital Humanities".
Timo Honkela: Ihminen+ -esitelmä, Mikkeli, 22.9.2016Timo Honkela
Timo Honkela discusses "Wise Machines" that help people to acquire knowledge and to understand and solve problems related to economy/business, emotions and health.
Timo Honkela: Tekoälyn ja koneoppimisen uhat ja mahdollisuudet, Turku, 27.10....Timo Honkela
Professor Timo Honkela's Studia Generalia presentation for the Society of Futures Studies at Turku University, Finland, on Thursday 27th of October, 2016. The title of the talk is "Threats and opportunities related to artificial intelligence and machine learning". The topics include 1) an introduction to AI and ML, 2) information on why AI and ML are societally relevant just in this moment of history, 3) natural language processing based on ML, 4) presentation of meta-analysis in humanities (cf. Helsinki Studia Generalia presentation a week earlier), 5) AI and intuition, and 6) discussion on the positive and negative scenarios related to AI and ML.
Esitelmä Eduskunnan tulevaisuusvaliokunnalle 4.5.2016 "Tieteen uuden valtatiet: Tiede 2.0 ja digitaaliset ihmistieteet".
A talk to the Committee for the Future of Finnish Parliament with the topic "New highways of science: Science 2.0 and Digital Humanities".
Current concepts in Asepsis and Infection control in a Dental ClinincArun1g
Current concepts in Asepsis and Infection control in a Dental Clininic. Lecture deilevered to Indian dental association Malanadu Branch.kerala, India.
By Dr Arun George MDS, Cosultant Maxillofacial surgeon India
La production cartographique pour les SIG version WebEric Lacoursiere
Présentation effectuée le 20 mai 2010 à Montréal. Elle couvre les différentes options disponibles aux utilisateurs des produits ESRI pour ajouter le volet de diffusion web à leur SIG.
Timo Honkela: Meaning negotiations as phenomenon and as languages technology...Timo Honkela
Abstract:
Models of linguistic semantics can be viewed through representation and reasoning. This distinction concerns questions on how do we represent the world that we refer to by linguistic expressions and what kind of reasoning do we apply based on these representations. It has been commonplace to assume that each word or expression has one or a limited number of different, distinct senses. The classification task of disambiguation has been devised to find the right reference in each case. It is also possible to represent the world using a high-dimensional continuous space. In that case, we do not need to assume that the world is represented as a network of nodes and their connections. These mathematical representations go beyond the capacities provided by symbolic logic. The word embeddings has history that stems from vector space representations in information retrieval. When a framework of multidimensional continuous spaces is available, it is possible to study nuances of meaning that go beyond conducting disambiguation or choosing between alternatives within a logical framework.
In the present work, it is postulated that semantic processes are essentially subjective and thus individual. When high-dimensional continuous spaces are used to represent meanings and defining contextual distributions, subjective aspects can be modelled. It is possible to measure subjectivity of meaning. This can be studied, for instance, in the framework of brain research (Saalasti et al. 2019) or motions tracking (Honkela & Förger 2013). The methodology or measuring subjective contextually grounded meaning has been been presented, for instance, in Raitio et al. 2014. Further methodological work and an empirical demonstration is presented in Sintonen et al. (2014). When it is possible to represent individually contextual meaning of expressions, it is consequently possible to analyse the differences of meaning between two individuals. A hypothesis is that suitable data for the purpose of meaning negotiation can be collected, computational algorithms devised and applied in real world contexts that helps in meaning negotiations. An alternative view is to aim at defining the meaning of words in a precise way and to teach all people to use these definition. In this present work, it is claimed that that objectivity can be reached only to a degree as it would require vast human cognitive and time resources and the mapping between words and the world is doomed to be partial. This concern has implications both in scientific and in real world communication and representation and has been applied in building the Peace Machine framework.
