Dec 27, 2023 · 17m · a16z

Big Ideas 2024: The Consumer AI Battleground Moves from Model to UX with Alex Immerman

Alex Immerman · 10m spoken Steph Smith · 4m spoken
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 Podcast, growth partner Alex Immerman explains why the primary competitive battleground for consumer AI applications in 2024 is shifting from model performance to user experience (UX). He outlines how founders can build defensible moats, capitalize on falling model inference costs, and leverage open-source models to create novel software experiences.

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 2.2 Guest teaching 2.0 Guest disagreement 0.5 The host pushing back 2.2
05100:0010:000:21–3:41 · The host as informed peer 0/10 Announcing the Big Ideas 2024 List This segment consists of an introductory promo and a monologue where Alex presents his thesis that consumer AI differentiation is shifting from models to user experience. Because there is no host interaction during the monologue, all host scores are set to zero.3:41–5:54 · The host as informed peer 3/10 Establishing Moats in AI via User Experience Steph pushes back by asking whether UX can truly serve as a defensive moat given that design can be easily copied. Alex politely responds by drawing a historical comparison to Instagram during the mobile shift.5:54–10:35 · The host as informed peer 4/10 Model Economics and Models as Primitives Steph presses on model commoditization and cites the industry meme about OpenAI demo days wiping out wrapper startups. Alex frames models as primitives similar to ACH rails or geolocation, using Uber versus simple compass apps to illustrate true UX value.10:35–12:56 · The host as informed peer 2/10 Creative AI UX: Model Aggregation, Multiplayer, and NPCs The conversation turns entirely collaborative as Steph asks for concrete examples of creative UX. Alex details products like Poe for model aggregation and potential voice and NPC interactions in multiplayer gaming.12:56–15:00 · The host as informed peer 2/10 The Impact of Open Source Models Steph prompts Alex on the role of open source models in driving industry parity. Alex references predictions from Hugging Face's CEO and highlights fine-tuning image apps like EverArt and Can of Soup.15:00–17:13 · The host as informed peer 2/10 Second-Order Effects and the Future of AI Applications Steph asks about the second-order effects of decoupling application layers from infrastructure. Alex outlines how lower barriers encourage more founders before Steph concludes the episode.0:21–3:41 · Guest teaching 0/10 Announcing the Big Ideas 2024 List This segment consists of an introductory promo and a monologue where Alex presents his thesis that consumer AI differentiation is shifting from models to user experience. Because there is no host interaction during the monologue, all host scores are set to zero.3:41–5:54 · Guest teaching 2/10 Establishing Moats in AI via User Experience Steph pushes back by asking whether UX can truly serve as a defensive moat given that design can be easily copied. Alex politely responds by drawing a historical comparison to Instagram during the mobile shift.5:54–10:35 · Guest teaching 4/10 Model Economics and Models as Primitives Steph presses on model commoditization and cites the industry meme about OpenAI demo days wiping out wrapper startups. Alex frames models as primitives similar to ACH rails or geolocation, using Uber versus simple compass apps to illustrate true UX value.10:35–12:56 · Guest teaching 2/10 Creative AI UX: Model Aggregation, Multiplayer, and NPCs The conversation turns entirely collaborative as Steph asks for concrete examples of creative UX. Alex details products like Poe for model aggregation and potential voice and NPC interactions in multiplayer gaming.12:56–15:00 · Guest teaching 3/10 The Impact of Open Source Models Steph prompts Alex on the role of open source models in driving industry parity. Alex references predictions from Hugging