Metin Seven 🎨3d ago

So where are all the AI apps?

https://www.answer.ai/posts/2026-03-12-so-where-are-all-the-ai-apps.html

#tech #technology #software #apps #BigTech #IT #AI #ArtificialIntelligence #LLM #LLMs #ML #MachineLearning #GenAI #generativeAI #AIAgent #AISlop #Fuck_AI #microslop #microsoft #copilot #meta #google #NVIDIA #gemini #OpenAI #ChatGPT #anthropic #claude

So where are all the AI apps? – Answer.AI

Practical AI R&D

Answer.AI
Show threadTillman Reuter3d ago
@metin the analysis is really shallow though. I don’t see any meaningful results
Show threadDr. Daniel Dizdarevic2d ago
@tillmanreuter @metin That is exactly the point. There is no meaningful evidence that AI improves productivity or innovation. I doubt a more in-depth analysis would change this, given that this conclusion is perfectly reasonable.
Show threadTillman Reuter2d ago
@daniel you don‘t strengthen a hypothesis by not finding evidence in favor of the counter-hypothesis. And of course you can doubt stuff, but you are no closer to knowing than before.
Show threadDr. Daniel Dizdarevic

@tillmanreuter While this may be true from a purely theoretical perspective, in practice such a hypothesis becomes increasingly unlikely. After all, in science, we can never know if a theory is correct. However, we must choose the most likely candidate based on current observations and knowledge.

Moreover, the specific hypothesis "AI will make software development more efficient" is based on naive assumptions, given that basic coding, which AI can currently do okayish, is only a small part of the process. Even if AI were to drastically improve this part, the overall impact would still be small.

@metin

Mar 26 at 10:03amWeb
Show threadMetin Seven 🎨1d ago

@daniel @tillmanreuter When I'm talking to coders these days, two observations keep returning:

1. AI-generated code is generally bloated, causing coders to spend time separating the wheat from the chaff.

2. AI-generated code causes coders to lose track of the code structure, as opposed to coding by themselves.

These two characteristics generally cause gen-AI to not significantly improve development time.

Show threadDr. Daniel Dizdarevic1d ago
@metin @tillmanreuter Good points, Metin! I would just add one more point: Current AI is incapable of understanding complex concepts and structures, particularly novel or unique ones. Therefore, the most challenging tasks must still be carried out by humans.