Meaning negotiations
as phenomena and
as LT challenges
Timo Honkela
University of Helsinki
with Iiro Jääskeläinen (Aalto University) on
the Study of Individualized Meanings
using Brain Research
University of Helsinki, Topelia, F211
4th of April, 2019
Timo Honkela: Meaning negotiations as phenomenon and as languages technology ...Timo Honkela
Abstract:
Models of linguistic semantics can be viewed through representation and reasoning. This distinction concerns questions on how do we represent the world that we refer to by linguistic expressions and what kind of reasoning do we apply based on these representations. It has been commonplace to assume that each word or expression has one or a limited number of different, distinct senses. The classification task of disambiguation has been devised to find the right reference in each case. It is also possible to represent the world using a high-dimensional continuous space. In that case, we do not need to assume that the world is represented as a network of nodes and their connections. These mathematical representations go beyond the capacities provided by symbolic logic. The word embeddings has history that stems from vector space representations in information retrieval. When a framework of multidimensional continuous spaces is available, it is possible to study nuances of meaning that go beyond conducting disambiguation or choosing between alternatives within a logical framework.
In the present work, it is postulated that semantic processes are essentially subjective and thus individual. When high-dimensional continuous spaces are used to represent meanings and defining contextual distributions, subjective aspects can be modelled. It is possible to measure subjectivity of meaning. This can be studied, for instance, in the framework of brain research (Saalasti et al. 2019) or motions tracking (Honkela & Förger 2013). The methodology or measuring subjective contextually grounded meaning has been been presented, for instance, in Raitio et al. 2014. Further methodological work and an empirical demonstration is presented in Sintonen et al. (2014). When it is possible to represent individually contextual meaning of expressions, it is consequently possible to analyse the differences of meaning between two individuals. A hypothesis is that suitable data for the purpose of meaning negotiation can be collected, computational algorithms devised and applied in real world contexts that helps in meaning negotiations. An alternative view is to aim at defining the meaning of words in a precise way and to teach all people to use these definition. In this present work, it is claimed that that objectivity can be reached only to a degree as it would require vast human cognitive and time resources and the mapping between words and the world is doomed to be partial. This concern has implications both in scientific and in real world communication and representation and has been applied in building the Peace Machine framework.
Meaning negotiations
as phenomena and
as LT challenges
Timo Honkela
University of Helsinki
with Iiro Jääskeläinen (Aalto University) on
the Study of Individualized Meanings
using Brain Research
University of Helsinki, Topelia, F211
4th of April, 2019
Timo Honkela: From early to later Wittgenstein and Artificial IntelligenceTimo Honkela
Professor Timo Honkela presented an argument that there is a analogy between the developments that took place in Ludwig Wittgenstein's philosophy and in the artificial intelligence when turned away from relying rule-based systems. Honkela also discusses more in general epistemological questions, the underlying questions regarding the objectives and motivations of formalisation. Formalization often relies on assumptions such as the basic or primary role of objects, relations and properties or truth values and propositions. Honkela proposed a pattern and distribution based epistemology as an alternative.
Current concepts in Asepsis and Infection control in a Dental ClinincArun1g
Current concepts in Asepsis and Infection control in a Dental Clininic. Lecture deilevered to Indian dental association Malanadu Branch.kerala, India.
By Dr Arun George MDS, Cosultant Maxillofacial surgeon India
La production cartographique pour les SIG version WebEric Lacoursiere
Présentation effectuée le 20 mai 2010 à Montréal. Elle couvre les différentes options disponibles aux utilisateurs des produits ESRI pour ajouter le volet de diffusion web à leur SIG.
Timo Honkela: Meaning negotiations as phenomenon and as languages technology...Timo Honkela
Abstract:
Models of linguistic semantics can be viewed through representation and reasoning. This distinction concerns questions on how do we represent the world that we refer to by linguistic expressions and what kind of reasoning do we apply based on these representations. It has been commonplace to assume that each word or expression has one or a limited number of different, distinct senses. The classification task of disambiguation has been devised to find the right reference in each case. It is also possible to represent the world using a high-dimensional continuous space. In that case, we do not need to assume that the world is represented as a network of nodes and their connections. These mathematical representations go beyond the capacities provided by symbolic logic. The word embeddings has history that stems from vector space representations in information retrieval. When a framework of multidimensional continuous spaces is available, it is possible to study nuances of meaning that go beyond conducting disambiguation or choosing between alternatives within a logical framework.
