· 5 min read · ESSAY

Seven days to WWDC — Apple is out of excuses

Six days from now I’ll be watching a prerecorded video. That’s the WWDC keynote in 2026: polished, edited, no stumbles, no awkward silence, no presenters recovering in real time from a demo that just died. I’ll watch it on my Mac, in the evening, with a second screen open and a cup of tea going cold.

I’ve been doing this for fifteen years. And this year, for the first time in a while, I’m genuinely excited.

That should probably worry me.

Why this time is different

Out of excuses.

I spent the past six months writing about Apple’s failures as much as their strengths. Siri’s three-year delay. A battery slider that took four years after third-party developers solved the same problem. The MacBook Neo’s 8GB ceiling. These weren’t nitpicks — they were a pattern: hardware firing on all cylinders, software lagging behind.

But patterns require a reckoning. Apple announced the AI-powered Siri at WWDC 2024. They called it a V1-to-V2 architecture reset. They inked a $1 billion-per-year deal with Google for Gemini because their own model wasn’t ready. Three WWDC keynotes to show something meaningful on the software side. If this isn’t the year, I don’t know what year is.

The “V2 architecture” excuse is gone. The “Gemini deal is being integrated” excuse is gone. There is no remaining architecture to blame. Apple is out of excuses.

What I actually want to see

Two things. Only two.

First: Siri holding a real conversation. Not a command. Not a timer. A multi-turn exchange where the second question references the first. I wrote in February that Siri in 2026 is functionally 2012 software with a better voice. One real multi-turn demo would be enough. I’d be satisfied with that.

Second: something that looks like the autonomous AI agent I described in March. I set up an open-source tool that sent me a morning briefing without being asked — calendar, weather, inbox summary, all on its own. I had to clone a repo and edit YAML files to make it work. Apple has your calendar, your email, your health data, your location, and on-device inference via Apple Silicon. They have everything needed to make this a native iCloud+ feature. Show me that Apple has connected the dots.

If Apple Intelligence becomes something people actually use unprompted, not a feature you hunt for in Settings, that’s a WWDC worth remembering.

But what stage?

What stage?

WWDC has been prerecorded since 2020. There’s no moment where a Siri response hangs and Craig has to fill twenty seconds with a joke. What ships in the video already works. Or at least, it worked that day in the editing suite.

This matters because it changes what a “successful demo” means. Apple can show Siri holding a five-minute conversation in a carefully constructed scenario and it’ll look perfect. The question is whether that same Siri will work on my HomePod six months later, or whether she’ll still be telling me to check my iPhone.

I miss live keynotes. The tension was real. When something goes wrong on stage and you watch a presenter make a real-time decision about how to recover, that’s honest. Prerecorded is cleaner but it tells you nothing about what actually ships.

Touch, foldables, and what Steve said

The foldable iPhone is a fall event story, not a WWDC one. Rumours are circulating, but WWDC is software-first. Hardware announcements come in September.

The MacBook Ultra touchscreen rumour is more interesting, since a touch-capable Mac would need new macOS APIs announced here. Steve Jobs said in 2010 that “touch surfaces don’t want to be vertical.” I noted in the MacBook Neo post that Apple has already moved past that position in practice: A18 Pro chip in a laptop, M-series scaling from a $599 machine to a workstation. The touchscreen Mac is the next push from above.

I’m not a day-one buyer. I want to see how touch actually feels in macOS before I commit. But Apple is allowed to evolve past one specific thing Steve said in 2010. The question is whether they’re doing it because it’s right, or because they need something new to show.

What the past six months told me

I’ve been paying attention.

Battery Toolkit got sherlocked: Apple added a native charge limit slider in macOS 26.4, four years after third-party developers built the same feature and years after iPhone already had it. The result is good. The timing tells you something about how Apple prioritises things.

The MacBook Neo proved that Apple’s chip architecture is extraordinary — an iPhone chip running macOS, matching M1 single-core scores, silent and fanless at $599. The 8GB ceiling is the price of that bet. It won’t change at WWDC, but it’s the hardware context for everything else.

Apple at 50 was my most honest piece: the admiration is real, the patience is wearing thin. The M-series chips prove what Apple can do when execution is right. Liquid Glass proved what happens when it isn’t — a design language unveiled at WWDC and then refined in public for the next year of minor updates.

The pattern is clear. Apple’s hardware team is operating at a different level than the software team. WWDC is the software show. This is the test.

Seven days

I’ll watch the keynote in the evening. If Siri holds a real conversation — even a prerecorded, edited, best-case-scenario one — I’ll take it. That would mean the V2 architecture actually exists and actually works, and the 2024 promise wasn’t vapor.

If it doesn’t happen this year, I’m not sure what I’d write next. I’ve been the patient critic for six months. I’ve given Apple the benefit of the doubt on timelines, on Gemini, on the architecture reset. At some point patience becomes something else.

Seven days. No more excuses.