Context :
Ø Behind Hume’s conversational AI with emotional intelligence … Eco Times… 01 April 2024
( Himanshi.lohchab@timesgroup.com )
Extract :
Artificial intelligence can now understand human emotions, pull-off sarcasm, and even express anger. New York-based startup Hume AI (https://www.hume.ai / hello@hume.ai )
last week launched the first voice AI with emotional intelligence which can generate conversations for emotional well-being of its users.
Founded in 2021 by Alan Cowen, a former researcher by Google DeepMind, the startup also raised $50 million in Series-B funding from EQT Group, Union Square Ventures, Nat Friedman, Daniel Gross, Northwell Holdings, Comcast Ventures, LG Technology Ventures, and Metaplanet days after the launch.
What is Hume AI?
Hume’s voice interface is powered by its empathic large language model (eLLM) which emphasises on tones of voice behind words to understand different emotions.
It can further emulate similar tones across 23 different emotions such as admiration, adoration, frustration etc, to generate human-like conversations.
The conversational AI chatbot is trained on data from millions of human conversations across the world to voice tonality, human reflexes and feelings. These responses are further optimised in real-time depending on user’s emotional state.
How is it useful?
While expressive AI chatbots in areas such as virtual dating have been around, Hume’s product is gaining accolades for its probable uses in robotics, healthcare, wellness etc.
Early predictions by some AI researchers show that AI assistants powered by Hume’s eLLM could not only make conversations but also help in daily tasks.
“Imagine an AI assistant that understands your frustrations or joys, a customer support agent that can empathize with your complaints, or even a virtual therapist capable of offering genuine emotional support,” according to a post on X.
Cowen in a LinkedIn post said, "Speech is four times faster than typing; frees up the eyes and hands; and carries more information in its tune, rhythm, and timbre.”
“That's why we built the first AI with emotional intelligence to understand the voice beyond words. Based on your voice, it can better predict when to speak, what to say, and how to say it."
Hume AI is preparing to release the platform APIs to developers next month in beta mode to integrate with various applications.
It can also integrate with other large language models such as GPT and Claude to add flexibility depending on enterprise use-case.
Besides empathetic feature, the voice assistant also offers transcription and text-to-speech capabilities.
My Take :
Background :
Ø In the US, the NIMH reports that 1 in 5 adults experiences mental illness in a given year https://www.nami.org/mhstats.
Ø This initiative reports a significant burden of mental disorders in India, affecting about 10% of the population https://nhm.gov.in/index1.php?lang=1&level=2&sublinkid=1043&lid=359.
Sure, it has took almost 8 years since I envisaged its arrival but that VIRTUAL THERAPIST has finally arrived !
Here is how I envisaged it :
Ø Share - Your - Soul / Outsourcing Unlimited .. ………..24 July 2016
Extract :
Here is an outline of my suggestion , re how young / educated / unemployed Indians can offer such service :
VEHICLE
This strictly " Online " service will have a platform called , www...COUCH...com ( supported by a Mobile App )
USERS
There will be two kinds of users who will register on this site , viz:
* " Talkers " , who want to engage someone who will listen to them / sympathize with them
* " Listeners " , who will listen patiently / ask occasional question / offer advice - sympathy - empathy
# REGISTRATION FORM
For both type of users , the Registration Form will require to submit following details :
* Personal Details ( Name / DOB / Gender / Nationality / Bank Account No / Photo / Short Video etc )
* Family Details ( Who are members of immediate family ? )
* Contact Details ( Address / Mobile No / Email ID / Skype - FaceTime ID / WatsApp..etc )
* Social Media Footprint ( Facebook / LinkedIn / Twitter : No of Friends - Contacts - Connections - Followers )
* Cultural Exposure ( Countries visited / lived-in , with stay-periods / Foreign friends )
* Educational Details ( Degrees / Colleges ) . Listeners having degree in Psychology will get ranked higher !
* Language Details ( Languages spoken / fluently - reasonably well )
* Experience Details ( Where worked / for how long ) . Retired / Worldly Wise , Listeners get ranked higher !