In the present work, it is postulated that semantic processes are essentially subjective and thus individual. When high-dimensional continuous spaces are used to represent meanings and defining contextual distributions, subjective aspects can be modelled. It is possible to measure subjectivity of meaning. This can be studied, for instance, in the framework of brain research (Saalasti et al. 2019) or motions tracking (Honkela & Förger 2013). The methodology or measuring subjective contextually grounded meaning has been been presented, for instance, in Raitio et al. 2014. Further methodological work and an empirical demonstration is presented in Sintonen et al. (2014). When it is possible to represent individually contextual meaning of expressions, it is consequently possible to analyse the differences of meaning between two individuals. A hypothesis is that suitable data for the purpose of meaning negotiation can be collected, computational algorithms devised and applied in real world contexts that helps in meaning negotiations. An alternative view is to aim at defining the meaning of words in a precise way and to teach all people to use these definition. In this present work, it is claimed that that objectivity can be reached only to a degree as it would require vast human cognitive and time resources and the mapping between words and the world is doomed to be partial. This concern has implications both in scientific and in real world communication and representation and has been applied in building the Peace Machine framework.
Meaning negotiations
as phenomena and
as LT challenges
Timo Honkela
University of Helsinki
with Iiro Jääskeläinen (Aalto University) on
the Study of Individualized Meanings
using Brain Research
University of Helsinki, Topelia, F211
4th of April, 2019
Timo Honkela: Meaning negotiations as phenomenon and as languages technology ...Timo Honkela
Abstract:
Models of linguistic semantics can be viewed through representation and reasoning. This distinction concerns questions on how do we represent the world that we refer to by linguistic expressions and what kind of reasoning do we apply based on these representations. It has been commonplace to assume that each word or expression has one or a limited number of different, distinct senses. The classification task of disambiguation has been devised to find the right reference in each case. It is also possible to represent the world using a high-dimensional continuous space. In that case, we do not need to assume that the world is represented as a network of nodes and their connections. These mathematical representations go beyond the capacities provided by symbolic logic. The word embeddings has history that stems from vector space representations in information retrieval. When a framework of multidimensional continuous spaces is available, it is possible to study nuances of meaning that go beyond conducting disambiguation or choosing between alternatives within a logical framework.
In the present work, it is postulated that semantic processes are essentially subjective and thus individual. When high-dimensional continuous spaces are used to represent meanings and defining contextual distributions, subjective aspects can be modelled. It is possible to measure subjectivity of meaning. This can be studied, for instance, in the framework of brain research (Saalasti et al. 2019) or motions tracking (Honkela & Förger 2013). The methodology or measuring subjective contextually grounded meaning has been been presented, for instance, in Raitio et al. 2014. Further methodological work and an empirical demonstration is presented in Sintonen et al. (2014). When it is possible to represent individually contextual meaning of expressions, it is consequently possible to analyse the differences of meaning between two individuals. A hypothesis is that suitable data for the purpose of meaning negotiation can be collected, computational algorithms devised and applied in real world contexts that helps in meaning negotiations. An alternative view is to aim at defining the meaning of words in a precise way and to teach all people to use these definition. In this present work, it is claimed that that objectivity can be reached only to a degree as it would require vast human cognitive and time resources and the mapping between words and the world is doomed to be partial. This concern has implications both in scientific and in real world communication and representation and has been applied in building the Peace Machine framework.
Meaning negotiations
as phenomena and
as LT challenges
Timo Honkela
University of Helsinki
with Iiro Jääskeläinen (Aalto University) on
the Study of Individualized Meanings
using Brain Research
University of Helsinki, Topelia, F211
4th of April, 2019
Timo Honkela: From early to later Wittgenstein and Artificial IntelligenceTimo Honkela
Professor Timo Honkela presented an argument that there is a analogy between the developments that took place in Ludwig Wittgenstein's philosophy and in the artificial intelligence when turned away from relying rule-based systems. Honkela also discusses more in general epistemological questions, the underlying questions regarding the objectives and motivations of formalisation. Formalization often relies on assumptions such as the basic or primary role of objects, relations and properties or truth values and propositions. Honkela proposed a pattern and distribution based epistemology as an alternative.
