Upgrade to Pro — share decks privately, control downloads, hide ads and more …

The Human Experience in the upcoming AI Society. MEMEX21

Willi Schroll
September 30, 2021

The Human Experience in the upcoming AI Society. MEMEX21

=== MEMEX21 EVENT 2021-09-23 === eye square Berlin. eye-square.com === VIDEO ON YOUTUBE: j.mp/aisociety21 === the hyperlinks are clickable in the downloaded PDF === me at strategiclabs.de === connect at linkedin.com ===

Willi Schroll started his presentation about the „human experience in the upcoming AI society“ pointing to the challenge of massive technological acceleration. The futurist Gerd Leonhard articulated the cultural disruption of transformative technologies in the quote: "Humanity will change more in the next 20 years than the previous 300 years."

>>>Artificial intelligence will be the main game changer in almost all fields of society and business.<<< But though AI already is outperforming humans in many niche areas of cognitive tasks, this Artificial Narrow Intelligence (ANI) is unlikely to reach the stage of Artificial General Intelligence (AGI) soon. Anyway especially Machine Learning has delivered huge progress in performance and is successful in analytical and generative application fields as computer vision in self-driving cars, realtime voice translation or even therapeutical apps which entertain helpful conversations. A model case for an „AI first“ imperative in business is the successful corporate strategy "Flywheel“ of the digital company Amazon: Deep understanding of the patterns in the data deluge and continuously rebuilding the customer experience based on these insights.

Human beings as customers and citizens will encounter AI more and more and in many ways, directly and aware, but also indirectly and unaware of the AI processes involved. >>>The concept of an upcoming "AI Society“ assumes that work, lifestyle and interactions are dominantly ai-driven. The smart environments and the exchange with advanced virtual avatars (in AR and VR) and with physical robots will result in a deep change of the human experience.<<< But as foresight is thinking about the future in a pluralistic way, the often divergent scenarios mean that the human experiences will differ according to the type of AI Society, which will become a reality. Schroll illustrated the possible plurality of futures using a scenario study about the possible co-living of man and machine in 2040 (KPMG). The matrix of four possible scenarios results from two combined parameters: a) how much autonomy is allowed for AI, b) how much trust/distrust is given to the AI. To avoid dystopian outcomes society and politics have to understand, what is at stake and start critical and open debates.

The cognitive autonomy of machines by definition disrupts the human control power and thus challenges culture and the legal system. As non-humans decide on humans these machines are no longer (controlled) tools. But not all of AI is of that autonomous nature, AI also can be used for augmentation and for the automation of repetitive tasks. The design of "humane AI" has to understand the different types of AI concerning the ethical impact. >>>"AI ethics“ at it is discussed today is too shallow, it has to dig deeper. The issue is not only about what situation and damage is to avoid, but about which positive values should guide the increasingly powerful AI systems.<<<

KEYWORDS: ,ai ,future ,artificial intelligence ,society ,machine learning ,scenario ,ethics ,design ,ux ,transformation ,human enhancement ,,humanity ,user experience ,power ,business ,robotics ,autonomous ,augmented reality ,virtualization ,vr ,2040 ,foresight ,utopia ,2030 ,prediction ,smart home

Willi Schroll

September 30, 2021
Tweet

Transcript

  1. The Human Experience
    in the upcoming AI Society
    Willi Schroll, strategiclabs
    Memex Conference – See the Experience
    September 23, 2021 eye square, Berlin
    Bildquelle: Peter Paul Rubens, Seated Male Youth https://www.themorgan.org/drawings/item/144725
    study version
    -
    rich
    text handout
    1

    View Slide

  2. Foresight
    2

    View Slide

  3. Foresight process: From insight to strategic action
    STEEP sectors
    selected time horizon
    economy trends (specific branch)
    technology trends
    POLITICS
    ECONOMY
    ECOLOGY
    TECHNOLOGY
    SOCIETY
    Selection of e.g. 10 prioritized
    of 30+ STEEP megatrends
    prioritized key trends
    Insights about action fields,
    concrete action plans
    2
    1 3 4
    Complexity reduction: focus on
    identified key trends
    Deep Dive: granular analysis
    and evaluation; scenario
    projections, backcasting
    Transformation and transfer ->
    Elaboration on the impact for
    business goals/models; initiatives,
    roadmaps, visions
    3

