Apple has released public test versions of iOS 27, macOS Golden Gate, and its other new software, letting everyday users try the next generation of Apple Intelligence and Siri for the first time. The release follows Apple's June 2026 announcement at its WWDC developer conference and several weeks of testing limited to app developers. Engadget

The latest episode of the Engadget podcast, hosted by Devindra and Nathan Ingraham, focuses on the question early adopters are asking: whether the new Siri is genuinely useful in iOS 27 and macOS Golden Gate. The hosts evaluate how well the assistant works in everyday situations, discuss performance improvements in Apple's new software, and talk about the competitive landscape, including OpenAI's rumored smart speaker. Engadget

Apple announced macOS 27 Golden Gate at WWDC 2026 on June 8, 2026. MacRumors The update brings Siri upgrades centered on two ideas: personal context, meaning the assistant can use information from your apps, messages, and habits to give more relevant answers, and on-screen awareness, meaning it can understand what you are currently looking at and act on it. macOS 27 Golden Gate also includes updated Liquid Glass controls, a visual design style Apple introduced in its previous software release. The Verge

On the productivity front, Siri in macOS Golden Gate can write drafts from scratch and give feedback on text you have written. Apple These features put Siri in direct competition with standalone AI writing tools, but built directly into the operating system instead of as a separate app.

The competitive context extends beyond Apple's ecosystem. Bloomberg reported on July 14, 2026 that OpenAI is developing its first consumer hardware device: a smart speaker with no screen that can move around, designed to serve as a humanlike AI companion that lives in your home. Bloomberg The device has built-in AI capabilities and can sync with ChatGPT, according to TechCrunch. TechCrunch

Initial rumors pointed to a 2026 launch for the OpenAI smart speaker, but the timeline has shifted. The device is now expected in 2027. MacRumors

The Engadget hosts' framing of the Siri question is worth noting for anyone tracking how AI assistants catch on. Asking whether an assistant is "actually useful" in a test version is different from checking its feature list. Real usefulness comes down to how fast it responds, whether it pulls up the right information from your personal data, and whether it correctly understands what is on your screen. Public testing will offer a clearer picture than developer testing alone, because it exposes the new Siri to a much wider range of devices, habits, and unexpected situations.

The OpenAI smart speaker points to a different approach to AI in the home, one built around voice and presence rather than the screen-based, app-focused model Apple is refining. The fact that OpenAI's timeline slipped from 2026 to 2027 suggests that building consumer hardware is genuinely hard, even for a company with strong AI capabilities. The delay also means Apple's Siri upgrades will reach the public well before OpenAI's device, giving Apple time to establish user habits around its own assistant.

The broader context here is that two credible visions for how AI assistants become part of daily life are taking shape at the same time. Apple is building AI deeper into its existing operating systems. OpenAI is building a separate, dedicated device for conversation and companionship. Having watched two children grow up alongside successive waves of mobile and voice technology, I have seen how quickly a well-integrated default can displace a superior standalone product. The assistant that is simply there, built into the text field and the system itself, often wins the daily habit over the one that requires buying a separate device. Apple's advantage is distribution — it already has the phones and laptops. OpenAI's challenge is proving that a screenless companion offers enough extra value to sit alongside a phone and a laptop.

What the public testing period establishes is a baseline for whether Siri's new features hold up under real-world conditions. Writing drafts, giving feedback on text, and understanding what is on your screen all depend on fast responses and accurate retrieval of your personal information, not just the quality of the AI model itself. If the public betas deliver on performance, Apple's approach of building AI into the operating system will be validated in the most practical way possible: by users choosing to rely on it instead of reaching for a separate tool.