Timo Honkela: Peace Machine: Peace from a difference perspective - Dialogue o...Timo Honkela
A presentation given in the National Dialogues Conference in Helsinki, Finland. The theme was how to use artificial intelligence, machine learning and other similar technologies to promote peace in the world. The three ares considered were language and meaning, emotions and society. Computers can help humans, for instance, by improving mutual understanding through meaning negotiations.
Timo Honkela: Turning quantity into quality and making concepts visible using...Timo Honkela
Professor Timo Honkela gave an invited talk in the Göran Mickwitz seminar that took place in Helsinki, 9th of February 2017. The event was organized in the honor of Doc. Jessica Parland-von Essen.
Timo Honkela: Tietokone lukemassa yli 100 miljoonaa eri kirjaa: Kielitieteen ...Timo Honkela
Professori Timo Honkelan esitelmä luonnonfilosofian seurassa 24.1.2017 aiheesta "Tietokone lukemassa yli 100 miljoonaa eri kirjaa: Kielitieteen ja filosofian näkökulmia". Tilaisuus järjestettiin Tieteiden talossa huoneessa 505. Erityisen tarkastelun kohteina olivat erilaiset merkitysteoriat ja niiden suhde koneoppimisen tutkimukseen. Yksi keskeinen johtopäätös oli, että koneoppimis- ja neuroverkkotutkimus tarjoaa tietoteoreettisille tarkasteluille uutta pohjaa.
Title in English:
Computer reading over one hundred books: Linguistics and philosophical views
Timo Honkela: Introducing the book Encyclopedia of Artificial Intelligence (i...Timo Honkela
The book "Encyclopedia of Artificial Intelligence" (Tekoälyn ensyklopedia), edited by Eero Hyvönen, Ilkka Karanta and Markku Syrjänen (1993), was an important landmark in the Finnish AI research and development. Finland has been and remains an important country in this field. Many of the authors were already prominent figures at that time, especially Professor and later Academician Teuvo Kohonen. Since then many authors have become processors in different universities. One of the authors, Tuomas Sandholm received in 2003 the IJCAI Computers and Thought Award and serves as a professor at Carnegie Mellon University.
Timo Honkela: Kohonen's Self-Organizing Maps for Intelligent Systems Developm...Timo Honkela
An invited talk given in the FODO'98, Foundations of Data Organization conference. The conference took place in Kobe, Japan, November 12-13, 1998. Main themes of the talk included Self-Organizing Maps (SOMs), Fuzzy Sets, context analysis, and systems of SOMs.
Timo Honkela: Analysis of Qualitative Data using Machine Learning MethodsTimo Honkela
A keynote talk given in the Arcada Analytics Workshop, Helsinki on 8th of June, 2016. The other keynote talks were given by Peter Sarlin (Machine Learning and Network Analytics for Measuring Systemic Risk) and Amaury Lendasse (ELMVIS+: Fast Nonlinear Visualization Technique based on Cosine Distance and Extreme Learning Machines).
Timo Honkela: Modeling evolution and dynamical systemsTimo Honkela
A short tutorial for Modeling Meaning and Knowledge, 18 April 2016, at University of Helsinki. Followed by presentations by Mika Pantzar and Sakari Virkki.
Timo Honkela: Metaphors, analogies and conceptual blending
Timo Honkela: Miten tekoäly muuttaa oppimista ja työtä? Kalajoen lukio, 17.8.2016
1. Timo Honkela, 17.8.2016, Kalajoen lukio
Timo Honkela
17.8.2016
Miten tekoäly muuttaa
oppimista ja työtä?
timo.honkela@helsinki.fi
Kalajoen lukio
http://375humanistia.helsinki.fi/humanistit/timo-honkela
3. Timo Honkela, 17.8.2016, Kalajoen lukio
Läsnäolo- ja
keskittymisharjoitus
● Olenko läsnä, hallitsenko ajatuksesi?
– Keskittymisen esteitä: ajatukset ovat muualla,
huomio kiinnittyy epäolennaisiin asioihin kuten
puhujan alkuasuun tai ääneen, itsessä elää
vastustamisen tarve, tms.
