I Nearly Let AI Talk Me Out of My Own Idea

Hello,

Hi, I'm Simon, and welcome to another edition of Plain AI. Here's what's worth knowing in AI this week: the stories and what's coming up.

P.S. - sharing this issue with a couple of people gets you a free guide. More on that partway down.

I spent some time this week trying to build out a new process at work. After thinking about it over a few days, I started to create a presentation that captured all of my thoughts, reasoning and the future plan. When I started, I had a pretty clear vision of what I wanted to do. As is often the case these days, I also turned to AI to help me produce this by being a sounding board. I created a new project to work in and gave it some context and specific instructions. Initially, this was all very helpful, and I was getting some good points that helped me update my presentation. However, something interesting happened when I connected my deck and asked it to review whether I was heading in the right direction. As is typical, I received some good praise complimenting me on the start I had taken, so I was happy. But as I asked it more questions, continued the conversation, and made further edits to the deck, I suddenly found myself a little confused as to what I was ultimately trying to say and what my plan was. I wouldn't go as far as to say I was going around in circles, but I was definitely questioning what I had produced so far. A further interesting point was that when I stepped back even further into my original problem and gave it the full scenario of the challenge, I got a response along the lines of "this changes my point of view somewhat…". And to be honest, its responses were much better, and I got some clarity back.

This whole thing made me question a couple of things. First, given that I was very clear on what I wanted to produce at the very beginning, was it habit or laziness that I still turned to AI to help me out? Did I really need it? Second, it was only when I explained the very core of my problem in detail that I truly got the kind of responses I was after.

After working for the best part of a full afternoon, I had my plan and a solid presentation explaining my new approach to present to colleagues, but I do have a slight nagging feeling that, for once, AI didn't actually save me time here — and in fact nearly derailed my original thought process, causing me to use more time than I might have needed.

Let this be a reminder that just because we have AI tools at our disposal, we shouldn't turn to them for every single challenge. Pay attention to your own thoughts and whether you actually need a helping hand or not. It would be a shame if, as humans, we lost the ability to handle original thinking without using AI as a crutch.

I also want to announce this week, that I’ve collated a number of the AI use cases that I use at work. I’ve put all of these into one document that includes how to replicate them in tools like Claude or ChatGPT. As a bonus, I’ve also made a simple prompt cheat sheet that details the exact 10 prompts I use in the examples.

Private real estate has long been a cornerstone of institutional portfolios, but individual investors have often had limited access to the asset class - or have had to invest through pooled funds and intermediaries.

Lightstone DIRECT offers a different approach. Accredited investors can invest directly in select real estate opportunities alongside Lightstone, a vertically integrated owner/operator with more than four decades of experience and a $12 billion real estate portfolio.

We invest 20% or more of the equity in every offering, aligning our interests with yours from day one. Our in-house team oversees acquisitions, asset management, and investment execution, with property operations managed by Lightstone and, where appropriate, experienced third-party operators.

No blind funds. No intermediaries. Just direct partnership.

Accredited Investors only. $100k minimum investment.

Explore Opportunities

^{This communication is for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to buy any securities. Any such offer or solicitation will be made exclusively through the definitive offering documents. All investments involve risk of loss, including the potential loss of principal. Past performance is not a guarantee of future results. Any targeted returns or projections are forward-looking statements, are based on current assumptions, and are not guarantees of future performance. Actual results may differ materially.}

Ryanair signed a five-year deal with Google Cloud on 12 August, giving the tech giant access to its operations across crew scheduling, fleet maintenance, and disruption management. The airline will deploy Gemini Enterprise and DeepMind models — including AlphaEvolve and WeatherNext — across its entire workforce of 35,000 people. Ryanair already uses AWS and is deliberately running a dual-cloud strategy to avoid a single point of failure. The deal is part of an ambition to carry 300 million passengers a year by 2034, supported by an order for 300 Boeing 737s, according to Computer Weekly.

Why it matters: This is one of the more concrete AI commitments from a major European business in recent months — not a pilot, not a proof of concept, but a five-year operational bet on AI-managed scheduling for a carrier that runs on razor-thin margins. Other operations-heavy industries — logistics, hospitality, manufacturing — will be watching to see whether it delivers.

When Spirit Airlines collapsed earlier this year, its business data was put up for auction, which Google won, paying $10 million for a treasure trove that includes more than 100 million company emails, 500 million Microsoft Teams messages, 30 million lines of code, and detailed employee productivity data. Customer records were excluded. Google says the data will be scrubbed of personal identifiers before use in AI model training, according to Bloomberg and Forbes. A US bankruptcy judge approved the sale this month.

Why it matters: This is a useful reminder that corporate data has real commercial value, and that value doesn't disappear when a company goes under. For any business evaluating AI training data strategies, or for employees wondering where their work communications end up, this is an instructive case study in how data assets are treated in insolvency proceedings.

A wave of reporting this week — led by CNBC on 23 August and TechCrunch on 19 August — documents a widening trust gap between employers pushing AI adoption and workers deeply sceptical of it. A Pew Research study found 52% of Americans are "more concerned than excited" about AI in daily life, up from 37% in 2021. CNBC found that 53% of workers across demographic groups worry AI will eliminate a job in their household. Meanwhile, Anthropic CEO Dario Amodei acknowledged this month that negative public sentiment is "fundamentally a crisis of trust." Corporate communications strategies, CNBC found, are lagging badly — very few companies have what one expert called "an honest plan for what AI is doing to their people."

Why it matters: The gap between executive enthusiasm and employee anxiety is a practical management problem, not just a PR one. Organisations that push AI tools without addressing the trust deficit risk higher attrition, lower adoption rates, and — as the CNBC piece notes — reputational damage that is now showing up in IPO filings as a disclosed risk factor.

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: