Apple, Qualcomm and Summit Day 3 – Peace Declared

Qualcomm vs. Apple – Armistice

• Apple and Qualcomm have signed a renewal of their global patent licensing deal and said nothing more, meaning that all the details and what this means are in the realm of conjecture.
• Given the size of Apple and its handset business, any change in the licence terms would be a price-sensitive event, and hence both companies, as listed entities, would be under some pressure from the SEC to disclose more details for the benefit of shareholders if something had radically changed.
• This, they have done in the past.
• Hence, I don’t think very much has changed from what has gone before, meaning that Apple is now a regular licensee, just like everyone else.
• The stony silence implies that this is probably a long-term agreement (I estimate 6 to 10 years) and that it has been struck at the usual royalty rate that has been the norm for many years.
• This is a substantially positive result for Qualcomm as 2027 was on track for a major resumption of hostilities, once Apple had been able to fully remove its dependence on Qualcomm modems.
• This was clear as the only reason that it caved in back in 2019 (see here) when it signed a 6-year extendable deal at the same royalty rate while it worked on its own in-house modems.
• These took a bit longer than expected, but the path is now clear for the Apple C-series of modems to take 100% of the Apple device portfolio in 2027, opening the way for Apple to resume legal hostilities.
• This deal is a sign that Apple has decided not to take this route and let bygones be bygones, which I think is a decision explicitly taken by John Ternus.
• As the new CEO of Apple, he has far bigger issues to deal with, and a messy and very public legal entanglement is not the best way to start one’s tenure.
• This does not move the needle for Qualcomm in terms of profits (as no one has forecast losing Apple royalties again in their numbers), but it does significantly reduce the risk profile of the company.
• Another large legal fracas would have undoubtedly cost hundreds of millions, increased the volatility of the share price, and wasted days and weeks of management time.
• None of this will now come to pass, which is great news for shareholders but bad news for the lawyers and industry observers who enjoy a good punch-up.

Summit Day 3: The themes (almost) brought to life through demos.

• Day 3 took the discussions of Days 1 and 2 and turned them into a series of live demos that were both quite effective and also demonstrated that agents have some way to go both in terms of reliability and trust.
• These included demos from memories.ai, AGI Inc, and Liquid AI, all of which have a take on how to make agents personal, private and on-device while having the context to be much more useful.
• All of these demos worked reasonably well, but at the same time they make compromises that demonstrate that the reality of having an agent doing everything for you remains quite some way off.
• For example, AGI’s agent can access all the apps and services on the device without an API by mimicking the user’s touch input, which is fine except that you have to grant the agent full access to your entire digital life.
• The net result is that having an agent do things for the user is much closer to reality than it was last year when the idea was first proposed, but the implementation needs a lot of finesse as well as trust and reliability.
• RFM Research has a standard test for the take-up of new technology, which is if a technology is better, faster and cheaper than the one that goes before, then it is likely to see adoption.
• Agents have yet to reliably tick all of these boxes, which is why I think the agentic takeover of consumer digital life is going to take a while.
• Apple remains way behind in this space, but I still think that it has time to catch up such that it could be in a position to defend the iPhone and its device ecosystem when the challenge matures.