Putting the strategy cart in front of the customer horse

One of the first significant projects I worked on was Windows for Workgroups (WfW).  It was a version of Windows 3.1 with networking built in — clients to access remote files, a peer file server, a peer print server, and other network utilities.  It was a good product — the networking was easy to install; it was all protected-mode networking, so it was fast and didn’t consume scarce low memory; and the networking features were fully integrated into the Windows UI and easy to use. 

As networking geeks, we on the team were all proud of the product.   And it was a strategic product for Microsoft — we were competing in networking with Novell, and in groupware with Lotus. Microsoft pushed the product as a competitor to Lotus Notes, which was pretty misdirected!.  But customers were mostly excited about WFW for a feature that had nothing to do with networking — WFW had a protected-mode filesystem that made disk access much faster, and thus made Windows faster.  That was a great lesson — you can have all kinds of fancy ideas about product strategy, but customers generally buy products for clear and straightforward reasons.

Later in my career, I worked on Internet Explorer, Microsoft’s browser to compete with Netscape Navigator.   IE versions 1 and 2 were just OK, but IE 3 was a great product.  It supported internet standards like CSS, was fast, had great Java and JavaScript support, looked great, and included a great bundled mail and news client.   It was an excellent browser.  And people used it because it was a darn good browser. 

But we got full of ourselves and got out over our skis on the next version, IE4.   We got all excited about pushing HTML deeper into Windows, with Active Desktop integrating the web right into the Windows Desktop.   This was a very strategic move by Microsoft and technically super cool, but it wasn’t a set of features customers asked for, and ultimately, many of those features were dropped in later versions.  The product was still good enough, but it wasn’t a significant step forward as a browser.  Probably one of the larger errors in my career.  I made team errors on this project as well; it was a learning experience for me.

Anyway, I’ve been trying out ChatGPT Atlas this week.   An interesting idea, but it is not yet a really good browser, nor is it a great AI workspace.  Some browser features have been removed or hidden to emphasize ChatGPT's features — for instance, there are no tab groups, which I use a lot, and there is no easy way to share a link with someone.  And the AI features are scoped to the current page you are looking at, which is rarely the scope I want for an agent; I want it to be looking across multiple sources.  Atlas does nothing to raise the bar for organizing AI work — it exposes ChatGPT projects as a sidebar, but they're very weak as project containers.  

Atlas puts OpenAI's strategy out in front of user needs, and that probably won't work.   I do want a better browser, I do want AI to be easier to use, I desperately want a better way to organize my AI projects, but this isn’t it yet.  The Diff posed an interesting question — why a browser and why not an office suite?  

College Football

When I am not dorking around with software, I am following college football.   And what a great year for college football.  The Athletic nails it — College football is absolutely unhinged right now. It’s exactly why we love it:

People involved with college football — coaches, administrators, fans, all of us — are simply not normal.  And thank goodness for that. Who wants normal?

College football is at its best when the passions and rivalries are completely out of control.  And the flood of money into the sport has just made it zanier — we are seeing an unprecedented number of mid-season coach firings.  NIL money has pushed expectations sky-high, without regard for rationality.  It is chaos, and that is what makes the sport great.

If you want well-executed play by the best athletes in the sport, the NFL is great.  But if you like drama, and passion, and excess, college football is the sport for you.

Feds vs the States

Paul Krugman wrote recently about the contempt that the Trump administration is showing for all of us, but particularly for rural America.  While Trump is dithering around with his ballroom and other interests, he is delivering blow after blow to the states that voted for him, and the month of November will be very painful with the loss of SNAP payments and the new health insurance pricing.

Mike Brock has written about the fundamental clash between the reactionary right and the democratic left.  While I don’t buy all of it, it is pretty clear that the reactionary federal government is not working in our favor, and we need to work hard to counter it with structures that defend our freedoms and rights and allow all our citizens to flourish.  

And so I am 1000% in favor of the soft succession discussions kicked off by Christopher Armitage and continued by Qasim Rashid.  The Federal Government has quit acting in our interests, and it is appropriate for our states and cities to start taking action to protect our rights.  Qasim’s summary:

When the federal government refuses to deliver on its end of the bargain—when it blocks food aid, infrastructure funds, and disaster relief to punish political opponents—states have the right to withhold their payments and redirect that money to protect their own residents.

Every time I write a tax check to the IRS and I mail it off to Cincinnati, it makes me angry. The West Coast states should start collecting those funds locally and then using that as leverage.

Shorts

San Francisco is back as the world’s leading tech hub per The Pragmatic Engineer. I need to try out Wispr Flow.

The scale of the Amazon layoffs is stunning.  As Marc Love says:

I’m a firm believer that any company that does mass layoffs in response to AI is in decline.

It’s a self-indictment of your:
– ability to train your employees
– remaining market opportunity
– imagination as company leaders