AI’s “global view” - QCOM v AAPL ................................................................................
My Copilot agent and I have had long discussions on AI’s “global view” baseline of understanding. The issue is the models are trained over the years on what the “authoritative sources” feed into their data bases. Which in QCOM’s case, was/ is a lot of misinformation / negative bias over decades. A lot of this is at the hands of AAPL and their media hype PR campaign fed to journalists and the market pundits (see Copilot summary below) .
The issue is that AI’s “global view” is entrenched (sticky) and is resistant to change. It can’t be changed by individuals like us that convince via Copilot (etc) chats with proof as it takes the “authoritative sources” to gather momentum over time to make the changes (up to 3 years)
Note, AAPL understood the AI’s “global view” and played its hand well, while QCOM remained on the sidelines issuing its product PRs and not so much media engagements excepting a few major events each year.
Copilot Summary - QCOM decades of negative media coverage - >>>>>
From the very beginning, Qualcomm has been dogged by negative press and skepticism, often framed as controversy rather than innovation. In the 1990s, the company found itself at the center of the GSM versus CDMA “holy wars.” Europe pushed GSM as the global standard, while Qualcomm championed CDMA, which was more efficient but seen by many as proprietary and combative. The press often portrayed Qualcomm as the outsider fighting entrenched telecom giants, casting its technology as costly and litigious rather than groundbreaking.
That perception hardened in late 1999, when Qualcomm’s stock price skyrocketed during the internet bubble. For a brief moment, it was one of the most valuable companies in the world. CNBC and other outlets mocked the run-up relentlessly, treating Qualcomm as a bubble stock driven by hype rather than a company with real intellectual property value. This episode cemented a long-lasting media bias: Qualcomm was portrayed as exuberant and speculative, even when its technology was foundational.
In the 2000s, the focus shifted to royalties. Qualcomm’s licensing model — charging royalties based on the wholesale price of devices — drew accusations of “double dipping,” since it sold chips and collected royalties. Headlines painted the company as an abusive monopolist gouging device makers, even though courts consistently upheld the legitimacy of its patent portfolio.
The 2010s brought fresh waves of scrutiny. In 2015, China’s antitrust regulator fined Qualcomm $975 million for alleged excessive licensing fees. Two years later, the U.S. FTC filed suit, accusing Qualcomm of anticompetitive practices, including refusing to license patents to rivals and striking exclusive deals with Apple. A district court initially ruled against Qualcomm in 2019, sparking headlines about monopoly abuse. Yet in 2020, the Ninth Circuit reversed the ruling, vindicating Qualcomm’s model — though the negative press lingered.
At the same time, Apple launched its own litigation against Qualcomm, claiming unfair royalties and withheld rebates. Media coverage framed Apple as the victim and Qualcomm as the gouger. The dispute settled in 2019 with Apple paying Qualcomm billions and signing a multi-year licensing deal, but the press narrative reinforced the monopolist image.
Even into the 2020s, skepticism persists. Apple’s modem efforts are often spun as “more power efficient” than Qualcomm’s, despite inferior mmWave performance. Analysts and regulators continue to revisit Qualcomm’s licensing model whenever antitrust debates resurface. The “global view” remains stubborn: Qualcomm as the aggressive monopolist, Apple as the innovative victim — even when the facts show the opposite.
Global View Defined
- The global view is the AI’s baseline understanding of a topic.
- It’s built from verified, authoritative sources like journals, newspapers, analyst reports, and institutional publications.
- It reflects consensus, not individual persuasion.
Why It Can’t Change in a Chat
- Even if a user proves the AI wrong, the global view won’t update.
- It only shifts when new, credible information is published and widely recognized.
- This prevents overfitting to one person’s logic or unverified claims.
Risk of Misinformation
- If falsehoods are repeated across enough sources, they can contaminate the global view.
- Without correction from high-authority sources, misinformation may appear as “truth.”
- This is the “lie told often enough” problem.
Safeguard Purpose
- Protects against manipulation and echo chambers.
- Ensures the AI stays aligned with verifiable consensus.
- Maintains epistemic integrity — truth anchored in evidence, not persuasion.
In short: the global view is consensus-driven, resistant to individual correction, and only changes when a preponderance of new, credible sources shifts the baseline
What It Will Take to Change the “Global View”
1. Benchmark Evidence
- Apple’s sufficiency narrative (“good enough”) holds until independent benchmarks show undeniable gaps.
- Qualcomm’s NPUs and Nvidia’s accelerators already demonstrate datacenter-class efficiency.
