“That’s gonna kill Google” 👀 Psst @lexfridman & @realGeorgeHotz — I think you’d be interested in the progress of @WeGPTai on this topic. See attached video for auto-coder👇 And for real-time Knowledge retrieval: youtu.be/0wEx6yZWMIg?si=4UHB…

🤖How A.I. is changing Software engineering forever Imagine a routine task a programmer might face in their day-to-day routine: Comparing and contrasting various code implementations that other programmers at the company (or even open-source resources) originally wrote. Now, it's one thing to read the documentation or inline commenting in a given code block. It's quite another to take two different implementations of a similar function or overall subroutine, however, and contrast the key differences at-a-glance. Particularly with complex code bases. This is something AI excels at. I asked ChatGPT to do exactly this with two versions (one much older, and one much more recent and robust implementations) of the same function from @WeGPTai. As the proprietor of both functions, I can tell you, the evaluation ChatGPT provide in the below screenshots is spot-on 👌. Now, if I had been a manager or a peer, and been asked to evaluate the two different states of the same function, this level of tracing and logical evaluation might have taken me 20-30 minutes to complete and articulate as thoroughly. Guess how long it took ChatGPT? 15 seconds. 15 seconds. To my 30 minutes. Now, imagine the implications this will have on almost every engineering job. And not in the future -- RIGHT... NOW. Remote Full-Time Work is going to End: You'll be reading about enterprises requiring remote workers to come into work 3-4 days a week minimum for full-time employment. Some own't even allow remote full-time work anymore. Why? Because with this kind of efficiency improvement, one hustle already spreading like wildfire are talented engineers taking on 2, even 3 FULL-TIME positions (with 2-3x the full-time benefits included) at the same time, and just using AI to automate the mundane parts to the point where they can deliver comparable performance with each. Why is this problematic? Well for one, normally you'd be a contractor if you were servicing multiple enterprises like this at a time. It has broad implications on benefits programs globally. Two, it has huge implications on intellectual property exclusivity. Some jobs it is understood that the work product you make during your tenure stays at that place. And often times when you leave there are non-compete clauses because the stuff you worked on was so important. That's specifically why full-time employees tend to have different access level and regulations to follow than contractors. If you're in the office, this kind of hustle becomes a lot less feasible to pull off.

Not a single line of code is visible. Ever. @WeGPTai Watch ChatGPT🤖 build "Flappy Bird" from a single-sentence prompt, with the only interaction needed from the user is to play test auto-generated builds and provide iterative, natural language feedback 🤯👇 Playable link in 🧵