Twice, well done!
Yep, looks like that is it. Recent patent from one of the founders: https://scholar.google.com/citations?view_op=view_citation&h...
[0] https://www.bloomberg.com/news/articles/2025-05-21/openai-to...
Pardon the AI crap, but:
> ...in most people, when they "talk to themselves" in their mind (inner speech or internal monologue), there is typically subtle, miniature activation of the voice-related muscles — especially in the larynx (vocal cords/folds), tongue, lips, and sometimes jaw or chin area. These movements are usually extremely small — often called subvocal or sub-articulatory activity — and almost nobody can feel or see them without sensitive equipment. They do not produce any audible sound (no air is pushed through to vibrate the vocal folds enough for sound). Key evidence comes from decades of research using electromyography (EMG), which records tiny electrical signals from muscles: EMG studies consistently show increased activity in laryngeal (voice box) muscles, tongue, and lip/chin areas during inner speech, silent reading, mental arithmetic, thinking in words, or other verbal thinking tasks
So, how long until my Airpods can read my mind?
This is an interesting acquisition given their rumored Echo Show / Nest Hub competitor (1). Maybe this is part of their (albeit flawed and delayed) attempt to revitalize the Siri branding under their Apple Intelligence marketing. When you have to say the exact right words to Siri, or else she will add “Meeting at 10” as an all day calendar event, people get frustrated, and that non-technical illusion of the “digital assistant” is lost. If this is the model of understanding Apple have of their customers’ perception of Siri, then maybe their thinking is that giving Siri more non-verbal personable capability could be a differentiating factor in the smart hub market, along with the LLM rebuild. I could also see this tying into some sort of strategy for the Vision Pro.
Now, whether this hypothetical differentiating factor is worth $2 billion, I’m not so sure on, but I guess time will tell.
https://www.macrumors.com/2025/11/05/apple-smart-home-hub-20...
The same basic developments of mental science that had brought about the development of the Seldon Plan, thus made it also unnecessary for the First Speaker to use words in addressing the Student.
Every reaction to a stimulus, however slight, was completely indicative of all the trifling changes, of all the flickering currents that went on in another's mind. The First Speaker could not sense the emotional content of the Student's instinctively, as the Mule would have been able to do – since the Mule was a mutant with powers not ever likely to become completely comprehensible to any ordinary man, even a Second Foundationer – rather he deduced them, as the result of intensive training.
Intel went through a phase in the 2010’s of buying gobs of companies with fancy tech and utterly failing to integrate those acquisitions.
And even more fundamental, Intel rested on its laurels of having good hardware and got bit hard in the end. Something similar seems to be happening at Apple.
...how is this not at the top of the page?
I wish the iphone had word prediction and autocorrect that was from the previous centruy