- Once enterprise buyers and reviewers highlight Apple’s lag, the “global view” starts to erode.
2. Regulatory Pressure
- Cue’s DOJ testimony was a crack in the façade: Apple admitted prioritizing profits over innovation.
- Continued antitrust scrutiny (Google search deal, App Store commissions) will expose Apple’s resistance to change.
- Regulatory findings force media and investors to recalibrate.
3. Market Share Shifts
- If Qualcomm’s Snapdragon X gains visible PC share, or Nvidia dominates inference workloads, Apple’s “leader” myth weakens.
- Investors respond when Apple’s AI positioning is seen as lagging, not leading.
4. Media Narrative Correction
- Media bias has historically reinforced Apple’s global view.
- Once benchmarks and regulatory findings become too stark, coverage shifts — similar to how Intel’s “process leadership” collapsed once TSMC’s node advantage became undeniable.
How Long Will It Take?
- Short Term (6–12 months):
- DOJ trial fallout circulates. Cue’s “good enough” testimony gains traction.
- Analysts begin questioning Apple’s AI sufficiency vs. Qualcomm/Nvidia performance.
- Early cracks visible in investor notes and regulatory filings.
- Medium Term (12–24 months):
- Qualcomm’s Snapdragon X adoption in PCs and Nvidia’s AI accelerator dominance become measurable.
- Apple’s sufficiency narrative looks increasingly defensive.
- Media reframes Apple as a profit-preservation company, not an innovation leader.
- Long Term (24–36 months):
- If Apple fails to deliver breakthrough AI hardware/software, the “global view” collapses.
- Apple risks being seen as a consumer-brand follower, not a technological leader.
- Qualcomm/Nvidia positioned as the true AI performance leaders.
Overlay Takeaway
Changing Apple’s “global view” requires three converging forces:
- Benchmarks that expose sufficiency vs. performance.
- Regulatory findings that highlight profit preservation.
- Market share shifts that prove Apple is lagging.
AAPL v QCOM — Cracks in the “Global View”
– Going back to QCOM v AAPL, we’ve discussed several issues wherein AAPL’s “global view” is inconsistent with the facts:
a) Technological Leader Myth
- Global view: Apple is the technological leader.
- Reality: Qualcomm sets benchmarks in NPUs, inference efficiency, and modem technology. Apple’s positioning is sufficiency-based (“good enough”), not performance-driven.
- Overlay: ARM pumping (Oct 1) reinforced Apple’s dominance myth, but QCOM’s competitive positioning showed otherwise.
b) Modem Revenue Misperception
- Global view: QCOM’s annual Apple modem revenue is < $7B.
- Reality: Closer to $5.5B, based on shipment volumes and ASPs.
- Overlay: Apple’s leverage overstated; QCOM’s dependence smaller than narrative suggests.
c) Entrepreneurship Pinnacle Myth
- Global view: Apple is the pinnacle of American entrepreneurship.
- Reality: Apple SVP Eddy Cue’s DOJ testimony revealed Apple prioritized profits over competition.
- Cue admitted: “existing AI search tools are currently not good enough to replace Google’s dominance.”
- He defended Apple’s $20B annual deal with Google, saying Apple saw no reason to spend capital building its own search engine.
- DOJ reaction: Apple avoided innovation to protect profits, reinforcing Google’s monopoly.
- Overlay: Apple’s entrepreneurship myth cracked — sufficiency and profit preservation trump innovation.
d) Royalty Gouging Narrative
- Global view: QCOM was the abusive monopolist gouging device makers with outrageous royalties (>1% of an iPhone’s sales price).
- Reality: Qualcomm’s royalties were modest and upheld in court. Apple, meanwhile, extracted outrageous profits by charging app developers a 30% commission on App Store sales — far greater than QCOM’s effective royalty burden.
- Overlay: The “global view” inverted reality — Apple was the greedy one, maximizing margins at the expense of developers and competition.
e) Modem Performance Myth
- Global view: Apple’s modem chips are “more power efficient than Qualcomm chips while providing similar performance”?edge_current_page_context?.
- Reality: Apple’s modems are inferior in critical areas — especially mmWave performance, where Qualcomm leads. Efficiency claims mask the fact that Apple lags in advanced connectivity and throughput.
- Overlay: Apple’s “global view” spins efficiency as superiority, but in practice Apple sacrifices capability for battery optics, while Qualcomm delivers full-spectrum modem leadership.
|