● Viritämme yhdessä motivaation
● Ajatelkaamme silmät suljettuna hetki
kysymyystä, mihin osaamiseen perustuu
kymmenen vuoden kuluttua se, mitä teen
työkseni
6. Timo Honkela, 17.8.2016, Kalajoen lukio
Esihistoriasta keskiaikaan
Eläimet apuna
Kirjapainotaito
Johannes
Gutenberg
1398-1468
https://fi.wikipedia.org/wiki/Johannes_Gutenberghttps://fi.wikipedia.org/wiki/Maanviljelyn_synty_L%C3%A4hi-id%C3%A4ss%C3%A4
7. Timo Honkela, 17.8.2016, Kalajoen lukio
Teknologinen murros:
koneiden aika
Koneet teillä,
tehtaissa ja pelloilla
James Watt
1736-1819
Henry Ford
1863-1947
https://fi.wikipedia.org/wiki/H%C3%B6yrykone https://fi.wikipedia.org/wiki/Traktori https://fi.wikipedia.org/wiki/Henry_Ford
8. Timo Honkela, 17.8.2016, Kalajoen lukio
Tietoverkkojen ja henkilökohtaisten
tietokoneiden yleistyminen
Henkilökohtaiset
tietokoneet
https://fi.wikipedia.org/wiki/Henkil%C3%B6kohtainen_tietokone
Apple II
(1977)
Commodore
64 C (1986)
https://en.wikipedia.org/wiki/World_Wide_Web
Netti (internet)
veppi (web)
Tim
Berners-
Lee1989
1960s
9. Timo Honkela, 17.8.2016, Kalajoen lukio
Valtavat ohjelmistomassat
yhteiskunnan peruspilarina
http://www.informationisbeautiful.net/visualizations/million-lines-of-code/
Ohjelmarivejä:
- Unix 1.0 ~ 10.000
- Windows 3.1 ~ 2.000.000
- Firefox ~ 10.000.000
- Facebook ~ 60.000.000
- Googlen palvelut ~
2.000.000.000
Suomalaislähtöisiä:
- MySQL ~ 13.000.000
- Linux 3.1 ~ 15.000.000
11. Timo Honkela, 17.8.2016, Kalajoen lukio
Tekoälystä
● Pyritään mallintamaan ja matkimaan ihmisen
älykkyyden ja osaamisen eri osa-alueita
– Lukeminen ja kirjoittaminen
– Puhuminen ja kuunteleminen
– Päättely, tunteiden hallinta ja ongelmanratkaisu
– Aistitiedon tulkinta ja hahmontunnistus
– Oppiminen
– Luovuus
– Maailmassa pärjääminen (robotiikka)
12. Timo Honkela, 17.8.2016, Kalajoen lukio
Tietokoneet ja tekoäly
ihmisten keskellä?
● Koneet eivät ole ihmisiä ja
ihmiset eivät ole koneita
● Koneiden käyttö saattaa pakottaa ihmisiä
toimimaan konemaisesti
● Niinpä ihmisen kykyjen matkiminen voi
auttaa siinä, että koneiden kanssa
voi toimia aiempaa luontevammin
● (Tietokoneiden käyttö on edelleen usein
varsin vaikeaa)
13. Timo Honkela, 17.8.2016, Kalajoen lukio
Perinteinen tekoäly:
“Ohjelmoitua älyä”
● Aiemmin ajateltiin, että päättelyä ja
ongelmanratkaisua saadaan siirrettyä koneelle
kysymällä parhailta asiantuntijoilta, mitä he
tietävät ja miten he päättelevät
● Tekoälyjärjestelmiä ohjelmoitiin keräämällä
sääntöjä, joita kirjattiin tietokoneelle
● Esimerkkinä voi tarkastella vaikkapa
sairauksien tunnistamista oireista ja
mittaustuloksista
14. Timo Honkela, 17.8.2016, Kalajoen lukio
Esimerkki asiantuntijasäännöstä
MYCIN-järjestelmässä 80-luvulla
Rule-Based Expert Systems
The MYCIN Experiments of the Stanford Heuristic Programming Project
http://people.dbmi.columbia.edu/~ehs7001/Buchanan-Shortliffe-1984/Chapter-04.pdf
15. Timo Honkela, 17.8.2016, Kalajoen lukio
Koneoppiminen:
“Oppiva tekoäly”
● Tilastollisen koneoppimisen avulla voidaan
matkia sitä