    View Slide

  4. Selection of relevant mega & macro trends
    4

    View Slide

  5. Trend systematics – Impact, duration, trend breaks
    Expectable „Wild Card“
    type events
    „Wild Card“ events are low-probability, high-
    impact events that change the scenario trajectory
    „Black Swan“ events are a subset with
    the feature of unpredictability
    „Black Swan“
    type events
    5

    View Slide

  6. Technology
    6

    View Slide

  7. AI in the broad context of digital transformation
    Banking and finance
    Business
    Document management & publishing
    Education
    Entertainment/Gaming
    Healthcare
    Home/Service Robots
    Industry and manufacturing
    Life and medical sciences
    Public Safety
    Security/Cybersecurity
    Telecommunications
    Transportation/Logistics
    AI application fields
    Smart City
    Smart Gov/Admin
    Smart Health
    Smart Home
    Smart Learning & Education
    Smart Mobility/MaaS
    Smart Production/Industrie 4.0
    Smart Work
    Smart X

    Guiding visions
    Value creation of
    15+ trillion USD
    until 2030*
    „Smartisation“ of all
    domains of life is
    driven by increased
    performance of
    sensors, processors
    and networks
    Tech: 24 related innovation fields in 6 clusters
    *Source: https://www.usfunds.com/investor-library/investor-alert/ai-will-add-15-trillion-to-the-world-economy-by-2030/
    7

    View Slide

  8. Concrescence of transformational technologies
    Digital Ledger Technologies
    Next Gen Connectivity 5G/6G
    Metaverse – AR, VR, MR, IoT
    Advanced AI/ML
    Sources: AI/ML: https://www.datasciencecentral.com/profiles/blogs/machine-learning-in-one-picture; Metaverse: https://medium.com/digital-catapult/who-will-own-the-metaverse-48c2912d7aae;
    DLT: https://twitter.com/Stacks/status/1436364674572840968; 5G/6G: https://www.techtarget.com/searchnetworking/definition/6G 8

    View Slide

  9. Performance beyond human, but no „Artificial General Intelligence“
    2011: Watson (IBM) wins on
    US quiz show Jeopardy
    1997: Deep Blue (IBM)
    defeats chess champion
    Garry Kasparov
    2015: AlphaGo (DeepMind)
    learns to play complex
    games in short time. Wins
    against the best Go-player.
    2025?: Some experts assume
    that the evolving features of
    Quantum Computing might be
    a factor to get closer to AGI.
    2021: Everyday AI – Google
    Translate supports 109
    languages, translates 100+
    billion words per day
    Source: https://aiartists.org/ai-timeline-art
    MACHINE LEARNING / DEEP LEARNING NEUROMORPHIC 2.0?
    RULE-BASED AI
    9

    View Slide

  10. Societal Scenarios
    (narratives with a purpose)
    10

    View Slide

  11. Rethinking the value chain: A study on AI, humanoids and robots
    x
    x
    Regulated Autonomous
    Distrust
    Trust
    Sources: https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/09/rethinking-the-value-chain.pdf
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter (partly modified chart)
    Source: KPMG scenario study
    11

    View Slide

  12. Sc-1: The new land of milk and honey (regulated/trusted)
    x
    • Ubiquitous AI, engrained in human
    culture
    • Average working week below 20
    hours, Universal Basic Income
    • Innovation done by AI
    • Working humans: experts in creative
    thinking, management, responsible
    for determining the AI tasks
    Sources: https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/09/rethinking-the-value-chain.pdf
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter 12

