5 ms·

I’ll go against the grain and say low code will eat software engineering. But not the conventional low code platforms like web flow or retool or power apps. I

I’ll go against the grain and say low code will eat software engineering. But not the conventional low code platforms like web flow or retool or power apps. It will be AI native and built in a visual+conversational manner.

toomuchtodo 3y ago
As someone who worked at a low code workflow SaaS, I agree with this. The automation is a commodity, the value was in discovering that people wanted to talk to the LLM to orchestrate their automation vs bespoke workflow design UX. You can also decouple models from API scaffolding like Gorilla LLM (or rather, broad models that dive deep into use case specific models on demand).
- flimsypremise 3y ago
  Congratulations you have just invented yet another query language with the special fun addition that the result is non-deterministic.
  - toomuchtodo 3y ago
    Doesn’t matter, as long as users are getting value. How you solve the problem is less important than solving the problem. I expect LLM output to improve over time, along with mechanisms to coax deterministic behavior.
    - nextworddev 3y ago
      Yeah developers sometimes overrate precision when in reality being precise is not as good as getting the job done
quickthrower2 3y ago
Software Engineering will be fine: it has been eating itself since Lisp invented macros. Maybe software engineering should be called “The operations of automation using data on general purpose machines”. That will not die. It will look different though.
- nextworddev 3y ago
  Sure - it will take on different form
contravariant 3y ago
That might work but it's going to require conversational AI with the audacity to talk back, which is going to be interesting...
doubloon 3y ago
Power Automate Cloud already has an AI builder, you type in "At 4pm every day check my Outlook inbox for an email from Suzie Snowflake, and if it says Critical, then check Sharepoint for a file from Integration Services and if its contents contain the string "Blorb" then send an email to Escalation" the result is often slightly wrong, but if someone in MS management is paying attention and could get it to ChatGPT-python level of accuracy, it would take over an enormous amount of stuff.

toomuchtodo 3y ago

As someone who worked at a low code workflow SaaS, I agree with this. The automation is a commodity, the value was in discovering that people wanted to talk to the LLM to orchestrate their automation vs bespoke workflow design UX. You can also decouple models from API scaffolding like Gorilla LLM (or rather, broad models that dive deep into use case specific models on demand).

flimsypremise 3y ago
Congratulations you have just invented yet another query language with the special fun addition that the result is non-deterministic.
- toomuchtodo 3y ago
  Doesn’t matter, as long as users are getting value. How you solve the problem is less important than solving the problem. I expect LLM output to improve over time, along with mechanisms to coax deterministic behavior.
  - nextworddev 3y ago
    Yeah developers sometimes overrate precision when in reality being precise is not as good as getting the job done

quickthrower2 3y ago

Software Engineering will be fine: it has been eating itself since Lisp invented macros. Maybe software engineering should be called “The operations of automation using data on general purpose machines”. That will not die. It will look different though.

nextworddev 3y ago
Sure - it will take on different form

contravariant 3y ago

That might work but it's going to require conversational AI with the audacity to talk back, which is going to be interesting...

doubloon 3y ago

Power Automate Cloud already has an AI builder, you type in "At 4pm every day check my Outlook inbox for an email from Suzie Snowflake, and if it says Critical, then check Sharepoint for a file from Integration Services and if its contents contain the string "Blorb" then send an email to Escalation" the result is often slightly wrong, but if someone in MS management is paying attention and could get it to ChatGPT-python level of accuracy, it would take over an enormous amount of stuff.