Today I stumbled across a thought-provoking post from Kyle Saunders of Colorado State University. In it he proposes that changes in the job market, driven in part but not completely by AI, are creating a choice for institutions between offering the traditional four year on campus model that Saunders calls Elite Mentorship, and an accelerated competency focused degree which he labels a Credential Factory. These two paths create graduates with very different experiences and skill sets.

Disclaimer — part one: Although I don’t currently work in higher ed, I spent more than two decades working in the sector after growing up a faculty kid at a public regional. I have four degrees, a bachelors and three masters. My last masters was employability driven and I completed it in 16 months, not quite a speedrun (to borrow Saunders’s term) but focused on efficiency rather than the experience.

Saunders points out that many employers don’t seem to be making a distinction between the two sorts of college experiences and treat the resulting credentials as equivalent. In doing so they undermine what he calls the formation layer, that experience of being guided over an intellectual hurdle during a faculty office hour or talking through a complex question in a class discussion or late at night in a residence hall lounge. It is this formation layer and wanting to be a part of creating it for others that drew many people , myself included, into higher education as a career. What happens when those paying the bills, either governments, parents, or employers (in the form of better paying and more secure jobs for college graduates) don’t actually value the formation layer. Instead, they value the sorting, taking the fact that you gained admission to and then completed a degree at ________ University as evidence of something. If the formation layer is truly important, we as a society would put a lot more effort into measuring it.

For all the think pieces about trust and value/ return on investment, maybe this is a key to understanding higher education’s current challenge. Perhaps for a long time, only a small proportion of stakeholders have prioritized the formation layer. but colleges have been able to “bundle” this with the things stakeholders care about like employable skills in the same way that general education requirements bundle low demand subjects like art and philosophy with higher demand majors. Now, just as subjects like foreign languages are shed from the curriculum in the name of efficiency, intellectually transformative experiences are dropped in favor of skills that can be quickly developed and demonstrated, in the name of efficiency.

Disclaimer — part two: I was a music and French major and then a music and French graduate student. It hurts every time I read about a music or foreign language department being sacrificed, and that seems to happen quite regularly these days.

Where does this leave room for intellectual transformation? Counterintuitively, AI may be a great opportunity. Now that your LLM of choice can create a perfectly serviceable email or presentation slide deck, the differentiator of the well educated is the ability to do deep and thoughtful analysis, the sort of thing the formation layer helps you learn to do. If we can measure that (admittedly no easy task) and convince employers that it’s worth hiring for (a perhaps even less easy task), there may be a way to put at least some of the bundle back together.

Yesterday I saw a post from Ben Werdmuller that sent me down a bit of a rabbit hole on the future of the open web.  Anil Dash presents a grim prognosis for the open web. Julien Genestoux points out that we ended up here by repeatedly choosing convenient and cheap (free) instead of open and portable.  He concludes with the idea that, “The web needs netizens again.” To his credit, he lists out some specific ideas on what that looks like.  Ben notes that ,as much as openness advocates don’t like to talk about funding, “If people are going to build good things, they need to be able to eat and pay rent so they can keep doing it.” All of this has me thinking.  What are the barriers to this sort of future and how might we  overcome them.  Julien points out that people pay as much for a streaming service every month as they would to support an open platform.

I’m thinking for the moment particularly about the social space. The trajectories of Activitypub and ATProto tell a difficult to hear story.  Friction is really important.  Even though Mastodon has been around much longer, Bluesky has almost three times the total number of users. My best guess is that this comes down to a trust issue.  Unless you know someone who runs a mastodon server, you’ll probably choose one of the biggest sites. To this point, ATProto alternatives to Bluesky are limited.  With size and hopefully stability come infrastructure demands.  Even without a need for ROI (Bluesky has taken VC money), there are bills to be paid and we’re right back at Ben’s questions about funding. Ideally the social networking space would be very decentralized.  I imagine the typical server as having a dozen users. Servers would be as populous as a city block.  

There is another option, of course.  You can do it yourself. I run a home server, so this approach must appeal to me.  However, while I briefly experimented with running Mastodon on a VPS, I abandoned DIY social networking. I did so because I wasn’t confident in my ability to be my own CISO, and I was concerned about the ramifications of people I might not know sending content to my server. Is there something we can do to lower the barrier to entry?

I also saw Dave Winer’s post suggesting that we just do this all with RSS. A chat could just be a subscription to a set of RSS feeds.  I’ve thought about using something like River to aggregate my Bluesky and Mastodon timelines.  The problem is there’s no easy way in RSS to reply to posts.  We feel close to and yet far from the point where your social network could live on your own device.