Could You Go a Day Without AI?

Hello,

This is Simon with the latest edition of The Weekly. In these updates, I share key AI related stories from this week's news, list upcoming events, and share any longer form articles posted on the website.

P.S. — sharing this issue with a couple of people gets you a free guide. More details below.

I was part of a conversation this week where we playfully discussed the idea of not using AI at work for a day, or even a week! The suggestion came from the fact that a lot of time is now spent just working with Claude, whether that's crafting prompts, re-prompting, connecting tools or writing Skills. Could we have done the task the way we originally did it, without AI, or have we already become completely wedded to it? We then went on to discuss the potential energy savings if the whole nation avoided AI tools for the day. A single prompt uses surprisingly little electricity, with Google putting it at about the same as watching TV for nine seconds, but it adds up quickly when millions of people are doing it all day long.

However, I mentioned that it is one thing to stop using a chat tool like Claude or ChatGPT, but quite another to stop all of the scheduled activities I have set up, or even to have a clear idea of all the different tools I have connected at some point over the last few years. There's probably a ton of AI that I rely on and make use of today that I've stopped being conscious of. And this is actually a problem.

Think about the data and information that you might have sitting in an AI tool somewhere. Is it secure? Could your colleagues or family see it?

If you’re not sure where to start, I’d begin with the places AI tends to hide. Have a look at the apps and accounts you have connected to your chat tool, any tasks you have set up to run automatically, and any browser extensions or meeting tools that record and summarise your calls. It’s also worth checking for shared chat links that are still live, as these can usually be opened by anyone who has the link. For anything you no longer use, delete the old conversations, remove the connection and close the account.

I’d suggest you take an hour this week to run a quick audit of the AI tools you are using, and stop anything you no longer need. The main benefit is your own security, although it may be a small win for the environment too.

When you type a prompt, you summarize. When you speak one, you explain. Wispr Flow captures your full reasoning (constraints, edge cases, examples, tone) and turns it into clean, structured text you paste into ChatGPT, Claude, or any AI tool. The difference shows up immediately. More context in, fewer follow-ups out.

Used by teams at Microsoft, Amazon, and Notion. Try Wispr Flow free on Mac, Windows, iPhone, and Android.

Start flowing free

AI stocks wobble, and a $50bn debt pile takes the blame

AI-linked shares fell sharply last week, with the Philadelphia semiconductor index down about 12% in October, according to Reuters. CoreWeave, an AI cloud provider carrying roughly $50bn of debt, saw its shares tumble, and Deutsche Bank and RBC both pulled back on tech recommendations. Goldman Sachs expects the five big AI hyperscalers to lift capital spending to about $1.2tn next year, which puts this week's Meta earnings in the spotlight.

Why it matters: The question markets are now asking is whether AI spending pays back, and who is funding it with debt. If you're signing multi-year AI contracts, it's worth knowing how financially exposed your vendor is.

EU AI Act: the compliance clock has been moved, not stopped

The Digital Omnibus changes to the EU AI Act push back the main deadlines, with Technology.org's 13 October guide setting out what applies between December 2026 and August 2028. The Parliament's amendment (passed 423 to 57, according to Freevacy) moves standalone high-risk systems to 2 December 2027 and embedded safety systems to 2 August 2028, with AI watermarking rules from 2 December 2026. Stephenson Harwood's October briefing notes that Article 50 transparency obligations are already in force.

Why it matters: UK firms selling into the EU get breathing room on high-risk systems, but not on transparency. If you generate content with AI, check now what labelling you'll need.

HubSpot cuts 660 jobs and says AI isn't the reason

HubSpot is cutting about 7% of its workforce, roughly 660 roles, as it reorganises around helping customers get results with AI, according to the Boston Globe. CEO Yamini Rangan said the cuts were not driven by AI-related efficiencies, and cofounder Dharmesh Shah later apologised after his LinkedIn farewell post was widely criticised as curt. Fast Company's October roundup shows the same pattern elsewhere: Netflix is cutting around 1,000 jobs and Amazon fewer than 1,000, and neither cited AI.

Why it matters: Job cuts are rarely as simple as "AI replaced them", and leaders are increasingly careful about saying so. Expect your own staff to read AI into any restructuring, so be clear about what is and isn't driving decisions.

Thanks for reading, and see you next Thursday.

Simon,

Was this email shared with you? If so subscribe here to get your own edition every Thursday.

You can also follow me on social media: