That is sad. She was with the web from its early days (her first book on the subject is from 1996), taught a lot of people, and bridged several communities: design, web standards, and accessability
David Graeber on "Tyranny of Structurelessness": "[Many] completely misread Freeman’s essay, and interpret it not as a plea for formal mechanisms to ensure equality, but as a plea for more transparent hierarchy."
It is his interpretation presented as if it were an objective truth.
As far as I read the essay, Jo Freeman advocates both equality and transparency. In particular, she makes strong points that equality is impossible without transparency.
Yes, that section seemed to be a bit incoherent to me. However, it was only a brief comment, not even on data science rather but rather the strong confidence in a picking-the-winning version idea of design via A/B tests.
I agree with the "single people can create good software without involvement of a UX expert". That does not only apply to UX experts, but any other expert role (You do not even need a person who sees themselves as a programmer; a person using Excel, Access, Godot… might create something useful).
A lot of roles and best practices are for making software creation work in large organizations.
I disagree with the "dumbing down" idea: "old UX" (90s, early 2000s) focussed a lot on beginner-friendlieness and at this time a lot of people might have seen this sofware as being "dumbed down" from its predecessors. Menus e.g. are a staple of making GUIs more discoverable than command line interfaces, and at that time there were already criticism that urged for building more efficient text based interfaces (See "canon cat" and the "anti mac interface")
No. People were not just "better" and I also think that the core motivation was and is making money.
However, how companies earn money changed. In the 80s and 90s computers were marketed at professionals working in offices. Software was sold shrink-wrapped on disks. It was very costly to change software later (send disks?), you better tested it really, really carefully. The promise was often to make experts more efficient at their tasks: Software was customizable, had macro recorders and had a familiar interface of buttons, menus and docking sub-windows. Computers were used with keyboards and (finally!) mouse: Two very efficient input devices. People (including me) are nostalgic for that time, and that makes memory very selective: We probably remember the best software created by large companies with great teams. We also remember software created for specifically for professionals working with computers, many HN readers will be that people. And, last but not least, there were a lot of terrible UIs too, that we gladly forgot and for every fondly remembered non-standard, fun UI (bryce?) there are sooooo many terrible ones.
A lot of what irks people can be explained with shifts in how software got made, for which audience and for which devices: Having many small teams working independently makes a coherent vision for a product more difficult (but has other advantages), user customization became less important since products were more targeted to non-experts and it also can mess with your automated testing; moving to the web as main way to deliver software meant that some types of interactions (drag and drop, using large amounts of data directly) were harder to create than others; on the web, advertisement and subscription models became popular, leading to very different ways to advertise software in contrast to the former "buy the bi-yearly upgrade in a big box" and the web was more OS independent so the old one-system, one UI-standard did not match anymore.
So, a lot of what "got worse" can be explained by changes in the ecosystem that software-creation and software-use happens in and by the position of people that assume it "got worse".
There are many things that I find worse now than in the 90s but this perspective can help to see if software got worse just for people similar to me and it can also help to find ways to make sustainable changes that fit the ecosystem that exists today.
"Products targeted to non-experts" bears emphasizing: prior to the early 2000s, "using computers" (for any purpose) was a very niche activity, not a mainstream thing that everyone did all the time every day.
Either you were a highly trained professional using a computer for serious work, or else you were a dedicated hobbyist. Either way, you expected to put in significant time to learn and understand how this magically complex machine worked before you could actually accomplish anything. "Normal people" without strong motivation to study the computer simply did not use computers much during that era.
Fast forward to the smartphone era, and "using computers" has become a casual everyday activity for normal people. Companies are now incentivized to produce software for a mass audience with UIs that require as close to zero thought, study, or technical skill as possible.
All the computer nerds were shocked when Google came out with just a single bare text input as its primary UI. Surely we needed the Baroque masterpiece that was AltaVista's UI to ever gain useful work from a search engine; there are so many parameters the user might wish to vary! But no; as it turns out, the new Eternal September mass computer user audience strongly prefers slightly less control in favor of less time spent thinking.
I think you're off by a decade there, in the early 2000s computers at home and in any kind of white collar (and many blue collar) professions were normal and common.
I agree, good things still exist when it is B2B or only designed for a highly trained or professional users. But, when we need to design things for 80 IQ crowd, everything suffers. RGB lighting and diamond studded so to speak in UI/UX field.
My point was that "good" is good only in connection to an ecosystem of tech, values, practices and the people judging what is "good". I personally do not think any group of users "drags the field down".
Yes it does. A certain group of people definitely drags the entire field down. It’s not a nice thing to think about but it’s a fact and the reality IMO. You can continue to lie to yourself but eventually you realize that you need to dumb things down for the below 2 std deviation folks or you won’t get them as users.
