If I have a large amount of capital to burn, and an idea that begins and ends with 'build me a Groupon clone for x market', then just pour money into a team of outsourced humans at a software sweatshop to build it. Is it still my work? Again, not many people would object if I were the face of that company, it is after all my app.

I'm struggling to see what the difference is if humans were swapped with LLMs in each case? Is the issue that the LLM models were trained on plagiarized code? I agree we should take issue this aspect, but when it comes to being honest about credit for the 'doing' part, this has never been the case, pre or post LLMs.

The issue is that simple and novel can also be copied easily (anyone can just point an LLM at the target and ask it to build a local personal version). The days of making money from this class of app are over.

Novel and very complex ideas will still make money, but outside of gaming are usually niche markets.

Novel and moderately complex apps will have to build their own moats to prevent copying - heavy server side rendering, server logic processing, and an idea that can't be re-implemented on simple CRUD architecture.

You have the right to believe what you like, in my opinion the consequences for blaming our own crimes on others is dire.

The semblance of an international order post world war 2 at least suggested that we as a species were attempting to move toward a better place. Many notable breaches occurred but there was a much stronger universal condemnation of each, with real political repercussions even for western powers.

Failure could be team burnout and attrition, it could be a major bug due to fast tracking validation (or not checking AI output), or a fundamental architectural issue with future repercussions that with a clearer mindset would have been considered .. or myriad other things in combination.

The person calling out the risk knows that it will 100% cause a failure somewhere. Exactly where and how the failure will occur cannot be predicted beforehand and it actually doesn't matter as the end result will be a missed delivery or incident.

Unfortunately the executive decision maker needs a concrete failure in order to implement a concrete solution.

The more features I add now will likely make the inevitable refactor much harder, but adding new features is so easy that I want want extend this illusion of productivity just a little bit longer!

Outside of my work - AI video has reached a point where creative artists are generating engaging, realistic content that is genuinely entertaining. IMO the common thread is that in the right hands it can produce amazing things. For anyone without the expertise, it defaults to slop.

I don't think any of the big streaming content providers have native apps on linux and no browser can pass through audio bitstreams to HDMI. Video quality is limited as well.

Having a dedicated streaming box is better in this regard

The Internet becomes primarily a passive stream of information vetted by government and Mega Corps, just like the TVs of old. Except for the nifty buy with one click button of course

If incorrect LLM output is a prompt issue then demand for experienced developers will remain, and demand may actually increase as time passes.