Stuff

    Subscribe to our newsletter

    What's Hot
    Xiaomi 12

    At least one of Xiaomi’s flagship 12-series devices coming to South Africa on 8 July

    July 1, 2022
    Vodacom

    Vodacom spends R460 million on expanding coverage in Limpopo

    July 1, 2022
    Top Five Tablets

    Stuff’s Top Five Tablets (at the moment)

    July 1, 2022
    Facebook Twitter Instagram YouTube SoundCloud
    Trending
    • At least one of Xiaomi’s flagship 12-series devices coming to South Africa on 8 July
    • Vodacom spends R460 million on expanding coverage in Limpopo
    • Stuff’s Top Five Tablets (at the moment)
    • Google’s South African domain went down briefly. But it’s back up, crisis averted
    • Niantic’s Campfire is a Pokémon Go social media app (with more to follow)
    • 2022 Ford EcoSport Titanium review – Outdone by its cheaper siblings
    • How recycling and reforestation are advanced with 4IR Tech
    • Samsung reportedly working on a cheaper range of folding smartphones
    Facebook Twitter Instagram YouTube
    Stuff Stuff
    • News
      • App News
      • Business News
      • Camera News
      • Gaming News
      • Headphone News
      • Industry News
      • Internet News
      • Laptops News
      • Motoring News
      • Other Tech News
      • Phone News
      • Tablet News
      • Technology News
      • TV News
      • Wearables News
    • Reviews
      • Camera Reviews
      • Car Reviews
      • Featured Reviews
      • Game Reviews
      • Headphone Reviews
      • Laptop Reviews
      • Other Tech Reviews
      • Phone Reviews
      • Tablet Reviews
      • Wearables Reviews
    • Columns
    • Stuff Guides
    • Podcasts & Videos
      • Videos
      • Stuffed
      • Stuffing Around
      • Tech Byte
      • T2S2
    • Win
    • Subscribe
      • Print
      • Digital
        • Google Play
        • iTunes
        • Download
        • Zinio
    • Stuff Shop
      • Shop Now
      • My Account
      • Downloads
    • Contact Us
      • Get In Touch
      • Advertise
    0 Shopping Cart
    Stuff
    Home » News » Industry News » AI: there’s a reason it’s so bad at conversation
    Industry News

    AI: there’s a reason it’s so bad at conversation

    The ConversationBy The ConversationSeptember 17, 2018Updated:October 1, 2021No Comments6 Mins Read
    Share
    Facebook Twitter LinkedIn Pinterest Email

    “I’m sorry, I didn’t quite get that.” Anyone who has ever tried to have a deeper conversation with a virtual assistant like Siri knows how frustrating it can be. That’s despite the fact that AI systems like it are increasingly pushing into our lives, with new success stories on an almost daily basis. Not only do AIs now help radiologists detect tumours, they can act as cat repellent and even detect signals of potential alien technology from space.

    But when it comes to fundamental human abilities, like having a good chat, AI falls short. It simply cannot provide the humour, warmth and the ability to build coherent and personal rapport that is crucial in human conversations. But why is that and will it ever get there?

    Chatbots have actually come a long way since their early beginnings, with MIT’s Eliza in the 1960s. Eliza was based on a set of carefully crafted rules that would give the impression of being an active listener and simulating a session with a psychotherapist.




    Eliza.

    Systems like Eliza were good at giving a sophisticated first impression but were easily found out after a few conversational turns. Such systems were built on efforts to collate as much world knowledge as possible, and then formalise it into concepts and how those relate to each other. Concepts and relations were further built into grammar and lexicons that would help analyse and generate natural language from intermediate logical representations. For example, world knowledge may contain facts such as “chocolate is edible” and “rock is not edible”.

    Learning from data

    Today’s conversational AI systems are different in that they target open domain conversation – there is no limit to the number of topics, questions or instructions a human can ask. This is mainly achieved by completely avoiding any type of intermediate representation or explicit knowledge engineering. In other words, the success of current conversational AI is based on the premise that it knows and understands nothing of the world.

    The basic deep learning model underlying most current work in natural language processing is called a recurrent neural network, whereby a model predicts an output sequence of words based on an input sequence of words by means of a probability function that can be deduced from data. Given the user input “How are you?” the model can determine that a statistically frequent response is “I am fine.”

    The power of these models lies partially in its simplicity – by avoiding intermediate representations, more data will typically lead to better models and better outputs. Learning for an AI is very similar to how we learn: digest a very large training data set and compare it with known but unseen data (test set). Based on how well the AI performs against the test set, the AI’s predictive model is then adjusted to get better results before the test is repeated.

    But how do you determine how good it is? You can look at the grammar of utterances, how “human like” they sound, or the coherence of a contribution in a sequence of conversational turns. The quality of outputs can also be determined as a subjective assessment of how closely they meet expectations. MIT’s DeepDrumpf is a good example – an AI system trained using data from Donald Trump’s Twitter account and which uncannily sounds just like him, commenting on a number of topics such as healthcare, women, or immigration.

