Samsung Unpacked lands July 22, GPT-5.6 goes default & the AI-picking playbook

Samsung Galaxy Unpacked is officially set for July 22, and it's shaping up to be a foldable-heavy show. Expect the Galaxy Z Fold 8 alongside a larger Z Fold 8 Ultra, a lighter Z Flip 8 (around 180g), the Watch 9 series, and possibly a first look at Samsung's Galaxy Glasses. The creases keep shrinking — but so does your wallet, with the Fold 8 starting near $1,900 and the Ultra around $2,100. Worth watching the value, not just the hype. 📱

The problem

Foldables have gotten thinner and more refined every generation, but that progress has consistently come with rising prices, making 'better' and 'more expensive' arrive together every single cycle.

What changed

Samsung Galaxy Unpacked is officially set for July 22 with a foldable-heavy lineup — the Galaxy Z Fold 8, a larger Z Fold 8 Ultra, a lighter Z Flip 8 (around 180g), the Watch 9 series, and possibly a first look at Galaxy Glasses — with the Fold 8 starting near $1,900 and the Ultra around $2,100.

The catch

The creases keep shrinking, but so does the value proposition at these prices — a thinner hinge doesn't automatically justify a bigger bill.

What it means for you

Before you get caught up in the reveal, decide your actual price ceiling now — at $1,900 to $2,100, this is a purchase worth budgeting for deliberately, not one to impulse-preorder on hype.

GPT-5.6 is now the default model inside ChatGPT. It's stronger at coding, reasoning and long context (a full 1M-token window), and the balanced 'Terra' tier reportedly matches the older GPT-5.5 at roughly half the cost. The best part: there's no setting to flip — the upgrade just landed for everyone. 🤖

The problem

Model upgrades usually require users to manually opt in or discover a new setting, meaning a lot of people keep using an older, weaker model without realizing a better one is available.

What changed

GPT-5.6 is now the default model inside ChatGPT for everyone, with no setting to flip — stronger at coding, reasoning, and long context with a full 1M-token window, and the balanced 'Terra' tier reportedly matches the older GPT-5.5 at roughly half the cost.

The catch

Because it rolled out silently, some users may not notice their experience changed at all, even though the underlying model and its cost profile did.

What it means for you

If you use ChatGPT regularly, you're already on GPT-5.6 whether you noticed or not — worth re-testing tasks you'd previously written off as too weak for the model, since the upgrade is free and automatic.

One Gmail trick most people miss: add a '+tag' before the @, like yourname+shopping@gmail.com. Mail still reaches your normal inbox, but now you can auto-filter it by tag and, crucially, spot exactly who leaked or sold your address when spam starts arriving at a tagged version. Zero setup, works today. 💡

The problem

When spam starts arriving at your email address, it's almost impossible to know which company or website actually sold or leaked it — every service you've ever signed up for is a suspect.

What changed

Adding a '+tag' before the @ in your Gmail address, like yourname+shopping@gmail.com, still delivers to your normal inbox, but lets you filter by tag and identify exactly which signup leaked your address once spam starts arriving at that specific tagged version.

The catch

Some sign-up forms reject the '+' character as invalid, so it doesn't work universally — and once you know who leaked it, you still can't undo the leak itself.

What it means for you

Start tagging new sign-ups today, especially ones you're wary of — it costs zero setup and turns 'who sold my email' from a mystery into a quick filter search.

'Which AI is best?' is the wrong question in mid-2026. The honest split: Claude leads for coding and agents (Fable 5 tops SWE-Bench Pro), ChatGPT is the everyday and voice pick (GPT-5.6 is now default), and Gemini wins if you live in Google apps and search. All three now handle 1M-token context for long documents, and Gemini 3.5 Pro is still delayed. There's no universal winner — keep two open and match the tool to the task. 🤖

The problem

The 'which AI is best' question keeps getting asked as a single-answer competition, when the honest answer depends entirely on whether you're coding, chatting, or living inside Google's ecosystem.

What changed

The honest mid-2026 split: Claude leads for coding and agents (Fable 5 tops SWE-Bench Pro), ChatGPT is the everyday and voice pick with GPT-5.6 now default, and Gemini wins if you live in Google apps and search — with all three now handling 1M-token context for long documents, while Gemini 3.5 Pro remains delayed.

The catch

Gemini's positioning here is weaker than it might otherwise be specifically because 3.5 Pro hasn't shipped — its ranking could shift the moment that changes.

What it means for you

Keep two of these open rather than committing to one: Claude for coding and agents, ChatGPT for daily use and voice, Gemini if your work already lives in Google apps.