Raw LLM Responses

To train and make AI, you need money. Tech companies are the only ones with enough money to realistically make human-like AI systems. For now, at least. Maybe a new AI math structure will be far more accessible to other companies in the future. Tech company labor is dominated by software engineers. Get rid of them and you'll cut no-value-added costs by like 80-95%. No office buildings, office furniture, or office hardware, or janitors for offices needed, or secretaries, or HR, if you have no devs in the office anymore. Ergo, tech company saves a ton of money by automating their labor. Even more money can go into training AI and selling AI (which the AI might be able to sell itself). Also, any AI capable of actually replacing a developer (not a copypasta dev, but someone who creates novel solutions to problems using software) is also an AI that can very, very likely be tuned to solve non-coding problems with code. AI practically flies planes from start to finish. What other things can AI+hardware do that humans already do? Warehouse workers are currently targets, as are drivers, and food makers. Teachers and tutors are already under attack. Can AI be used to replace those humans, too? Definitely—if it is advanced enough to replace devs, it's advanced enough to replace most people's jobs given the right hardware. The only problem these companies have is... they released the right product under the wrong promise. The product that is LLMs is great for what they do, but a farcry from replacing people. They're now all scrambling to prove their LLMs can replace people, like they promised it would. The goalposts on expectations are still getting walked back to, "Enhancing productivity," rather than replacing workers.