Face's CEO and highlights fine-tuning image apps like EverArt and Can of Soup.15:00–17:13 · Guest teaching 1/10 Second-Order Effects and the Future of AI Applications Steph asks about the second-order effects of decoupling application layers from infrastructure. Alex outlines how lower barriers encourage more founders before Steph concludes the episode.0:21–3:41 · Guest disagreement 0/10 Announcing the Big Ideas 2024 List This segment consists of an introductory promo and a monologue where Alex presents his thesis that consumer AI differentiation is shifting from models to user experience. Because there is no host interaction during the monologue, all host scores are set to zero.3:41–5:54 · Guest disagreement 1/10 Establishing Moats in AI via User Experience Steph pushes back by asking whether UX can truly serve as a defensive moat given that design can be easily copied. Alex politely responds by drawing a historical comparison to Instagram during the mobile shift.5:54–10:35 · Guest disagreement 1/10 Model Economics and Models as Primitives Steph presses on model commoditization and cites the industry meme about OpenAI demo days wiping out wrapper startups. Alex frames models as primitives similar to ACH rails or geolocation, using Uber versus simple compass apps to illustrate true UX value.10:35–12:56 · Guest disagreement 0/10 Creative AI UX: Model Aggregation, Multiplayer, and NPCs The conversation turns entirely collaborative as Steph asks for concrete examples of creative UX. Alex details products like Poe for model aggregation and potential voice and NPC interactions in multiplayer gaming.12:56–15:00 · Guest disagreement 1/10 The Impact of Open Source Models Steph prompts Alex on the role of open source models in driving industry parity. Alex references predictions from Hugging Face's CEO and highlights fine-tuning image apps like EverArt and Can of Soup.15:00–17:13 · Guest disagreement 0/10 Second-Order Effects and the Future of AI Applications Steph asks about the second-order effects of decoupling application layers from infrastructure. Alex outlines how lower barriers encourage more founders before Steph concludes the episode.0:21–3:41 · The host pushing back 0/10 Announcing the Big Ideas 2024 List This segment consists of an introductory promo and a monologue where Alex presents his thesis that consumer AI differentiation is shifting from models to user experience. Because there is no host interaction during the monologue, all host scores are set to zero.3:41–5:54 · The host pushing back 4/10 Establishing Moats in AI via User Experience Steph pushes back by asking whether UX can truly serve as a defensive moat given that design can be easily copied. Alex politely responds by drawing a historical comparison to Instagram during the mobile shift.5:54–10:35 · The host pushing back 5/10 Model Economics and Models as Primitives Steph presses on model commoditization and cites the industry meme about OpenAI demo days wiping out wrapper startups. Alex frames models as primitives similar to ACH rails or geolocation, using Uber versus simple compass apps to illustrate true UX value.10:35–12:56 · The host pushing back 1/10 Creative AI UX: Model Aggregation, Multiplayer, and NPCs The conversation turns entirely collaborative as Steph asks for concrete examples of creative UX. Alex details products like Poe for model aggregation and potential voice and NPC interactions in multiplayer gaming.12:56–15:00 · The host pushing back 2/10 The Impact of Open Source Models Steph prompts Alex on the role of open source models in driving industry parity. Alex references predictions from Hugging Face's CEO and highlights fine-tuning image apps like EverArt and Can of Soup.15:00–17:13 · The host pushing back 1/10 Second-Order Effects and the Future of AI Applications Steph asks about the second-order effects of decoupling application layers from infrastructure. Alex outlines how lower barriers encourage more founders before Steph concludes the episode.