* Availability Details ( Available from - to / GMT- Local time )
Based on the completeness of the Registration Form Details , a software will rate and rank the Listeners , which will be visible to the Talkers
There will be facility to update / edit the data submitted
Upon registration , users will be assigned USER ID and PASSWORD
In addition , Listener will be assigned a unique COUCH / INTERVIEW-CABIN number
# SEARCHING DATABASE OF LISTENERS
Talkers will be able to search the database of the registered Listeners , except for their " Contact Details "
Talkers can , then select / shortlist , a few listeners of their preference
# SERVICE CHARGE
Talkers will pay $ 2 per hour to the portal , which will credit this amount to the Bank Account of the
Concerned Listener , after deducting 10 % as its own commission
# PAYMENT MECHANISM
Using online payment gateway , Talkers will deposit a minimum of $ 20 , on the portal as PRE-PAID amount
As Talker continues using the service , credit balance will get displayed ( in $ and in " Hours " terms )
There will be facility for online ( or through Mobile ) re-charging of the account
# PROCESS
A Listener can login anytime and occupy his own virtual COUCH / CABIN ( " I am now available for
listening " )
As soon as he does , a GREEN light will shine on the CABIN , showing the online availability of the concerned
Listener .
This green light will tell the Talkers : " Welcome ! I am ready to listen "
The light will turn RED , as soon as a Talker walks into Listener's CABIN ( " I am engaged right now " )
Any time a Talker logs in , he will find if any Listeners ( that he had previously shortlisted ) are available
online
If he finds one , he simply CLICKS on the CABIN icon and enters that VIRTUAL cabin !
Simultaneously , both the Talker and the Listener , turn on their Skype ( on Mobile or Tablet ) to start the
talk
Remember , Skype ID of neither the Talker , nor the Listener , is ever visible to each other !
All conversation / transaction , can ONLY take place through www...COUCH...com ( no bypassing ! )
The entire conversation will get recorded ( Video + Audio ) and can be downloaded by the Talker ( but not
By the Listener ) , if he so desires
Portal will be obliged to make this recording available to a Court of Law , in case of any litigation
Portal will carry a WARNING that it reserves the right to remove any Talker or a Listener , if it finds that its
service is being misused / abused ( will need defining , in detail )
# REPUTATION SYSTEM
At the end of each " talk / conversation " , Talker will be obliged to " Rate " the concerned Listener
on a 5 point scale ( Excellent > Horrible ) .
Cumulative / Average " Rating " will be prominently displayed for guidance of all Talkers .
Of course , a Listener can see his own rating as soon as he logs in
At some future date , it should be possible ( through appropriate software ) , to introduce following variations
In pricing of the service ( ie ; Hourly Rates ) :
* Surge Pricing ( depending upon the DEMAND of any given Listener ) ie: No of Talkers waiting for a given
Listener at a given point of time
* Reputation Pricing , based on points accumulated by a given Listener from all past ratings
# USAGE HISTORY
For each user ( Listener or Talker ) , there will be a Usage History page of all the past transactions / talks ;
As also Credit Balance ( for the Talker ) and the Earnings ( for the Listener )
# PRIVACY
The portal will NOT reveal any info / data ( including Audio-Video recording ) of any user to anyone else.
However , portal will reserve the rights to subject those Audio recordings ( but not Video recordings ) to
an Artificial Intelligence ( AI ) software , which can , over a course of time , come up with a SOFTWARE
ROBOT that can take over the role of the HUMAN listeners ! If you have any doubts , ask Ray Kurzweil !
When that happens , this portal may morph into a PPO ( Psychology Process Outsourcing ) !
The portal will also reserve the rights to use the Audio recordings for offering Voice-to-Voice language
translation mobile app for the benefit of world-travellers
# PROMOTING THE SERVICE
To an extent , the portal may affect the jobs of local Psychologists / Psycho-Analysts who offer low level
consulting in any country. They will be in danger of being " Bangalored " ! So , it is bound to face resistance
from those vested / threatened interests !
But foreign Hospitals / Educational Institutions / NGOs / Medical Colleges , etc could be targeted for
promoting
# BUSINESS MODEL
Business Model will be in the nature of " Sharing Economy " , where those owning / possessing " Idle /
Spare / Under-utilized " assets / resources , will offer the same to those in need / when in need , for a price
Eg:
Millions of private car-owners use their cars for ( may be ) two hours per day . Uber aggregates this spare
capacity and makes it available to travelers who do not own ( or wish to own ) their own cars
Both parties benefit . Economy also benefits by fuller utilization of the spare/surplus capacities of millions of
assets
All in all , I think this is a great opportunity for some Indian Start Up to seize
Dear Alan Cowen :
Congratulations for your innovation which is bound to REVOLUTIONIZE , conversational AI
I would love to integrate it to enable my VIRTUAL AVATAR ( www.HemenParekh.ai ) to answer 51,400 questions with appropriate EMOTIONS ( - in all of 26 languages ? )
With regards,
Hemen Parekh
www.HemenParekh.ai / 01 April 2024
No comments:
Post a Comment