Whenever this story (and all the related stories about PARC) re-surface, I am (writing now in late 2026) mostly sad about the incredible language that should have 'won' .. well .. everything!

The term 'footgun' unfortunately matches all too well: something so great, so powerful yet most of the time, even in the hands of well-intended folks, proceeds to shoot oneself in the foot. I was for more than a decade a professional Smalltalk developer (Telecommunications, Financial/Derivatives) but we were replaced by technically inferior systems, mostly Java, JavaScript and even VisualBasic. The HR person at the Financial company explained (after my exit interview!) that for the cost of a single Smalltalk developer, they could hire 4 or 5 lesser experienced programmers (less cognizant of how systems actually worked!). The HR folks preferred these hires: more selection, cheaper . etc. Sigh!

Am I a horrible person to enjoy so much (schadenfreude!) the upcoming AI/LLM footgun massacre?

▲

pjmlp 4 hours ago | parent | next [-]

Blame IBM and other Smalltalk vendors.

They were the ones that almost on day one pivoted their Smalltalk offerings into Java.

Eclipse still has the Visual Age Smalltalk object browser.

If you look into OS/2 documentation, and SOM (OS/2 version of COM), you will find out Smalltalk on OS/2 had a role similar to .NET on Windows.

I think AI is the revenge of Lisp, dealing with Python without JIT, and the whole 4 GLs, CASE tooling that tried to automate programming in the 1990's, this time around it goes.

▲

bch 3 hours ago | parent [-]

> I think AI is the revenge of Lisp, dealing with Python without JIT, and the whole 4 GLs, CASE tooling that tried to automate programming in the 1990's, this time around it goes.

Perhaps it’s because I haven’t had a coffee yet, but I’m not following/parsing this last paragraph - can you expand?

	▲	pjmlp 2 hours ago \| parent [-]
		Sure. Lisp was the AI language, alongside Prolog (see Japan's 5th generation computing project). Instead of having nice Interlisp-D and Symbolics machines with their graphical tooling, hot code reloading and machine code generation, we got Jupiter notebooks and C/C++/Fortran libraries on a pure interpreted language [0]. Additionally in the 1990's there were several efforts to move into the next level of programming, with more human like languages, CASE tooling, RUP method (with UML), most of them failed, SQL is one of the survivors, think its declarative nature. https://en.wikipedia.org/wiki/Programming_language_generatio... https://en.wikipedia.org/wiki/Computer-aided_software_engine... https://en.wikipedia.org/wiki/Rational_unified_process So while not all of those ideas caught on as expected, now we get to program in natural language, use Markdown waterfall plans for agents, integrate them with Web services, seat reading email and Internet surfing until their activities are ready for review. In low code/no code cloud platforms like Workato, Boomi and co, classical programming is already past tense. [0] - Granted now JIT is being part of CPython roadmap, and there are some nice IDEs for Python actually.

▲

dharmatech 2 hours ago | parent | prev | next [-]

Interesting!

Were you using IBM Visual Age Smalltalk?

▲

rootnod3 5 hours ago | parent | prev | next [-]

I for one already have the popcorn at the ready. Can't wait for it.

▲

CrimsonRain 5 hours ago | parent | prev [-]

You're not a horrible person to enjoy that. However, the schadenfreude is only in your head. 2026 is the most amazing year of building software in human history, so far.

▲

gobdovan 4 hours ago | parent | next [-]

Well, programming feels that it will get closer Smalltalk. For webdev you get pretty mature HMR, you look at the interface and sometimes just point at things and tell a model to modify it. So pretty close to live medium +short feedback loops.

If the big labs wouldn't have the incentive of hiding the code and making you rely solely on the models, you'd probably also see more inspection tools and contextual changes when the model does them.

▲

dormento 3 hours ago | parent | prev [-]

It depends, IMHO. If op is talking about "programming", then I think they are right. If they are talking about "development", then I think you are right.

	▲	CrimsonRain 2 hours ago \| parent [-]
		I still don't agree. Nothing is stopping you from doing programming without AI and enjoy it like an art form just like people still paint even though they can just print. On the other hand, the amount of languages, tools, paradigms available are crazy and there's always something for everyone to enjoy.