InvestAI Transparency is overrated

Transparency is over-rated

I wonder how relevant your Stagecoach example would be today. Would internet searches make the actual trip less useful? Granted, there is always something special about an in-person viewpoint, but that argument applies against AI information scraping as well. Would an exhaustive AI-enabled search uncover anything useful compared to an on-the-ground physical visit?

Microsoft used to encourage its employees to blog regularly and officially about their work (transparency = trustworthiness and all that). I suppose an outsider who followed all those blogs could gain some insights that might be investment-worthy, but one thing I realized as a manager at the time was how few non-confidential things were worth writing about. Anything an employee has to say that would be of outside interest is confidential; conversely, anything non-confidential is uninteresting.

There are newsletters/blogs that follow a specific company, often written by a former employee or close business partner who has lots of internal sources. Similarly, there are niche bloggers who follow particular industries or products (I just saw https://pangolinswithpacks.com today, a guy who reviews backpacks). I suppose what we’re trying to do is to surface all of those kinds of information sources and find a way to present the results to decision-makers who care.

Investors don’t know

Stuff I bet investors don’t know

  1. Rodney Brooks, among the most robotics-savvy people in history (inventor of Roomba and more), thinks full self-driving cars are decades away, if that. Click here for the surprising reasons why.
  2. Those glowing reports of how ChatGPT can pass medical exams, law exams, and more? Sounds great, but click here for why it’s more complicated than it sounds.
  3. The main thing business users like about AI is something you can use as an investor. Click here for a case study.

A couple of meta-thoughts about this project:

The best podcasts, I find, are conversations between two people who go wherever their talking takes them. Email newsletters, on the other hand, work best when they’re well-edited, crisp, with clear intro, body, conclusion. At another extreme are WhatsApp or Facebook group chats, which are free-for-alls that anyone joins, and can be somewhat messy — but often informative too.

Maybe we can do something that combines elements of each of these, but fitted to the newsletter genre. Something like:

  1. One big subject, with a few paragraphs of back-and-forth.
  2. Conclude with a question, unanswered, that teases the conversation for the following week.

Week 1: “What is AI?” * Short definition, with back-and-forth explaining LLM basics and terminology, why Nvidia is rising, and the “no backtracking” limitation. * Leading question for next week: e.g. “So what do investors care about and why?”

A few other thoughts:

  1. Should we spice up the formatting a bit? At least to me, an email longer than a few paragraphs can be intimidating to read. Maybe we should highlight key phrases throughout, or insert pull-quotes or other features that allow easier skimming.
  2. We don’t know the long-term best format, so why not be honest and up-front with our readers. Have this conversation in front of them.

A key aspect of engagement is to create a story that draws people in and makes them want to hear more.

About Us

Since graduating from Wharton with Sami, Richard Sprague has been a senior executive at numerous technology firms, including Apple (where, as an early employee in Japan, he was responsible for the launch of numerous Mac software products) and Microsoft (where he led the Beijing-based development of Mac Excel). After more than 10 years in Japan and China, he now lives in Seattle where he works with startups building what he calls “personal science”: applying the latest technology to personalized health and wellness.