QandA / 1988 / Symantec era / natural language and databases

QandA / 1988 / Symantec era / natural language and databases

When I was 3-4 years old, I played around with computers that looked like this. Magical times. Maybe 10m later I went to the playground. In the 80s era, you'd always find other kids there. Weird, right?

In some ways today and 1988 are much alike. People don't want to learn how to query a database. And use natural language instead.

Symantec in '88 wasn't all about snake oil. They actually produced text processing / database software. QandA or F&A (Frage und Antwort), Q&A, QnA. It was available for Microsoft DOS.

Let me teach you about the database....

In 1988 (to be fair, early 90s), you were able to ask the database assistant natural language questions.

Agentic coding, 1988, DOS era
It doesn't know the word "Median" but it can learn
And just like GenAI it does something...

A lesson to learn

Of course, median / average / math... all the same. But you could teach the tool new words. I understand people don't want to hear that, but:

This would work on a

  • Intel 80286 CPU 16 bit, 8 MhZ
  • 512 KB RAM
  • HD recommended, but not needed. You could store data on a floppy.
  • Multi-user features could have required a network, but ... this was before the internet era

We think that we need LLMs to query databases with natural language. GPUs with gigabytes of memory, parallel vector processing, and so on. Developers in 1988 have proven that you don't. Don't use AI if you can solve things with simplicity.

Share This Article