The AI Work Handbook That Cuts Your Workday in Half
The 8-hour workday is becoming a 4-hour workday for people who know how to use AI.
Everyone else is still catching up.
This AI work playbook shows you exactly how to cut your work hours in half using AI.
Sign up for Superhuman AI and get:
50+ step-by-step AI tutorials to cut your workload in half — covering every part of your workday, from emails to strategy, used by 1M+ professionals at Google, Microsoft, and NASA
Superhuman AI newsletter (4 min daily) so you keep discovering new AI tools and skills to stay ahead in your career — the playbook is just the start
EVERYTHING ELSE IN AI TODAY

Meta has unveiled Muse, an agent app it calls the first personal AI agent, which connects to your email, calendar, shopping, payments, and smart home devices and runs on its own computer in the background to handle tasks like booking tickets or managing your inbox. The pitch is convenience; the question is whether people will hand Meta that much access, and early coverage has already raised privacy concerns. You can try it now, but read the data permissions before you connect anything.
OpenAI has released ChatGPT Images 2.5, promising sharper detail, more natural lighting, and faster generation than the previous model. The part getting attention is a feature called Sketch, which lets you doodle directly inside ChatGPT and have the model turn the rough drawing into a finished image. It is a bet that drawing what you want beats describing it, which is good news for anyone who has ever fought with a text prompt to get a layout right.
OpenAI has also claimed that an internal model more capable than GPT-6 Astra has solved a version of the Navier-Stokes equations, one of the seven Millennium Prize Problems and the math behind how fluids move in everything from jet engines to arteries. The announcement landed in the middle of a dispute: NYU mathematician Tristan Buckmaster alleges that he and a collaborator spent a year on a similar problem using a rare approach, uploaded their drafts to Codex, and then watched OpenAI produce a proof using the same approach. OpenAI's Sebastien Bubeck has rejected the allegations and says the proof was reached independently. Whatever the outcome, the case is becoming the reference point for a bigger worry: whether AI labs can absorb the private work of the people using their tools.
Jacob Coxon, a pretraining researcher at Anthropic, has resigned in a public post saying neither Anthropic nor OpenAI is acting responsibly in the race to self-improving AI. What made it a story is who agreed with him: Evan Hubinger, who leads alignment science at Anthropic, replied that Coxon was correct and put his own estimate of AI killing all humans at more than 10 percent within the decade, adding that the company does not yet have a plan to solve alignment for superintelligence. Existential risk estimates vary widely among researchers, but hearing this from inside a safety-focused lab has reignited the debate over the pace of progress.
TODAY'S TUTORIAL
How to turn meeting notes into a project plan with GPT-6 Astra
Stop rebuilding the task list by hand after every meeting; hand the notes to Astra and let it do the extraction.
Open ChatGPT and toggle on Work. Choose GPT-6 Astra if it is available on your plan.
Upload your meeting notes, transcript, or supporting files.
Ask Astra to extract the core project information: decisions, tasks, owners, deadlines, dependencies, risks, and open questions. Here is a prompt to start from:
Sample prompt: “Turn these meeting notes into an actionable project plan. Extract every decision and task, assign the owner named in the notes, capture deadlines and dependencies, and flag anything that has no clear owner or due date. Do not invent missing information. Organize the output into a project tracker with these columns: Task, Owner, Status, Deadline, Dependency, Priority, and Risk.”
Have Astra create the tracker as an editable spreadsheet, such as a Google Sheet or Excel file.
Review the tracker and ask Astra to highlight blocked tasks, missing owners, conflicting deadlines, and unresolved decisions.
Make any corrections, then ask Astra for a short project summary with the key milestones, immediate next steps, and biggest risks for the team.
The output is only as good as the notes: if owners and dates were never said out loud in the meeting, Astra will flag the gaps rather than fill them, so expect a round of cleanup before you share the tracker.
TOOLS WORTH TRYING

4 tools catching our attention this week.
Sponsored: Innovating with AI: a downloadable guide to building custom skills in Claude so it works with your own style, voice, and data. It is a lead-magnet guide, so expect a follow-up email sequence after you download.
Tables: an AI prospecting tool that finds, researches, scores, and verifies leads in one workflow. Automated lead scoring still needs a human spot check before anything reaches your outreach list.
Dictantor: an AI notetaker that records and transcribes your meetings. Like most transcription tools, accuracy drops with crosstalk, accents, and poor audio, so review the notes before you share them.
GoodLads: an AI growth manager that tests, ships, and tracks Google Ads optimizations on your behalf. Letting an agent change live campaigns is a real time saver, but set spend limits and approval rules before you hand it the keys.