speaking balance: gold is the host, purple is the guest (3 minute bins)

0:00 · the host 0% · guest 100%0:00 · the host 0% · guest 100%3:00 · the host 0% · guest 100%3:00 · the host 0% · guest 100%6:00 · the host 0% · guest 100%6:00 · the host 0% · guest 100%9:00 · the host 0% · guest 100%9:00 · the host 0% · guest 100%12:00 · the host 0% · guest 100%12:00 · the host 0% · guest 100%15:00 · the host 0% · guest 100%15:00 · the host 0% · guest 100%
Sharpest disagreement ▶ 4:09 Alex counters premise on UX copyability

Alex offers his most direct pushback against the host's premise that design is easily copied, explaining how superior user experience raises switching costs and builds network effects.

Hardest push from the host ▶ 7:38 Host pushes on wrapper startups getting wiped out

Steph directly challenges Alex's optimism about app-layer startups by invoking the widespread industry risk of OpenAI demo days rendering wrapper startups obsolete.

Biggest teaching moment ▶ 8:38 Alex explains the compass app vs Uber primitive framework

Alex educates the host on how to think about model wrappers by contrasting basic compass apps with Uber's transformative use of smartphone geolocation primitives.

The host holds their own ▶ 7:38 Host demonstrates market awareness with OpenAI demo day reference

Steph demonstrates strong domain awareness by bringing up current founder anxieties and industry memes regarding incumbent model providers displacing thin wrapper products.

the scores for every segment, with the reasoning behind each
ChapterTopicThe host as informed peerGuest teachingGuest disagreementThe host pushing backWhy
Announcing the Big Ideas 2024 List 0000 This segment consists of an introductory promo and a monologue where Alex presents his thesis that consumer AI differentiation is shifting from models to user experience. Because there is no host interaction during the monologue, all host scores are set to zero.
Establishing Moats in AI via User Experience 3214 Steph pushes back by asking whether UX can truly serve as a defensive moat given that design can be easily copied. Alex politely responds by drawing a historical comparison to Instagram during the mobile shift.
Model Economics and Models as Primitives 4415 Steph presses on model commoditization and cites the industry meme about OpenAI demo days wiping out wrapper startups. Alex frames models as primitives similar to ACH rails or geolocation, using Uber versus simple compass apps to illustrate true UX value.
Creative AI UX: Model Aggregation, Multiplayer, and NPCs 2201 The conversation turns entirely collaborative as Steph asks for concrete examples of creative UX. Alex details products like Poe for model aggregation and potential voice and NPC interactions in multiplayer gaming.
The Impact of Open Source Models 2312 Steph prompts Alex on the role of open source models in driving industry parity. Alex references predictions from Hugging Face's CEO and highlights fine-tuning image apps like EverArt and Can of Soup.
Second-Order Effects and the Future of AI Applications 2101 Steph asks about the second-order effects of decoupling application layers from infrastructure. Alex outlines how lower barriers encourage more founders before Steph concludes the episode.

Statements from this episode (10)

Prediction Not checkable as stated
Immerman: Consumer AI battleground will shift from models to UX in 2024
“And my big idea for 20, 24 is the consumer AI battleground is going to move from the model to UX.”
Alex Immerman Dec 27, 2023 ▶ 1:44
Insight
Immerman: Traditional moats will drive long-term winners in consumer AI
“LLMs can be a source of differentiation. Today, they might provide a first-mover advantage, but old-fashioned modes like network effects, high switching costs, scale, and brand are still what will likely win long term.”
Alex Immerman Dec 27, 2023 ▶ 3:24
Insight
Immerman: Tech paradigms always progress from technical features to UX
“Over the history of technology paradigms, we often see technology progress and features first, and then a movement toward prioritizing design and experience.”
Alex Immerman Dec 27, 2023 ▶ 4:11
Prediction Held up
Immerman: Access costs for quality AI models will fall
“What's changing now is we expect the costs to access quality models, whether closed source or open source to fall.”
Alex Immerman Dec 27, 2023 ▶ 6:42
Insight
Immerman: AI models are software primitives similar to payment rails
“One framework to consider is that models are just a new primitive to build applications on top of the same way storages or geolocation or payment rails are today.”
Alex Immerman Dec 27, 2023 ▶ 7:04
Insight
Immerman: Defensible AI businesses do not require building proprietary models
“Building a defensible business does not have building a model as a prerequisite.”
Alex Immerman Dec 27, 2023 ▶ 9:16
Opinion
Immerman: Perplexity proves UX value can be built on third-party models
“In the AI context, I'm a fan of perplexity, which is an AI powered research assistant. It does work with third party models. But it also offers a lot of seemingly small features that add up to a compelling user experience. Lower latency, providing sources, sug…”
Alex Immerman Dec 27, 2023 ▶ 10:04
Prediction Not checkable as stated
Alex Immerman: Shared AI experiences will evolve from text to multimodal
“Looking ahead, these shared experiences are going to go from just being text to being multimodal.”
Alex Immerman Dec 27, 2023 ▶ 12:04
Prediction Not checkable as stated
Immerman: Open-source AI models are heading toward parity with closed-source
“Clem, CEO at Hugging Face, a few days ago, forecasted that in 2024, open source models are going to reach the level of the best closed source models. That may be aggressive for 20, 24, but they are headed towards parity.”
Alex Immerman Dec 27, 2023 ▶ 13:31
Prediction Not checkable as stated
Immerman: Decoupling AI apps and infrastructure makes startup creation easier
“So as the application and infrastructure layers are split, That'll make it a lot easier for new companies to be built. Founders can just build for applications and not have to worry as much about the underlying model as an input to their own company creation.”
Alex Immerman Dec 27, 2023 ▶ 15:25
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 1,000 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.