I think his criticism of the economics and business sense is pretty reasonable, even though he is definitely being pretty credulous about the capabilities of the underlying tech. One of the fun side effects of the diminishing returns in raw scaling is that the competitors are rapidly catching up with the capabilities of ChatGPT, which is going to be bad news for Saltman and the gang. What goes unaddressed is the bigger underlying problem; these systems don't actually do what they're being advertised for and burn an unsustainable and unconscionable amount of money (and actual resources in case anyone forgot) to do it. That's going to be the difference between OpenAI falling apart and being overtaken by another company with better monetization or the entire tech sector facing a recession, and I'm pretty sure the latter is more likely.
TechTakes
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
Fits a pattern I've seen before. Kinda critical of OpenAI and not buying their PR wholesale, but also accepting the framing that AI is some kind of critical foundational tech instead of another shitty magic trick.
I stopped reading Gruber years ago to preserve my blood pressure, but this particular piece is not that bad. In particular the Netscape analogy rang true.
People were formatting text only emails and README text files following loose conventions and Gruber came along, documented common practice and called it Markdown. I don't know if I'd call that "inventing" anything.
Any good invention looks obvious in hindsight.
Before Markdown there was setext, Textile, reStructuredText and atx, which were basically the same thing but less trendy. Gruber had the huge advantage of being popular with the Macbook Starbucks coder crowd.
and many you can tell they're dogshit from the moment they're introduced
so, for instance, in the case of markdown specifically: it's actually a shitshow format to implement. the reference implementation driving the "spec" as done by gruber was (iirc) essentially a big stack of regexes with language-specific behaviours assumed in, and lots of by-design implementation presumptions rather than by-specification dictation
want to make an implementation in another language? better hope you know every single corner case and behaviour!
(this is one of the reasons why markdown impls varied for so many years, and why even now "commonmark" is still trying to fix the issue)
Commenting to push the discussion to more people, since I don't have a take