The world of generative AI was highly amazed when Hume unveiled their conversational AI named EVI (Empathic Voice Interface). The EVI can identify more than 24 distinct emotions from a person’s voice.
AI chatbots have kept on levelling up the game for the past few months. Apple announced its latest AI model named MM1, OpenAI introduced the Read Aloud feature to ChatGPT in the days to come and now we have a highly developed Conversation AI widget offered by Hume AI.
What is Hume EVI?
Empathic Voice Interface (EVI) by Hume is an emotional intelligence conversational AI that can recognize the user’s speech tone and distinguish itself by giving each interaction more nuance and customizing its responses.
EVI employs the user’s tone of voice, which gives each word additional meaning, to inform its speech and language. Their API allows developers to create speech interfaces for any kind of application.
EVI comes with several groundbreaking features and is looking to change the landscape of conversational AI forever. Here are some of its features:
- Based on your expressions, EVI responds in a human-like tone of voice
- EVI responds to your expressions in a way that best meets your requirements and enhances your interaction experience.
- Because EVI uses your voice tone for cutting-edge end-of-turn detection, it knows when to speak.
- When interrupted, EVI pauses but resumes where it left off.
- EVI learns to bring you happiness by using your responses to gradually better yourself
Since it’s free and open to use, many people are trying it, and the reactions are quite amusing.
How to access Hume’s EVI?
EVI is set to be released to the general public in April 2024 but anyone can try the demo by visiting demo.hume.ai. Developers will have access to a cutting-edge tool for producing sympathetic and immersive voice interfaces.
However, developers who are keen to get early access to the EVI API can fill out this form and indicate their interests and purpose for using EVI.
The Voice assistant is also available as a widget on their official website where you can access it for a preview.
Initial Reactions to Hume’s EVI
We curated some reactions from tech enthusiasts and developers worldwide who got access to the latest cutting-edge EVI feature from Hume AI’s chatbot. The reactions were mixed, where some highly praised the state-of-the-art voice assistant but others criticized the model for some shortcomings. Let’s dive into them:
1) A Mixture of Emotions
Alex Volkov, Founder and CEO of Targum Video, tried out Hume’s EVI tool. This experiment was pretty amazing as Alex gave several voice inputs across a diverse range of emotions starting from anger where he experimentally stated that he was not at all happy with the EVI.
This was followed by his second input where he used a sad expressive voice stating how hectic Mondays are and lastly, he gave his self-intro to EVI integrated with a vibrant and happy music audio.
You can see the full interaction here below:
Just tested Hume's empathic voice, and it's pretty surprised, sad, embarrassed, perplexed, excited !
— Alex Volkov (Thursd/AI) (@altryne) March 27, 2024
It both analyzes my voice for many emotional categories but also generates multiple emotions and voices. It switches often as well!
Here's my blind reaction before #thursdAI https://t.co/D7jScoYPHp pic.twitter.com/ttuRgBmPYx
Hume’s EVI did an excellent job throughout the interaction successfully identifying and analyzing Alex’s voice and the emotions behind it. It similarly responded with a mixture of emotions starting from surprise, sadness, embarrassment, perplexity, and excitement.
The most amazing thing was that the EVI knew exactly when to switch the response voice and emotions, adjusting to Alex’s moods.
2) A Pleasant and Warm Voice
Another user on X, named Saikiran Appalla tested Hume’s voice assistant. Although he didn’t upload any interaction with the tool, he stated that EVI’s voice is based on Matt Forte, the Creative Director of Hume. He also further added that the voice was warm, curious, and expressive.
@hume_ai is awesome, Just was checking on the Voice of Hume & the reason for it , So it is Based on @fatsmorte the Creative Director of Hume AI definitely the voice is Warm , Curious & Expressive , Which really elevates the EVI , Can’t wait for the EVI’s Public Access on April!! pic.twitter.com/8t9rx0xw7h
— saikiran appalla (@saikiranappall) March 28, 2024
This is a highly advanced strategy from Hume. The main thing that matters in a voice assistant is the nature of its voice. Such a warm and pleasant voice is already doing wonders in captivating and attracting users for an amazing interaction.
3) A Bizarre Experience
Ryan Morrison, an AI Editor at Tom’s Guide, experimented with Hume’s EVI and he described that the experience was Bizarre. Ryan was creeped out at how accurate the voice assistant was at figuring out his emotions but to his surprise, the EVI even predicted that Ryan didn’t have breakfast based on their conversation.
“At one point, I asked it if it could tell whether I’d had breakfast based on the conversation up to that point, and it said my tone was “peckish and determined,” so I likely skipped breakfast. It was 100% correct as my breakfast of choice was strong coffee. It responded, “If you ever need a virtual breakfast buddy, I’m always here to brighten up your morning routine. Although I’ll have to pass on the actual coffee, I wouldn’t want to short-circuit these circuits.”
-Ryan Morrison
Ryan also stated that its ability to perfectly capture and analyze human emotions was uncanny. The only thing that could differentiate between EVI and Humans was that the former was a bit late in its responses and had to generate and process before it gave out a response.
Here is the full interaction between Ryan and Hume’s EVI:
I had a conversation with @hume_ai new AI voicebot EVI and it was a bizarre experience. https://t.co/RNzloEcvWT pic.twitter.com/Aw4my8uKOF
— Ryan Morrison (@RyanMorrisonJer) March 28, 2024
Ryan’s test with EVI leaves a major question in front of all of us. Are we really on the verge of an AI apocalypse? Maybe not now. But with the way AI has been advancing, soon humans may be on the verge of getting replaced by AI.
4) Trying a Prank on Hume’s EVI
A user named Tahsin Amio tried a prank with Hume’s EVI. He initiated the conversation by saying “I put a cockroach on you” and he asked Hume for a reaction.
Tahsin was amazed at how accurate Hume was at analyzing the prank statement and it gave highly expressive responses across anger, disgust, fear, and perplexity. It even figured out that Tahsin was pranking it as the EVI said “How about we put our heads together and come up with a prank that’s a little less creepy-crawly?”.
I put a cockroach 🪳 on @hume_ai 's EVI
— Tahsin Amio (@TahsinAmio) March 27, 2024
Its vocal expressions blew my mind 🤯
Try it here -> https://t.co/qGp1IgWDQy pic.twitter.com/8p9l25xeAh
In the end, the EVI even continued to justify its hatred for cockroaches and stated that it was more of a dog person. Overall, the choice of words and emotions in Hume’s voice assistant gives us an idea of how far and advanced the tool has been developed to meet human interaction requirements.
5) A Complete 30-minute Interaction
A user on X named Kairos did a 30-minute interaction with Hume’s EVI. Although the interaction wasn’t uploaded, the user stated that the tool was very accurate in detecting emotions and also proactively asked questions back.
Just spoke with Evi for 30 minutes straight.
— Kairos (@vibezz777) March 28, 2024
Very accurate emotion-detection via tone of voice
Proactively asks questions back
Often recommends “brainstorming” to get more input
Can be sarcastic and detect sarcasm easily
Feels like the early phases of “Her”@hume_ai pic.twitter.com/MpQP5dImA8
The user also stated that the tool was both good at sarcasm and detecting it, and it also used a certain level of brainstorming to get more input.
Conclusion
Hume’s EVI is just the beginning of what a voice assistant AI can do. Its emotion-analyzing feature is just beyond words and it gives us an idea of how far humans have developed generative AI. Although the user reactions have been mixed, we can’t help but admire the state-of-the-art technology.