LondonFintechPodcast.jpg">LondonFintechPodcast.jpg" alt="" width="767" height="402">

This takes not just the biscuit but the whole packet. We’ve interacted with computers via switches, via punched cards, via teletypes, via audio. But Trulience have taken this to the next level – imagine a video version of Alexa that also responds to your movements (eg maintains eye contact as you move round the room. This (short of actual robots) must be nigh on the final frontier of lifelike human-machine interaction and has a near infinite number of use cases.

Truli (the avatar above) exists as a hologram or as in in-browser 2D version – amazingly in this latter case it is entirely in-browser requiring no download of software. But have a look for yourself (the clip is only 40secs and Truli keeps her eyes on the cameraman not the young lady asking the questions):

150x150.jpg" alt="" width="150" height="150">

In this episode I have the pleasure of a deep-dive into this near-miraculous new technology just about to emerge from its development into commercial application with Marek Zwiefka-Sibley who co-founded Trulience along with Ben Weekes.

Topics discussed on the show include:

• https://www.notjordanpeterson.com/ (not currently working but it was)
• we are in the world of Star Trek – we can talk to our phones and they will talk back
• Trulience have taken us one massive step further – to the world of Red Dwarf where you can talk to a visual Holly who responds to your actions and talks back
• Trulience was formally incorporated January 2019
• Trulience stands for True Likeness Science
• Marek’s career journey
• Ben and Marek’s first software company, founded in 2003 – Requestec
• their technological know-how and problem-solving in comms as a base when they pivoted from SMS to video conferencing and then today doing in-browser high bandwidth avatars
• Google open-sourcing of their WebRTC technology
• their later spinning-out of another company
• they were eventually acquired by US firm Blackboard which was the biggest online educational firm in the world
• the real-world problems that led them to think of digital humans/avatars
• the importance of having lunch (esp with Will Wynne) – Marek’s lunch in 2016 that led to proto-Trulience forming within Smart Pension, and mine last week that led me to finding out about them
• how the world has changed – when I started in FS pensions was about the most boring sector of FS – now the real nextgen technologies are created in a pension firm!!
• being able to integrate data, charts etc with an Avatar
• Smart Pension span-out Trulience and are now a major shareholder
• the commercial and creative importance of lunches – they are not “for wimps” but (provided they are with the right Übermensch) they can create new things in the world
• not hiding behind your computer
• comparison with Deep Fakes – audio, visual – similarities and differences
• SSML – speech synthesis mark-up language
• the Deep Fake production/detection arms race
• Deep Fake takes a long time to produce and polish a brilliant short video clip. But this is “just”, no matter how impressive, a form of CGI – zero interactivity with the watcher
• how to gather the data to produce avatars
• where avatars deviate from say gaming company
• detecting the users motion with client-side libraries to enable genuine two-way interaction
• capturing audio, translating into words, creating mapped output and using NTTS to crete speech
• all of this needs to be done within 200-300 milliseconds to make it “real-time enough” to not seem laggy
• “video is the final piece in the jigsaw in terms of human-machine interaction”
• robots, virtual robots, newtech is always used in gambling and pornography
• Marek’s shoutout for Ian McEwans’ book “Machines like Me”
• the use of Unreal – the engine used in Fortnite
• the connection of their “no download” systems in the past to the current “no download” Avatars
• this tech as a very rare type of tech – there is no doubt whether it works as you can see it real time!
• tech in search of an application vs tech whose application is obvious and infinite
• Trulience first customer is Smart Pension, their customers have already expressed an interest in having their own avatars – so FS is an “obvious” route
• other avenues – healthcare, educational firms, hardware companies (who want to integrate into kiosks)
• avatar receptionists
• sci-fi futures are very near now
• refining the technology – challenges and constraints
• ray-tracing developments (used for reflections of light on the avatar)
• why Trulience have not gone down the cartoon route – Amazon’s Sumerian technology in re
• differential eye-contact behaviour in men and women
• the early origins of their work in using avatars to help with loneliness in old people
• Eliza the 1960s (!!!) chatbot therapist – phenomenally ahead of its time
• Trulience’s services are designed for companies who already have, or could easily build Alexa-style chatbot services
• different personalities for different avatars
• Trulience’s commercial offerings

And much much more

Share and enjoy!