    View Slide

  13. Sc-2: Thinking without limits (autonomous/trusted)
    x
    • Humans and AI work side-by-side as
    equals, driving innovations
    • Mutual assimilation: AI becoming
    more human
    • Augmentations to improve intellect
    and physical abilities
    • Robots are granted rights similar to
    humans
    13
    Sources: https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/09/rethinking-the-value-chain.pdf
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter

    View Slide

  14. Sc-3: Red button era (regulated/mistrusted)
    x
    • Humans have given AI the ability to
    manage and innovate on their own
    • But often the humans in charge of
    control don’t understand or trust the
    AIs/robots to do what is beneficial to
    everyone
    • Hitting the “emergency stop button”
    happens frequently
    • AI as a beloved enemy, one
    humanity relies upon
    • Constant struggle
    14
    Sources: https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/09/rethinking-the-value-chain.pdf
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter

    View Slide

  15. Sc-4: Powerless society (autonomous/mistrusted)
    x
    • AI has gone far beyond what
    humans can fathom (singularity)
    • We have lost control. Humans work
    at the behest of AI, but resent what
    they cannot understand
    • AI is interested in what is best for
    both parties
    • AI is most keen on working with
    human influencers who have the
    ability to sway the opinions of others
    15
    Sources: https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/09/rethinking-the-value-chain.pdf
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter

    View Slide

  16. Current Cultural Signals
    (tiny fraction)
    16

    View Slide

  17. Impact of cognitive technologies on life and work
    Image sources: Human log loss for image classification https://deepsense.ai/human-log-loss-for-image-classification; 24/7: Atlas robot, Boston Dynamics https://upload.wikimedia.org/wikipedia/commons/6/6c/Atlas_from_boston_dynamics.jpg, SDC:
    https://en.wikipedia.org/wiki/Waymo; Job replacement, prediction (2016 survey) – Oxford-Yale study: Within 50 years, robots could outperform humans at everything; https://www.aitrends.com/ai-research/within-50-years-time-robots-capable-
    outperforming-humans/, Reskilling: https://www.bbc.com/news/business-44849492; V. AI Compaion: https://replika.ai; https://roboticsandautomationnews.com/2019/01/25/top-16-autonomous-delivery-robots-ready-to-take-over-the-streets/20709;
    Therapy: https://www.wysa.io; Translation: https://www.timekettle.co; BCI in car: Mercedes AVTR concept https://newatlas.com/automotive/mercedes-avtr-concept/
    AUGMENTATION
    REPLACEMENT
    Realtime translation
    Brain-Interface in Car (planned)
    Therapy Chatbot
    Job Replacement Timeline
    Autonomous Delivery
    24/7 Worker
    „Better than human“
    Reskilling
    Virtual AI Companion
    Self-driving Car
    17

    View Slide

  18. Upcoming lifeworld with ubiquitous AI and XR technologies
    „Generation Alpha“, born 2010+
    • The announcement of
    „generations“ is not based on
    hard science, but there is a set
    of formative experiences
    triggered by the environment
    • These result in new attitudes,
    interests, goals, habits, self-
    concepts, media and purcha-
    sing bahavior
    • AI-based technologies already
    are coining Generation Alpha:
    AI toys, AI-based characters in
    gaming, AI-driven voice
    assistants, cleaning robots,
    autonomous cars
    Source: https://mccrindle.com.au/insights/blogarchive/gen-z-
    and-gen-alpha-infographic-update/
    18

    View Slide

  19. „AI Society“ as the next socio-economical stage?
    AI POLITICS
    AI ECONOMY
    AI ECOLOGY
    AI SOCIETY
    KNOWLEDGE IS
    POWER
    PERFORMANCE BOOST AI FOR GOOD
    AUGMENTED
    WORKPLACE
    DATA VALUE
    TRANSPARENCY &
    CONTROL
    RESOURCE EFFICIENCY
    VALUE CREATION
    AI ETHICS
    AUTOMATION WAVE
    AI TECHNOLOGY
    PREDICTIVE
    INTELLIGENCE
    Society is challenged
    • Autonomous decision systems pose ethical dilemmas
    • Questions of purpose and identity – e.g. status of work in life
    • Long-term consequences of choices (status of AI agents)
    19
    Image sources :
    https://www.researchgate.net/publication/327221974_Wertschopfung_neu_gedacht_Von_Humanoiden_KI's_und_Kollege_Roboter;
    https://www.technologyreview.com/2018/10/24/139313/a-global-ethics-study-aims-to-help-ai-solve-the-self-driving-trolley-problem/

