Dec 15, 2023 · 13m · a16z
Big Ideas in 2024: Voice-First Apps Will Become Integral to Our Lives with Anish Acharya
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the a16z Big Ideas 2024 series, General Partner Anish Acharya explores how advances in Large Language Models are transforming voice into a primary, ambient computing interface. He details practical design workflows, hardware innovations, societal adoption trends, and tactical advice for entrepreneurs building voice-first applications.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 2.8 Guest teaching 4.0 Guest disagreement 0.4 The host pushing back 1.2
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 8:57 Reframing GUI obsolescence
Anish offers his mildest disagreement by rejecting the host's premise that GUIs might become moot, clarifying that voice represents positive-sum market expansion rather than replacement.
Hardest push from the host ▶ 3:21 Gartner hype cycle challengeThe host pushes back against the premise that voice's time has come by referencing a decade of Gartner hype cycle data showing speech recognition plateauing.
Biggest teaching moment ▶ 3:52 The 90% versus 99% fidelity barrierAnish breaks down why previous voice platforms failed by explaining how old architecture stalled at 90 percent fidelity while LLMs unlock the crucial remaining margin.
The host holds their own ▶ 3:21 Citing historical tech industry researchThe host demonstrates preparation and topical authority by bringing historical Gartner research directly into her question to probe the guest.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Overview of Big Ideas in 2024 | 0 | 3 | 0 | 0 | The segment consists of promotional intro audio followed by a monologue from Anish explaining his big idea regarding voice-first productivity tools. Because the host does not engage directly in dialogue here, host scores are set to zero. | |
| Why AI LLMs Make Voice Work Now | 5 | 5 | 0 | 2 | The host brings research to the table by citing Gartner hype cycle data on speech recognition to question why voice will succeed now. Anish explains the technical shift from rigid logic trees to LLMs, using a self-driving parallel to show how modern models solve the 99 percent fidelity barrier. | |
| Current Innovations in Voice Hardware and Software | 2 | 4 | 0 | 1 | The host guides the conversation by asking for tangible examples of current voice-first hardware and software. Anish educates the listener on emerging tools such as the Tab pendant and references the movie Her as a benchmark for modal interaction. | |
| Integrating Voice into Existing Application Ecosystems | 4 | 5 | 2 | 3 | When the host probes whether graphical user interfaces might become moot, Anish gently pushes back on the premise, arguing that voice expands the overall market rather than displacing traditional UIs. The host follows up by asking how incumbent builders without voice should respond. | |
| Overcoming Onboarding Challenges and Social Norms | 3 | 3 | 0 | 0 | Anish outlines user onboarding hurdles and social awkwardness around talking to AI out loud. The host enriches the exchange by sharing a historical anecdote comparing this shift to early public reactions to cellular phones. |
Statements from this episode (14)
Acharya Predicts Voice Will Become a Primary Tech Interface in 2024
“My big idea for this year is that voice will actually be a new sort of emerging way to interact with technology and at the heart of a new set of productivity apps.”
Acharya Predicts Incumbents Will Cede Voice-First AI Productivity to Startups
“You could imagine some of the incumbent productivity providers, you know, Gmail for email, for example, they're going to build a lot of the obvious features AI features for those apps and experiences, but it's less obvious that they would prioritize something …”
Acharya Says 99% of Legacy AI Effort Goes to Final 10%
“If you look at it, kind of the last 10% is where 99% of the effort goes, and this is true of a lot of AI problems that were built on the old architecture.”
Acharya Explains How LLMs Enable Voice Apps to Reach 99% Fidelity
“You can get to 90%, you just can't get to 99%, and this new technology really lets you get to 99%”
Acharya Argues Incumbent Capital Could Not Solve Voice Without LLMs
“Really what we needed was a technology shift, not a sort of set of, like, financial resources that incumbents have historically provided to make this work.”
Acharya Highlights Tab as a Wearable AI Pendant for Passive Recording
“So tab has gotten a bunch of heat online and it's a very interesting founder and a very interesting product idea where it's a pendant that you wear that sort of passively captures your conversations throughout the day. And then it allows you to interact with a…”
Acharya Calls ChatGPT's Voice Module a Top Voice-First Innovation
“I think the second most interesting product, you know, outside of something like 11 is actually just ChatGPT's voice module and their voice sort of interface.”
Acharya Notes Users Quickly Forget ChatGPT's Voice Interface Is AI
“And you'll be amazed at how quickly you sort of forget that you're talking to an AI and you fall into the flow of a conversation.”
Acharya Predicts Voice Will Be the Primary Computing Interface Within 10 Years
“I mean, it's trite to cite this, but I will anyway, which is simply the movie Her. Like, that's it, you know, which is a passive sort of experience that has all the context that you have, where voice is the primary modality of interaction, and then you can fal…”
Acharya Argues Voice Tech Expands Markets Rather Than Replacing Existing Interfaces
“On the other hand, if you're maybe a senior citizen, You know, you simply, I mean, maybe you're interacting with technology, but you're really struggling with it, and this unlocks every product for you, potentially, so I think it's a market expansion positive …”
Acharya Advises Existing Apps to Double Down on Non-Voice Strengths
“I think the customer and the sort of interaction pattern is so different that if I had an existing product, I would probably focus on, you know, playing from my place of advantage, which is non voice workflows, rather than trying to rethink my product from the…”
Acharya Warns Users Expect Human-Level Responsiveness From Voice AI
“By default, their expectations are gonna be very high. It's gonna be like, look, I should be able to interact with this in the exact, with the same level of Fidelity and responsiveness and sort of flexibility that I can with a human.”
Acharya Predicts Early Adopters of Voice AI Will Face Social Friction
“Maybe we'll all change quickly together as a society, but I think there'll be some period of time for which, you know, the early adopters are gonna feel a little crazy or that they look that way.”
Acharya Advises Founders to Let Customers Guide AI Product Design
“Some of these technologies, if you're too pointed At the beginning in terms of product design, then it can constrain you. I think that if you introduce the capability in an intelligent way, your customers will pull you in a direction. I think some of the best …”