Apple owning an open-weight LLM hosting strategy would be the single best product strategy for their hardware.
Predictor: Ben Horowitz · ep#232 "Ben Horowitz: xAI Executive Exodus, Apple's AI Crisis, The Pace of AI | EP #232" · source
Prediction text
Apple owning an open-weight LLM hosting strategy would be the single best product strategy for their hardware. | 100% I think that because it you know it's going to go way beyond open claw as we just discussed but I it is it would be so such a breakthrough for Apple uh in their thinking and organizationally and culturally for them to go for it on like you know farms losters that like it would be surprising if they did it. It's very obviously a fantastic idea. It's probably the single best product strategy idea.
Watch events: Apple WWDC 2026 June 8-12 keynote
Verbatim quote
100% I think that because it you know it's going to go way beyond open claw as we just discussed but I it is it would be so such a breakthrough for Apple uh in their thinking and organizationally and culturally for them to go for it on like you know farms losters that like it would be surprising if they did it. It's very obviously a fantastic idea. It's probably the single best product strategy idea.
Predictor: Ben Horowitz
Evidence about this node from Ben Horowitz is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).
Reference class
This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.
Probability over time
Milestone chain
- 2025-09-15hitApple ships on-device foundation model in iOS 26 with developer frameworkHow: Apple releases iOS 26 with Foundation Models framework exposing ~3B parameter on-device LLM to third-party developersSource: https://developer.apple.com/documentation/FoundationModelsconf 99%Notes: HIT — Apple shipped Foundation Models framework with ~3B parameter on-device LLM in iOS 26.
- 2026-04-05hitApple model remains primarily walled-garden, not open-weightsHow: Apple's foundation model weights remain proprietary (no Hugging Face or GitHub open release) >=12 months after iOS 26 launchSource: https://www.agent-wars.com/news/2026-04-05-apfel-apple-foundation-model-access-toolconf 95%Notes: HIT — As of April 2026 Apple's model weights still locked inside walled garden; third-party tools like Apfel only expose access, not weights. Contradicts prediction.
- 2026-06-01 → 2027-06-30pendingApple announces formal open-weights or open-licensing strategy for a foundation modelHow: Apple publishes weights of any foundation model (>=1B params) under permissive license (Apache 2.0, MIT, or Llama-style) on Hugging Face or apple.com/researchSource: Apple Machine Learning Research blog, WWDC keynoteconf 20%Notes: Low confidence — Apple historically guards weights; Horowitz prediction is normative ('would be best'), not directional.
- 2026-06-01 → 2027-06-30pendingApple ships server-side foundation model API to developers (Private Cloud Compute)How: Apple opens Private Cloud Compute or equivalent server-side foundation model API to third-party developers beyond Apple IntelligenceSource: https://machinelearning.apple.com/research/apple-foundation-models-2025-updatesconf 55%
- 2026-09-01 → 2027-12-31pendingApple iPhone hardware sales beneficiary of on-device AI: services + hardware revenue inflectionHow: Apple reports >=2 consecutive quarters where Mac+iPhone unit growth attributable to AI features explicitly cited in earnings call or 10-Q narrativeSource: Apple 10-Q filings, earnings call transcriptsconf 45%
- 2026-09-01 → 2027-12-31pendingOpen foundation model strategy adopted at competitor (Samsung/Google/Microsoft) hardware tierHow: At least one major Apple competitor (Samsung, Google Pixel, Microsoft Surface) ships open-weight foundation model bundled with hardware as primary differentiatorSource: Press releases, Reuters, MIT Tech Reviewconf 50%Notes: If competitor leads here, would pressure Apple to follow per Horowitz thesis.
What if this resolves?
Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"
Evidence chain
Network propagation neighbors
Top incoming (parents)
Edges that influence THIS node's belief
| Kind | Node | Their prob | P(c|s=T) | P(c|s=F) | Δ implied |
|---|---|---|---|---|---|
| killer | TK09 Energy Grid Cap (Data Center Power Wall) | 35.0% | 0.050 | 0.550 | -0.088 |
| prereq | 234_012 Anthropic revenue will cross OpenAI revenue in middle of 202 — Peter Diamandis | 67.1% | 0.550 | 0.050 | -0.081 |
| prereq | SEM_042 2025 will be the definitive year that agentic systems finall — Kevin Weil | 73.8% | 0.550 | 0.050 | -0.049 |
| killer | TK06 China-Taiwan Military Conflict | 8.0% | 0.050 | 0.550 | +0.047 |
| prereq | 235_002 Anthropic will exceed OpenAI in revenue this year (2026). — Dave Blundin | 74.6% | 0.550 | 0.050 | -0.045 |
Top outgoing (children)
Predictions THIS node influences
| Kind | Node | Their prob | P(c|s=T) | P(c|s=F) | Δ implied |
|---|---|---|---|---|---|
| prereq | CMQ_002 By 2028, AI systems will reach 'independent researcher' leve — Sam Altman | 31.4% | 0.550 | 0.050 | -0.035 |
| prereq | 241_043 ASI will arrive within 2 years to 5 years to this next decad — Peter Diamandis | 35.9% | 0.650 | 0.050 | -0.034 |
| prereq | 235_030 Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 203 — Ray Kurzweil | 39.2% | 0.750 | 0.050 | -0.022 |
| prereq | SEM_034 True artificial general intelligence will be achieved betwee — Demis Hassabis | 28.7% | 0.550 | 0.050 | -0.008 |
| prereq | 232_055 We're exiting the industrial age permanently as recursive se — Peter Diamandis | 35.5% | 0.700 | 0.050 | -0.007 |
Ticker exposure
Beneficiaries (23)
Adverse (6)