– Miten ihminen oppii sääntöjä esimerkkien avulla,
tai
– Miten ihminen kerää intuitiivista tietoa
hermoverkkojensa rakenteisiin
Kielellinen,
sääntömuotoinen tieto
Intuitio,
kokemusperäinen tieto
16. Timo Honkela, 17.8.2016, Kalajoen lukio
Tekoälyn kehittäminen lisää
ymmärrystä ihmismielestä
● Tekoäly- ja hermoverkkotutkija ymmärtää
hyvin, miksi “kannettu vesi ei kaivossa pysy”
● On myös selvää, miksi kokenut ihminen ratkoo
ongelmia yleensä paremmin kuin pelkästään
“kirjanoppinut”; parasta on yhdistelmä
taustatietoa ja kokemusta
17. Timo Honkela, 17.8.2016, Kalajoen lukio
Esimerkki keinotekoisesta hermoverkosta ja
pieni osa ihmisen aivojen hermoverkosta
http://arxiv.org/pdf/1507.02672v1.pdf
Rasmus, Valpola,
Honkala. Berglund, Raiko
https://en.wikipedia.org/wiki/Biological_neural_network
18. Timo Honkela, 17.8.2016, Kalajoen lukio
Tietokoneiden ohjelmoinnista
oppiviin koneisiin
● Suomi on ollut yksi edelläkävijä
koneoppimisen ja erityisestä keinotekoisten
neuroverkkojen alueella
● Uranuurtaja on ollut akateemikko Teuvo
Kohonen, joka muotoili vuonna 1981
maailmanmenestyksen saavuttaneen
itseorganisoiva kartta (Self-Organizing Map,
SOM) -menetelmänsä, joka on samalla
erinomainen malli aivokuoren
järjestäytymisestä
19. Timo Honkela, 17.8.2016, Kalajoen lukio
Koneoppiminen ja tiedon louhinta
● Tilastollisen koneoppimisen avulla matkitaan
ihmisen oppimista
● Kone käy läpi sille annettuja esimerkkejä
● Annetun datan perusteella kone oppii
esimerkiksi luokittelemaan, järjestämään,
ryhmittelemään, hahmottamaan,
käsitteistämään tai laittamaan
paremmuusjärjestykseen erilaisia tietoalkioita
http://www.tynka.fi/
20. Timo Honkela, 17.8.2016, Kalajoen lukio
Koneille opetetaan kieltä
Koneet oppivat kieltä
● Perinteisesti koneet on pyritty saamaan
“kielitaitoisiksi” kirjoittamalla kielen sääntöjä
koneen ymmärtämään muotoon
● Tämän on kuitenkin osoittautunut
ongelmalliseksi
● Nykyisin hyödynnetään
koneoppimismenetelmiä
21. Timo Honkela, 17.8.2016, Kalajoen lukio
Sanojen suhteet
paljastuvat niiden käytöstä
● Kun käytettävissä on suuria tekstiaineistoja,
mielivaltaisen kielen sanojen välisiä suhteita
voidaan selvittää tilastollisesti
● Perusidea on se, että kahta sanaa käytetään
tyypillisesti samaan tapaan (samanlaisessa
lauseyhteydessä), jos niiden merkitykset ja/tai
kieliopillinen rooli on samankaltainen
22. Timo Honkela, 17.8.2016, Kalajoen lukio
Itsenäistä
pohdiskelua
Keskustelua
vierustoverin
kanssa
Miten tekoäly ja koneoppiminen
muuttaa alaa, josta olen kiinnostunut
24. Timo Honkela, 17.8.2016, Kalajoen lukio
Tekoäly muuttamassa työtä:
Automaattinen kirjastonhoitaja?
25. Timo Honkela, 17.8.2016, Kalajoen lukio
Tekoäly ja koneoppiminen
kirjastossa
● Automaattinen asiasanoitus
● Dokumenttien automaattinen luokittelu
● Kunkin dokumentin sijoittaminen yhteen tai
usempaan luokkaan; ehkä erilaisilla
jäsenyysasteilla
● Dokumenttien ryhmittely
luokittelun sijaan tai lisäksi
● Virtuaalinen
kirjasto
WEBSOM: Honkela, Kaski,
Kohonen, Lagus (1996...)
