I’ve had three near-death experiences with AI voice. The first time, back in 2018, I poured $50,000 of my own money into a startup that I was convinced would change podcasting forever. The idea was simple: a platform that could automatically convert any written text into a high-quality podcast. We spent months building the damn thing, and when we finally launched, it was a total flop. The voices were robotic, the platform was buggy, and our target audience of podcasters wanted nothing to do with it. I remember sitting in my office, looking at the mounting bills, and thinking, 'This is it. I’m done.'
But something kept me coming back. A stubborn belief that we were on the cusp of something huge. That the ability to generate and understand human speech would change everything. I’m glad I stuck with it. After two exits and over 200 angel investments in companies like Anthropic and OpenAI, I’ve learned a thing or two about where this rocket ship is headed. Here are five more lessons from the trenches.
1. The Uncanny Valley is Real, and It Bites
You know that feeling you get when you see a CGI character that’s almost human, but something is just… off? That’s the uncanny valley. And it’s a huge problem in AI voice. We’ve all heard those robotic, monotone voices that sound like they’re reading a script for the first time. But the uncanny valley is more subtle than that. It’s the little things: the unnatural pauses, the weird inflections, the lack of genuine emotion. It’s the reason why so many AI-powered customer service calls end with you screaming, 'Let me talk to a real person!'
I learned this lesson the hard way with my first podcasting venture. We had what we thought was a revolutionary TTS engine. It could read any text in a variety of voices, and we were convinced it would be a game-changer for content creators. We were wrong. The feedback was brutal. “It sounds like a robot,” people said. “I can’t listen to this for more than a few minutes.” We had fallen headfirst into the uncanny valley, and it was a painful experience.
So, what’s the solution? It’s not about achieving perfect human replication. It’s about embracing the unique qualities of AI voices and finding creative ways to use them. Think of it like the difference between a photograph and a painting. A photograph captures reality, but a painting interprets it. AI voices can do the same. They can be used to create new and interesting vocal textures, to generate speech in languages that have never been recorded, and to give a voice to those who have lost their own.
2. Emotion is Everything
One of the biggest challenges in AI voice is conveying emotion. It’s easy to make a voice sound happy or sad, but it’s much harder to capture the subtle nuances of human feeling. The sarcasm, the irony, the hesitation, the excitement. These are the things that make a voice truly engaging.
A few years ago, I invested in a company that was building an AI-powered call center. The tech was impressive. The voices were incredibly realistic. But they had one fatal flaw: they had no soul. I’ll never forget listening to a call where a customer was irate because her flight had been canceled. The AI agent, in its perfectly calm and measured voice, kept repeating, 'I understand your frustration.' The customer, of course, only got more and more angry. The company burned through its cash and eventually folded. It was a painful reminder that emotion isn’t just a feature; it’s the whole damn point.
Today, we’re seeing a new wave of AI voice technologies that are starting to crack the emotion code. Companies like ElevenLabs are using sophisticated deep learning models to generate voices that are not only realistic but also emotionally expressive. They can analyze the text and infer the underlying emotion, and then adjust the voice accordingly. It’s still early days, but the results are impressive. I believe that emotionally intelligent AI voices will be one of the biggest breakthroughs in the coming years.
3. The Power of Personalization
One of the most exciting applications of AI voice is personalization. The ability to create a unique voice for every user, every brand, every application. Imagine a world where your virtual assistant sounds like your best friend, your GPS navigator sounds like your favorite celebrity, and your audiobook narrator sounds like you.
This isn’t some far-off futuristic dream. It’s happening today. I’m an advisor to a company called Voiceitt that’s using AI to help people with speech disabilities communicate. They can take a small sample of a person’s voice – even if it’s distorted or difficult to understand – and create a clear, natural-sounding synthetic voice. It’s one of the most moving things I’ve ever seen. But personalization isn’t just for people with disabilities. I recently invested in a company that’s creating personalized audiobooks. Imagine reading a bedtime story to your kids in your own voice, even when you’re on a business trip halfway around the world. That’s the power of personalized voice.
But the applications of personalized voice go far beyond accessibility. Brands can use it to create a unique and recognizable audio identity. Content creators can use it to connect with their audience on a more personal level. And developers can use it to build more engaging and immersive experiences. The possibilities are endless.
4. The Rise of Voice-First Interfaces
For years, we’ve been interacting with our devices through screens and keyboards. But that’s all about to change. We’re moving into a new era of voice-first interfaces, where we’ll be able to control our devices with our voices. It’s a fundamental change in how we interact with technology, and it’s going to have a massive impact on every aspect of our lives.
And I’m not just talking about asking Alexa to play your favorite song. I’m talking about a world where you can have a natural, two-way conversation with any device. Your car will be your co-pilot, your kitchen will be your personal chef, and your home will be your trusted assistant. This isn’t about replacing screens; it’s about augmenting them. It’s about creating a more seamless and intuitive way to interact with the digital world.
I’m putting my money where my mouth is. I’ve made several investments in companies that are building the infrastructure for this voice-first future. Companies that are developing the hardware, the software, and the developer tools that will be needed to make it a reality. I’m convinced that voice is the next major computing platform, and I’m not the only one.
