"To do unreasonable things, you need to be unreasonable"
I just signed the waitlist for an app that augments our coding environment with AI, called Kite. Exciting times.
Kite sounds cool, and from what I can see it looks like it needs a lot more work. But I love where this is going. Rather than just code completion and hinting, I want it to tell me (and have the option to dismiss):
- Your database queries appear to be vulnerable to SQL injection.
- You’re vulnerable to Heartbleed.
- You’re coding in a non-standard way. eg. You’re mixing capitalization/spacing.
- You’re not using common conventions, camelCase in JS, underscores for Postgres fields, etc.
- You’re using a version of PHP that has X critical vulnerabilities.
- You’re coding MySQL in a non-recommended way. eg. using mysql instead of mysqli in PHP.
- Your forms appear to be vulnerable to XSS.
There’s really no limitation to how far we can go with this until AI is doing the programming for us. Thoughts?
As a stopgap, I hope they’re doing something clever like Honey, rather than traditional methods (scraping, manual input). They promise something (coupon codes) and can only fulfill that promise because they get access to user data (users applying coupons on other sites). Initially they can’t fulfill the promise, but soon grow quickly & organically, if the promise is worthwhile to enough people.
And if you liked this post please consider donating some Bitcoin :)
>> Follow me on Twitter for more posts like this.