27. Timo Honkela, 17.8.2016, Kalajoen lukio
Robottiautot tulevat
https://fi.wikipedia.org/wiki/Google_driverless_car https://fi.wikipedia.org/wiki/Robottiauto
28. Timo Honkela, 17.8.2016, Kalajoen lukio
Tekoäly muuttamassa työtä:
Kirjoittava ja piirtävä kone?
29. Timo Honkela, 17.8.2016, Kalajoen lukio
Luovat koneet
http://deepdreamgenerator.comhttps://www.cs.helsinki.fi/en/story/82156/brain-poetry
Toivonen, Toivanen, Kantosalo,
Xiao, Kantosalo, Valitutti, Gross et al.
31. Timo Honkela, 17.8.2016, Kalajoen lukio
Tunteita analysoivat koneet
● Koneita voidaan tiedon ja järjen sisältöjen lisäksi yhä
lisääntyvässä määrin käyttää myös tunteiden
analysointiin
● Nykyään on suosittua tehdä ns. sentimenttianalyysia
esimerkiksi asiakaspalautteen selvittämiseen: mistä
tuotteista tai palveluista asiakkaat ovat olleet
tyytyväisiä, vihaisia, tms?
● Tunteiden maailmaa voi mallintaa
ilmiöiden taustojen, dynamiikan ja keskinäisten
tekijöiden näkökulmasta
33. Timo Honkela, 17.8.2016, Kalajoen lukio
Lääketiede ja hyvinvointi
● Lääketieteessä suuret tietoaineistot ja
koneoppimismenetelmät mahdollistavat
aiempaa tarkemmat ja nopeammat diagnoosit
ja koneen antamat yksilölliset hoitosuositukset
● Elintapojen ja hoitojen vaikutuksia voidaan
tutkia aiempaa tarkemmin ottaen huomioon
jopa tuhansia tekijöitä
● Elintapojen vaikutusten selvittäminen ja niiden
suhde geeniperimään voi parhaimmillaan estää
joitakin sairauksia puhkeamasta
35. Timo Honkela, 17.8.2016, Kalajoen lukio
Robotit kehittyvät – kuinka pitkälle?
https://en.wikipedia.org/wiki/Robot_%26_Frank https://en.wikipedia.org/wiki/Robot-assisted_surgery
37. Timo Honkela, 17.8.2016, Kalajoen lukio
Koulutuksen ja oppimisen muutos
● Tieverkkojen ja -haun ansiosta monenlainen tieto on
helposti saavutettavissamme
● Ulkoa oppimisen merkitys on dramaattisesti
vähentynyt
● On edelleen tärkeää ymmärtää käsitteitä, asioiden
välisiä suhteita ja soveltaa tietoa
● Näyttää myös siltä, että erikoistuminen on entistä
tärkeämpää
● Ajatus kannattaa myös uhrata sille,
miten ihminen ja kone tekevät
yhteistyötä
39. Timo Honkela, 17.8.2016, Kalajoen lukio
Koneoppimisen hyödyntäminen
on olennaista kilpailukyvyn
ylläpitämisessä
40. Timo Honkela, 17.8.2016, Kalajoen lukio
Miten käy, jos emme
hyödynnä koneita (koneoppimista)
nykyistä paremmin?
https://www.pinterest.com/pin/509680882801748515/
Emme pärjää kilpailussa
junaan käyttävälle
kilpailijallemme, vaikka
annamme hevosille
vähemmän kauraa,
ratsastajille pienempää
palkkaa, …
… tuotamme
halvempaa
energiaa,
karsimme
kustannuksia,
pienennämme
työntekijöiden
palkkoja, ...
41. Timo Honkela, 17.8.2016, Kalajoen lukio
Kiitos!
http://375humanistia.helsinki.fi/humanistit/timo-honkela
http://www.slideshare.net/timohonkela
https://www.youtube.com/watch?v=UXwkGPMMZdk