Skip to content

Debates

Is Apple positioned to win in AI, or is it falling behind?

15 recorded positions from 10 people, first said Jun 30, 2023. They do not agree — the readings below are what each one actually argued.

Apple has strong theoretical ai positioning

Des Traynor · Nov 15, 2023 · hedged

Apple will make major strides in AI and win consumer AI specifically — an on-device, privacy-preserving Siri as conversational as ChatGPT that can take actions, changing the interaction model across iOS and desktop

Apple is a well-run, privacy-focused company that must have a head of AI training models designed to run natively on its hardware and integrate with AirPods, Watch and phone; Siri is currently useless only because it lacks the smarts

Scope: 'who owns consumer AI could well be Apple'; does not rule out OpenAI's or Amazon/Anthropic's success; disappointed at how long Apple has taken

38:47 20VC: How to Survive and Thrive in a World of OpenAI, Are LLMs Being Commoditised, Where Does the Value Lie; Infrastructure or Application Layer, How Apple Could Win in a World of AI, How Amazon Could Threaten OpenAI and Why Google Struggle with Des Trayn

Jeff Seibert · Nov 22, 2023 · speculative

Apple is the most overlooked player in AI because its control of custom silicon could let it pioneer on-device small models with outlandish performance relative to other platforms

They control the silicon, so pairing small on-device models with custom chips designed to run them could yield performance no other platform can match

Scope: conditional on Apple pioneering small on-device models and building custom silicon for them

0:00 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits

Harry Stebbings · Jan 10, 2024

Apple is more exciting than ever

Its ability to run models locally on device gives it unparalleled access and advantage at the end-consumer point

58:28 20VC: Did Figma Kill M&A Markets in 2024, The Three Biggest Mistakes Made in Growth Investing, The Three Requirements Companies Need to Go Public in 2024 with Ed Sim and Jamin Ball

David Luan · Jun 24, 2024

Apple has a massive advantage in the one of the two sources of leverage in ML — running smart models for free at the edge — and will completely crush use cases that are private or fine-tuned on personal data but don't require massive reasoning

Model capability comes in concentric rings: small models already do things like tweet sentiment perfectly, so those skills will live at the edge and frontier providers can't monetize them, while a billion-parameter model won't design a 3D car part for some time

Scope: whether edge advantage alone is enough is a hard open question; does not cover heavy-reasoning tasks

27:25 20VC: Why Foundation Model Performance is Not Diminishing But Models Are Commoditising, Why Nvidia Will Enter the Model Space and Models Will Enter the Chip Space & The Right Business Model for AI Software with David Luan, Co-Founder @ Adept

Richard Socher · Apr 18, 2025 · hedged

Apple would in theory be extremely well positioned to win in AI

Scope: in theory

19:43 20VC: Foundation Models: Who Wins & Who Loses | How Economies and Labour Markets Need to Change in a World of AI | China vs the US in an AI Race: What You Need to Know | Rich Socher, Founder @ You.com

Apple has not executed on its ai potential

Douwe Kiela · Jun 30, 2023

Apple has so far failed to produce anything interesting in AI despite being best positioned via Siri as an early personal assistant.

Siri backed by a powerful language model could do very interesting things, but nothing interesting has come out of Apple yet.

33:03 20VC: Why Data Size Matters More Than Model Size, Why The Google Employee Was Wrong; OpenAI and Google Have the Advantage & Why Open Source is Not Going to Win with Douwe Kiela, Co-Founder @ Contextual AI

Ed Sim · Jan 10, 2024 · hedged

Apple is a short for 2024

There's an iPhone growth problem, and although on-device/edge machine learning is exciting, he doesn't see it translating into revenue this year

Scope: on-device AI with built-in privacy, compliance and speed is genuinely interesting; timeframe limited to this year

57:40 20VC: Did Figma Kill M&A Markets in 2024, The Three Biggest Mistakes Made in Growth Investing, The Three Requirements Companies Need to Go Public in 2024 with Ed Sim and Jamin Ball

Harry Stebbings · Apr 18, 2025

In practice Apple has not done much with its AI position

19:46 20VC: Foundation Models: Who Wins & Who Loses | How Economies and Labour Markets Need to Change in a World of AI | China vs the US in an AI Race: What You Need to Know | Rich Socher, Founder @ You.com

Luke Harries · May 23, 2025

Apple is the worst current offender at marketing an unready product, pushing Apple Intelligence as the reason to upgrade devices

No one knows what Apple Intelligence is, how to use it, or gets value from it, and it still isn't shipped — yet it anchors their upgrade marketing, despite Apple's legacy as a marketing and product beacon under Steve Jobs

64:38 20VC: ElevenLabs Head of Growth on Why You Do Not Need PMs | The 7-Part Launch Playbook That Gets 700K+ Views Per Product | The Truth About CAC, Payback & Performance Marketing in AI with Luke Harries

Ai task automation turns phone into command center opening new monetization surface for apple

Harry Stebbings · Nov 15, 2023 · speculative

An Apple intent layer would open up a paid-search-style ad market where providers like Spotify pay Apple for default placement in responses

40:28 20VC: How to Survive and Thrive in a World of OpenAI, Are LLMs Being Commoditised, Where Does the Value Lie; Infrastructure or Application Layer, How Apple Could Win in a World of AI, How Amazon Could Threaten OpenAI and Why Google Struggle with Des Trayn

Aaron Levie · May 22, 2024

AI turns the phone into a task automation command center, opening an effectively infinite set of new transactions Apple could facilitate and monetize

If a billion-plus users start doing entirely new things — booking flights, ordering rides, getting answers — through Apple's interfaces and APIs, that's a powerful new monetization surface Apple isn't currently capturing

Scope: how they'd actually monetize is unknown

48:08 20VC: Box's Aaron Levie on Predictions for the Next Wave of AI: Will Foundation Models Be Commoditised | How the Business Model of SaaS Changes Forever | Startups vs Incumbents: Who Wins | App vs Infrastructure Layer: Where is the Value?

Also on the record

Jeff Seibert · Nov 22, 2023 · hedged

Apple will be able to run a sufficiently large LLM on-device within about five years, which would cut OpenAI out of the picture entirely for iPhone users

Apple is focused on privacy and doesn't want data leaving the device, and the only way to deliver AI under that constraint is an on-device model

29:00 Apple on device llm within five years cuts out openai for iphone users

Jerry Murdock · Aug 22, 2026 · hedged

Apple's AI position is only defensible if they've been watchful observers with something up their sleeve; if they are merely dumb consumers of AI they are in big trouble

The answer depends on the internal AI culture at Apple, which outsiders can't see — they're a massive Claude Code customer but it's unclear whether they have anything innovative behind it

59:35 Defensible only if hidden capability exists unknowable outside

Aaron Levie · May 22, 2024

Apple partnering with OpenAI for Siri's underlying engine is not a weakness — it's just a normal technology supply chain decision

As a consumer, you don't care what engine powers Siri; you just want to click your phone, ask a question and get something done with the best AI on your device

50:35 Outsourcing siris engine to openai is a normal supply chain decision not a weakness

Richard Socher · Apr 18, 2025 · hedged

For some consumer use cases, Google and Apple in theory already hold the data needed to make agents work well

18:18 Google and apple already hold agent data if they choose to use it

Your assistant can query this graph directly — 15 positions here, 19,646 across the corpus. Add 996.fm over MCP.