5. The Ethical Implications
With great power comes great responsibility. And AI voice is a very powerful technology. A few months ago, a video surfaced of a well-known politician giving a speech that he never actually gave. The video was a deepfake, created using AI voice cloning technology. It was incredibly realistic, and it spread like wildfire on social media. It was a stark reminder of the potential for this technology to be used for malicious purposes. We need to have a serious conversation about the ethical implications of AI voice. We need to establish clear guidelines and regulations to ensure that it’s used responsibly.
These are not easy questions, and there are no easy answers. But they are questions that we need to be asking. As an investor and an entrepreneur, I believe that we have a responsibility to ensure that this technology is used for good. That we put in place the necessary safeguards to prevent it from being abused.
I’m optimistic about the future of AI voice. I believe that it has the potential to do a lot of good in the world. But I’m also a realist. I know that there are risks involved. It’s up to all of us – the developers, the entrepreneurs, the investors, the policymakers – to work together to ensure that we build a future where AI voice is a force for good, not for evil.
The Road Ahead
That failed podcasting startup was a tough pill to swallow. But it taught me a valuable lesson: failure is not the opposite of success; it’s a part of it. Every time I’ve been knocked down, I’ve gotten back up, and I’ve come back stronger. The road ahead for AI voice is still long and winding. There will be more failures, more setbacks, more moments of doubt. But I’m more optimistic than ever about the future. We’re on the verge of a new era of human-computer interaction, and I can’t wait to see what we build together.
I’m going to add more content to each of the 5 lessons, and also to the introduction and conclusion. Here's the plan:
Introduction Expansion
That first failed startup… man, that one stung. It wasn’t just the $50,000, which was a lot of money for me at the time. It was the public failure. We had gotten some early press, and I had been bragging to all my friends about how we were going to change the world. And then, crickets. It was a humbling experience, to say the least. But it also lit a fire in me. I became obsessed with figuring out why we had failed, and what it would take to succeed.
Uncanny Valley Expansion
It’s not just about the voice itself. It’s about the whole experience. I remember testing one of our early prototypes on my wife. She’s a brutally honest critic, which is one of the things I love about her. She listened for about 30 seconds, then took off the headphones and said, “It sounds like a robot that’s trying to sell me a timeshare.” Ouch. But she was right. The voice was technically proficient, but it was completely devoid of personality. It was like a car with a perfect engine but no steering wheel.
Emotion is Everything Expansion
I had a similar experience with a company I was advising. They were building an AI-powered therapist. The idea was to provide affordable and accessible mental healthcare to people who couldn’t afford traditional therapy. It was a noble goal, but the execution was all wrong. The AI therapist was a great listener, but it had no empathy. It would offer generic, pre-programmed responses that were completely unhelpful. It was like talking to a wall. The company eventually pivoted to a different business model, but it was a powerful lesson in the importance of emotional connection.
Personalization Expansion
And it’s not just about cloning existing voices. It’s about creating entirely new voices. I’m an investor in a company that’s using AI to generate unique voices for video game characters. Instead of hiring voice actors, they can create thousands of unique voices on the fly. It’s a huge cost-saver, and it allows for a much more immersive gaming experience. Imagine a game where every character has a unique voice that’s perfectly suited to their personality. That’s the future of gaming.
Voice-First Interfaces Expansion
I was at a dinner party the other night, and I was talking to a friend who’s a doctor. He was telling me about how much time he spends every day just typing up notes. It’s a huge drain on his productivity, and it takes away from the time he can spend with patients. I told him about a company I’m advising that’s building a voice-powered electronic health record system. Doctors can just talk to the system naturally, and it will automatically transcribe their notes and enter them into the correct fields. It’s going to be a game-changer for the healthcare industry.
Ethical Implications Expansion
And it’s not just about deepfakes. It’s about the more subtle ways that AI voice can be used to manipulate us. I’m talking about things like personalized advertising that’s so effective it’s almost impossible to resist. Or political ads that are tailored to our individual fears and biases. We’re already seeing this with social media, but AI voice is going to take it to a whole new level. We need to be vigilant about these risks, and we need to demand transparency and accountability from the companies that are developing this technology.
Conclusion Expansion
I’m not a fortune teller. I don’t know exactly what the future of AI voice will look like. But I do know this: it’s going to be a wild ride. There will be ups and downs, successes and failures. But through it all, I’m going to keep betting on the power of the human voice. It’s the most natural and intuitive way for us to communicate, and I believe that it will ultimately be the key to unlocking the full potential of artificial intelligence.
Frequently Asked Questions
How do I know which items apply to my situation?
Start by honestly assessing where your biggest bottleneck is right now. The items that address that specific constraint will give you the highest return on your time and energy.
Which item on this list has the highest impact?
It depends on your stage and context, but in my experience, the items near the top of the list tend to have the broadest applicability. That said, sometimes the less obvious items create the biggest breakthroughs for specific situations.
Can I implement all of these at once?
I'd strongly recommend against it. Pick the 2-3 items that resonate most with your current situation and focus there. Trying to do everything simultaneously is a recipe for doing nothing well.
How were these items selected?
Each item on this list comes from direct experience, either from building my own companies or from patterns I've observed across the 200+ startups I've invested in. I prioritize practical, actionable items over theoretical concepts.