Prerequisites (10)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| prereq | SEM_008 | Training runs costing $10 billion for a single model will commence sometime in 2025. | AI | — |
| prereq | 234_012 | Anthropic revenue will cross OpenAI revenue in middle of 2026 | Markets/Stocks | — |
| prereq | 235_002 | Anthropic will exceed OpenAI in revenue this year (2026). | AI | — |
| prereq | SEM_012 | Nvidia quadrupled chip production output while only doubling human headcount — achieved by deploying AI coding tools (Cursor, Claude Code) across engineering. | AI/Manufacturing | — |
| prereq | SEM_042 | 2025 will be the definitive year that agentic systems finally hit the mainstream. | AI/Agents | — |
| killer | TK09 | Energy Grid Cap (Data Center Power Wall) | — | — |
| killer | TK01 | AGI Capability Plateau (2026-27 Training Stall) | — | — |
| killer | TK02 | AI Compute Supply Shock (TSMC/Taiwan Disruption) | — | — |
| killer | TK03 | AI Regulatory Moratorium (EU/US Capability Freeze) | — | — |
| killer | TK06 | China-Taiwan Military Conflict | — | — |
Dependents (5)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| prereq | 235_030 | Ray Kurzweil predicts Longevity Escape Velocity (LEV) by 2033. | Biotech/Longevity | — |
| prereq | 232_055 | We're exiting the industrial age permanently as recursive self-improvement unfolds. | AI | — |
| prereq | 241_043 | ASI will arrive within 2 years to 5 years to this next decade | AI | — |
| prereq | CMQ_002 | By 2028, AI systems will reach 'independent researcher' level — driving autonomous scientific discoveries without human intervention. | AI | — |
| prereq | SEM_034 | True artificial general intelligence will be achieved between 2032 and 2042 — 'first we solve AI, then use AI to solve everything else'. | AI/AGI | — |
Linked documents (2)
| Sim | Source | Title | Market prob | Polarity | Reviewed | Published |
|---|---|---|---|---|---|---|
| 0.662 | manifold | Will a US open-weights LLM outrank every Chinese open-weights LLM on LMArena at the end of 2027? | 32% | mentions | pending | 2026-05-31 |
| 0.559 | manifold | Will Android succeed in making it harder to side-load apps | 85% | mentions | pending | 2026-05-14 |
Raw metadata
{
"nia": false,
"url": "https://www.youtube.com/watch?v=C1GLT9_tag0",
"mode": "ASPIRATION",
"role": "Guest-VC",
"caveats": "Conditional on Apple actually doing it; 'would be surprising if they did'",
"context": "do you think like if if you were Tim Cook would you be pivoting and changing course and and like owning the open claw strategy for Apple hardware? 100%. I mean 100% I think that because it you know it's going to go way beyond open claw as we just discussed but I it is it would be so such a breakthrough for Apple uh in their thinking and organizationally and culturally for them to go for it",
"to_year": 2028,
"verbatim": "100% I think that because it you know it's going to go way beyond open claw as we just discussed but I it is it would be so such a breakthrough for Apple uh in their thinking and organizationally and culturally for them to go for it on like you know farms losters that like it would be surprising if they did it. It's very obviously a fantastic idea. It's probably the single best product strategy idea.",
"conv_cues": "100%; single best product strategy idea",
"direction": "HAPPEN",
"from_year": 2026,
"timeframe": "unspecified",
"conv_level": "HIGH",
"milestones": [
{
"kind": "llm_pre_event",
"label": "Apple ships on-device foundation model in iOS 26 with developer framework",
"notes": "HIT — Apple shipped Foundation Models framework with ~3B parameter on-device LLM in iOS 26.",
"source": "https://developer.apple.com/documentation/FoundationModels",
"status": "hit",
"weight": 0.4,
"ordinal": -11,
"source_id": null,
"confidence": 0.99,
"source_url": "https://developer.apple.com/documentation/FoundationModels",
"expected_date": "2025-09-30",
"observed_date": "2025-09-15",
"research_origin": "deep_research",
"measurement_criterion": "Apple releases iOS 26 with Foundation Models framework exposing ~3B parameter on-device LLM to third-party developers"
},
{
"kind": "prereq",
"label": "Nvidia quadrupled chip production output while only doubling human headcount — achieved by deploying AI coding tools (Cursor, Claude Code) a",
"status": "hit",
"weight": 0.5,
"ordinal": -10,
"source_id": "SEM_012",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "Training runs costing $10 billion for a single model will commence sometime in 2025.",
"status": "hit",
"weight": 0.5,
"ordinal": -9,
"source_id": "SEM_008",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "Anthropic revenue will cross OpenAI revenue in middle of 2026",
"status": "hit",
"weight": 0.5,
"ordinal": -8,
"source_id": "234_012",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "Anthropic will exceed OpenAI in revenue this year (2026).",
"status": "hit",
"weight": 0.5,
"ordinal": -7,
"source_id": "235_002",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "prereq",
"label": "2025 will be the definitive year that agentic systems finally hit the mainstream.",
"status": "hit",
"weight": 0.5,
"ordinal": -6,
"source_id": "SEM_042",
"expected_date": "2026-04-29",
"observed_date": "2026-04-29"
},
{
"kind": "llm_pre_event",
"label": "Apple model remains primarily walled-garden, not open-weights",
"notes": "HIT — As of April 2026 Apple's model weights still locked inside walled garden; third-party tools like Apfel only expose access, not weights. Contradicts prediction.",
"source": "https://www.agent-wars.com/news/2026-04-05-apfel-apple-foundation-model-access-tool",
"status": "hit",
... (truncated)