    However, problems start when models receive “wrong” inputs. Microsoft’s Tay was an attempt to build a conversational AI that would gradually “improve” and become more human-like by having conversations on Twitter. Tay infamously turned from a philanthropist into a political bully with an incoherent and extremist world view within 24 hours of deployment. It was soon taken offline.

    As machines learn from us, they also take on our flaws – our ideologies, moods and political views. But unlike us, they don’t learn to control or evaluate them – they only map an input sequence to an output sequence, without any filter or moral compass.




    This has, however, also been portrayed as an advantage. Some argue that the recent successes of IBM’s Project debater, an AI that can build “compelling evidence-based arguments” about any given topic, is down to its lack of bias and emotional influence. To do this, it looks up data in a large collection of documents and pulls out information to express the opposite view to the person it is debating with.

    Next steps

    But even if more data can help AI learn to say more relevant things, will it ever really sound human? Emotions are essential in human conversation. Recognising sadness or happiness in another person’s voice or even text message is incredibly important when tailoring our own response or making a judgement about a situation. We typically have to read between the lines.




    Conversational AIs are essentially psychopaths, with no feelings or empathy. This becomes painfully clear when we are screaming our customer number down the phone for the seventh time, in the hope that the system will recognise our agony and put us through to a human customer service representative.

    Similarly, conversational AIs usually don’t understand humour or sarcasm, which most of us consider crucial to a good chat. Although individual programs designed to teach AI to spot sarcastic comments among a series of sentences have had some success, nobody has managed to integrate this skill into an actual conversational AI yet.

    Clearly the next step for conversational AIs is integrating this and other such “human” functions. Unfortunately, we don’t yet have the techniques available to successfully do this. And even if we did, the issue remains that the more we try to build into a system, the more processing power it will require. So it may be some time before we have the types of computers available that will make this possible.

    AI systems clearly still lack a deeper understanding of the meaning of words, the political views they represent, the emotions conveyed and the potential impact of words. This puts them a long time away from actually sounding human. And it may be even longer before they become social companions that truly understand us and can have a conversation in the human sense of the word.

    • Nina Dethlefs is Lecturer of Computational Science, University of Hull
    • Annika Schoene is PhD Candidate, Computer Science, University of Hull
    • David Benoit is Senior Lecturer in Molecular Physics and Astrochemistry, University of Hull
    • This article first appeared on The Conversation

    AI artificial intelligence conversation robots speech The Conversation
    Share. Facebook Twitter Pinterest LinkedIn WhatsApp Reddit Tumblr Email
    The Conversation

      Related Posts

      Xiaomi 12

      At least one of Xiaomi’s flagship 12-series devices coming to South Africa on 8 July

      July 1, 2022
      Vodacom

      Vodacom spends R460 million on expanding coverage in Limpopo

      July 1, 2022
      Top Five Tablets

      Stuff’s Top Five Tablets (at the moment)

      July 1, 2022

      Leave A Reply Cancel Reply

      In The Mag
      Stuff June-July 2022 Latest Issue

      In This Issue – The Outdoors (June-July 2022) Issue

      By Brett VenterMay 30, 20221

      Once again, we are asking you to check out a new issue of Stuff Magazine.…

      2021 Wish List
      wish list Stuff Wish List 2021

      Stuff Wish List: for the tech impaired

      By Duncan PikeDecember 22, 20210

      Are you from the time before being glued to a smartphone was considered normal? Here’s…

      Wishlist DIY Stuff tech

      Stuff Wish List: for the DIY Diehard

      December 21, 2021
      Wish List Gearhead

      Stuff Wish List: For the petrol-soaked gearhead

      December 20, 2021
      outsiders

      Stuff Wish List: for the Outsiders

      December 17, 2021

      Latest Video

      Sonos

      SONOS Roam SL unboxing by Toby Shapshak

      March 30, 2022
      Mini Cooper

      The Mini Cooper SE Electric with Toby Shapshak

      March 18, 2022
      MSI Crosshair 15 Rainbox Six Extraction Edition unboxing

      MSI Crosshair 15 Rainbox Six Extraction Edition unboxing

      March 16, 2022
      Samsung Galaxy S22 Ultra Unboxing

      Samsung Galaxy S22 Ultra unboxing with Toby Shapshak

      March 16, 2022
      Contact

      South Africa's Consumer Tech News Hub

      General: [email protected]
      Subscriptions: [email protected] or 087 353 1291
      Editorial: 072 735 2614
      Sales: 083 375 2418

      Facebook Twitter Instagram YouTube SoundCloud

      Subscribe to Updates

      • Terms and Conditions
      • Privacy & POPI
      • My account
      © 2022 Stuff Group. Designed by Chronon.

      Type above and press Enter to search. Press Esc to cancel.