# Helping Airtable team to add AI features - AI code suggestions 😁

**URL:** <https://air.tableforums.com/t/helping-airtable-team-to-add-ai-features-ai-code-suggestions/381>\
**Category:** Show and Tell\
**Created:** [March 8, 2023, 3:21pm UTC](https://air.tableforums.com/t/helping-airtable-team-to-add-ai-features-ai-code-suggestions/381 "2023-03-08T15:21:13Z")\
**Posts on this page:** 1\
**Showing post:** 3

<div class="post-metadata">

**Author:** ![bfrench](https://yyz1.discourse-cdn.com/flex029/user_avatar/air.tableforums.com/bfrench/32/46_2.png) [@bfrench](https://air.tableforums.com/u/bfrench)\
**Post date:** [March 9, 2023, 2:22pm UTC](https://air.tableforums.com/t/helping-airtable-team-to-add-ai-features-ai-code-suggestions/381/3 "2023-03-09T14:22:31Z")

</div>

> [@Greg\_vonF](#):
>
> Chrome extension, which adds formula suggestions based on ChatGPT API

This is good stuff. Certainly, a fun direction to take LLMs. Leaning into easier pathways for everyday users to create more complex formulas is a good use of AI. Ultimately, though, there are far bigger fish to fry up - I think LLMs will [change everything for users](https://air.tableforums.com/t/suddenly-rows-can-now-have-meaning/386), not just solution-builders.

But two dimensions of this approach concern me.

1. Prompt engineering must be used, and that requires a fair bit of maintenance because prompts (that work) for GPT-3.5 are vastly different from GPT-3.0. Where will this go in GPT-4 and GPT-5?
2. The nature of Airtable formulas is unique; the patterns are inconsistent, so how could AI provide consistent (and smart) recommendations?

[This response](https://air.tableforums.com/t/any-idea-of-chatgpt-can-check-for-airtable-formulas-errors/364/14) is also relevant. It explains why trying to do anything AI-related wth formulas may be extremely difficult.

---

_[View the full topic](https://air.tableforums.com/t/helping-airtable-team-to-add-ai-features-ai-code-suggestions/381)._
