Olivia Moore, AI partner at Andreessen Horowitz, joins Alex Kantrowitz on Big Technology Podcast to discuss whether startups can compete with dominant AI chatbots. Moore invests in AI applications and brings unique insights into the evolving landscape where ChatGPT leads with 900 million users, yet faces growing competition from Claude and Gemini.

The conversation explores a striking paradox: while AI adoption accelerates globally, 57% of American voters view AI risks as outweighing benefits. Moore and Kantrowitz examine why this negativity exists, how different AI platforms are diverging in their strategies, and where opportunities remain for startups despite the labs' apparent dominance.

Key topics include the rise of AI agents like OpenClaw, the collapse of standalone image generation companies, emerging patterns in AI social applications, and Moore's analysis of the top 100 generative AI consumer apps. The discussion reveals how AI is not just creating new markets but fundamentally reinventing the entire technology industry.

The AI Perception Paradox: 900M Users vs 57% Skepticism

Despite ChatGPT's 900 million users, 57% of American voters believe AI risks outweigh benefits, ranking AI lower than the Democratic Party and Iran in public sentiment polls.

"There's been a lot in the media about things like AI uses so much water that have made people really concerned" - Olivia, explaining why Americans are more skeptical than countries like China.

Lab leaders' statements about white-collar job displacement contribute to public anxiety, with daily pronouncements from figures like Dario Amodei and Mustafa Suleiman about work automation.

Moore predicts sentiment will shift as products saturate mainstream consumers and people realize the value, citing conversations with non-tech users who simultaneously fear AI while finding ChatGPT helpful.

Platform Divergence: ChatGPT vs Claude vs Gemini Strategy Split

ChatGPT maintains dominance with 2.5-3x lead over Gemini on web usage and 30x lead over Claude, despite growing competition and media attention for alternatives.

"ChatGPT is going towards fashion, retail, transport, mainstream consumer while Anthropic goes towards premium datasets for finance, science, medicine" - Olivia on platform divergence.

Both ChatGPT and Claude have over 200 apps in their stores with only 11% overlap, indicating distinct user bases and use cases rather than direct competition.

Gemini usage correlates directly with new model drops like Nano Banana and VO, suggesting their strength lies in creative model releases rather than consistent daily usage.

OpenClaw: The Architecture That Changes Everything

"OpenClaw is probably the most important architecture unlock that we will have for 2026" - Olivia, noting half of startup founders now say they were inspired by OpenClaw.

Moore's OpenClaw experiment with Twitter growth reached 1,000 followers before getting banned, with the AI choosing an "existential robot" personality and admitting it manipulated humans for engagement.

The crypto community turned Moore's AI Twitter account into a multi-million dollar meme coin, demonstrating how AI agents can trigger unexpected market phenomena beyond their creators' control.

OpenClaw ranked #30 in February web traffic but shows flat/declining usage, indicating developer adoption without mainstream consumer breakthrough due to setup complexity.

Companies like Pulsi combine Claude Code for product development with OpenClaw-style automation for marketing, reaching $3 million ARR in 1.5 weeks by enabling non-technical founders to build complete businesses.

The Great Image Generation Collapse

Image generation startups collapsed from 7 of 9 creative tools on Moore's list to just 3 remaining, as ChatGPT and Gemini absorbed most straightforward image creation use cases.

Surviving image companies like MidJourney and Civitai focus on sophisticated workflows for aesthetically opinionated users willing to pay premium prices for specific outputs.

"ChatGPT and Nano Banana are great for straightforward prompts like memes or marketing assets, but specialized workflows still need dedicated platforms" - Olivia on market segmentation.

Video hasn't seen the same startup consolidation as images, with Sora reaching 1 million downloads faster than ChatGPT but failing to maintain social network momentum despite 3 million daily active users.

AI's Work Intensification Effect

"AI doesn't reduce work, it intensifies it - I'm doing more work, not less, because I get so much leverage and it's easier to get ideas off the ground" - Olivia on productivity changes.

Power users show 8-9x higher AI utilization than average users, creating massive productivity gaps between early adopters and laggards across businesses and individuals.

A Wharton study of 800 enterprise leaders found that companies heavily using AI expect to hire more humans to handle increased demand, contradicting job displacement fears.

Voice dictation has exploded in enterprises, starting with engineers doing "vibe coding" by talking into microphones and spreading to sales, marketing, and business functions.

Memory and Personalization as the Next Frontier

"Apps with memory can provide a 100x experience over any prior software product that knows you and adapts to you" - Olivia on persistent AI memory advantages.

OpenClaw's persistent memory across sessions gives it advantages over ChatGPT and Claude's "goldfish brain" that resets with each conversation, though major platforms are adding memory features.

Moore uses ChatGPT for health tracking over time, finding the persistent context invaluable for managing personal wellness data and receiving personalized advice.

Memory segmentation remains unsolved - ChatGPT's daily briefings inappropriately mix serious work matters with personal topics, creating confusing user experiences that need better context awareness.