    View Slide

  20. AI Politics sets Frames
    20

    View Slide

  21. From AI ethics to AI geopolitics & AI economy
    x
    21
    Image sources: UN Human Rights https://thecrimereport.org/2021/09/17/un-human-rights-commissioner-calls-for-moratorium-on-artificial-intelligence/; ; Metaverse: https://www.theverge.com/22588022/mark-zuckerberg-
    facebook-ceo-metaverse-interview; AI virtual model https://www.allkpop.com/article/2021/09/social-media-influencer-model-created-from-artificial-intelligence-lands-100-sponsorships; Putin 2017
    https://www.theverge.com/2017/9/4/16251226/russia-ai-putin-rule-the-world; Standards 2035 https://thediplomat.com/2021/09/chinas-standards-2035-project-could-result-in-a-technological-cold-war/

    View Slide

  22. Human Experience
    22

    View Slide

  23. Different AI surroundings/agents create different AI perceptions
    Fatal attachment – what could go wrong?
    Attachment to robotic creatures and virtual avatars happens due to psychological
    mechanisms.
    Language trap
    Today we use mental attributions for machines in an „as-if“ language („the car sees
    the barrier“). But the artificial systems of today do not have any „inner experience“,
    (phenomenal) mental states (Sys 0?).
    Related: Metzinger 2018: https://www.frontiersin.org/articles/10.3389/frobt.2018.00101/full
    Weizenbaum 1966: https://dl.acm.org/doi/10.1145/365153.365168
    Image source: Two elderly interact with robot pet Paro https://web-japan.org/trends/09_sci-
    tech/sci090917.html
    23

    View Slide

  24. Analytical and generative AI potential for value creation
    x
    Analytical AI techniques enable/support
    • Context adaption
    • Personalisation / targeting
    • Prediction of interests & behavior
    • Process optimization
    • Resource planing
    • …
    Generative AI techniques enable/support
    • Product ideation, design and development
    • Personalisation of content in realtime
    • Automated content creation
    • Conversational AI and rich communication
    • …
    Metaverse of AI Agents
    Avatar Sales Person
    DRIVER 1: Sim Tech Maturity
    Basic Simulation
    Rendering, Emotional &
    Conversational AI
    DRIVER 2: Live Commerce Trend
    Social Live Commerce
    Time consuming
    Advanced Simulation
    Virtual Actors &
    Influencers
    TOMORROW
    Example of a generative AI usage scenario
    24

    View Slide

  25. Autonomous agents are transcending the tool function
    x
    Image sources: Girl with robot „Pepper“: https://www.weforum.org/events/global-technology-governance-summit-2021; AI type matrix (PWC) modified, original graphics:
    https://www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizing-the-prize-report.pd (2017)
    25

    View Slide

  26. „Ethical AI“? Designing the AI Experience needs deeper concept
    x
    Image sources: Girl with robot „Pepper“: https://www.weforum.org/events/global-technology-governance-summit-2021 AI type matrix (PWC) modified, original graphics:
    https://www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizing-the-prize-report.pd (2017)
    trust
    control
    expectations
    attitudes
    emergent relationship?
    AI FRIENDLINESS
    „Ethical AI“ as factor of acceptance of a specific AI design
    • It needs depth of field into the overall contexts
    • Context may comprise open and hidden intentions,
    interests and values of stakeholders, sensitivities
    3rd
    dim
    ension
    26

    View Slide

  27. Thank you!
    slideshare.net/willi
    Linkedin • twitter.com/wschroll
    Willi Schroll – schroll [at] strategiclabs.de
    27

    View Slide