AI News You Should Know About This Week: Apple launched the iPhone 18 Pro this week with a genuinely rebuilt Siri, built on Apple Intelligence and described by Apple as more personal, capable, and conversational. The new Siri pulls context from your messages, emails, and photos to actually answer what you're asking, has on-screen awareness so it can act on whatever you're looking at, and includes a new camera mode where you point your phone at something and ask Siri about it. It's rolling out in beta with iOS 27 on September 14th, English only at first, with French, Japanese, Korean, Portuguese, and Spanish coming in October. Some of it runs on-device, and some routes through Apple's private cloud servers with daily limits on those server-dependent features. It's paired with new hardware too: the A20 Pro chip, an upgraded 48-megapixel camera, and faster charging. Ashley's take: this is the first version of Siri that sounds like it could genuinely save you time instead of just being a novelty, but it's launching in beta, in English only, with usage caps on the good stuff. If you're the type to hand your calendar and inbox over to an AI assistant, give this a few months before you restructure how you work around it. Then: OpenAI shipped ChatGPT Images 2.5, generating images up to 50% faster than the previous version, with sharper results and more natural lighting and texture. The bigger upgrade is precision: it can make a focused edit to one part of an image without messing up the rest, and it holds up across multiple rounds of edits instead of degrading. New tools include a sketch feature for drawing a rough reference directly in ChatGPT, templates for things like posters and product photos, the ability to leave comments on an image to guide edits, and prompt sharing so someone else can try your exact prompt with their own photo. It's live now for everyone on ChatGPT, ChatGPT Work, and Codex on desktop, mobile, and web, with two new API options for developers: Flare (faster) and Sunburst (built for precision work). Ashley's take: this is the update that matters for anyone making their own marketing assets — fewer redo cycles because edits actually stay consistent, and templates for the stuff you make over and over, like product shots or social graphics, mean less time waiting on a designer for something small. If your team is already using GPT, it's worth testing on your next batch of content. And a reminder: give credit where credit is due and support real graphic designers — there's a balance here, and with AI slop everywhere, it's worth asking whether a project actually calls for a designer before you generate your next poster. Finally: Meta launched Muse, a personal AI agent that lives inside WhatsApp and proactively helps with your goals. It's not just there to answer questions — it can send emails, book travel, fill out forms, and even negotiate on your behalf, learning your preferences over time, like turning a saved Instagram recipe into a grocery list or remembering someone's dietary restrictions when you're planning an event. It also has its own standalone app at news.ai, with support for Meta's AI classes coming soon. It's rolling out now in the U.S. on iOS, Android, and web, free for basic tasks with a paid tier for more. On privacy, Meta says each user's data runs in an isolated, secure environment overseen by a "sentinel" agent that has to approve any action that touches the internet. You can control access app by app, it can't see your passwords directly, your data isn't used for ad targeting, and you can opt out of training or ask it to forget specific information. Ashley's take: handing an agent the keys to book travel and send emails on your behalf is a real time saver for anyone drowning in admin work, and the permission controls Meta built in are more thoughtful than expected. That said, it's U.S.-only for now, and she'd want to see it in the wild for a few months before letting it touch anything financial. Worth watching, not worth rushing into. Also this week: the full recap of Apple's keynote — everything announced on stage that didn't fit into the regular show — is available now as a bonus episode exclusively for subscribers on Apple Podcasts, along with ad-free listening on every regular episode. AI news you should know about, in under 10 minutes. In this episode: (00:04) Welcome and this week's headlines: iPhone 18 Pro/Siri, ChatGPT Images 2.5, and Meta Muse (00:33) Subscriber bonus: full Apple keynote recap episode on Apple Podcasts (01:15) Apple launches the iPhone 18 Pro with a rebuilt Siri (01:39) What the new Siri can actually do (02:04) Rollout timeline, on-device vs. cloud processing, and new hardware (02:28) Ashley's take: promising, but give it a few months (02:50) OpenAI ships ChatGPT Images 2.5 (03:11) New editing tools: sketch, templates, comments, and prompt sharing (03:39) Ashley's take: fewer redo cycles, and a note on supporting designers (04:32) Meta launches Muse, an AI agent inside WhatsApp (04:57) What Muse can do and how it learns your preferences (05:19) Privacy controls and rollout details (05:43) Ashley's take: worth watching, not worth rushing into (06:07) Wrap-up and where to follow the show Links: Apple launched the iPhone 18 Pro with a rebuilt Siri (Apple Press Release)OpenAI shipped ChatGPT Images 2.5 with generation up to 50% faster (OpenAI Press Release)And Meta dropped Muse, an agent that books your travel from inside WhatsApp. (Meta Press Release)Follow Ashley at @ashleyrcoffey89 on Instagram. Send us a topic you want covered, and follow the show for AI news you should know about, in under 10 minutes! Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
