Not only are there going to be languages optimized for LLMs, but there are going to be new high-level languages for humans to quickly, but precisely understand the giant code stack created by the LLM. An agent-generated markdown won't be enough. So Python and byte-code will morph to Python++ for humans and byte-code-- for LLMs. Otherwise programmers will gradually lose track of the details of their code.

The problem with these executive orders (AI and now robotics) is that they have no funding behind them. The administration likens these initiatives to the Manhattan and Apollo programs, but those each had prodigious funding levels. The best this can do is reprioritize existing funds. However, since Trump re-took office, with DOGE attempted to eviscerate gov't programs including science funding, this momentum has been stifled rather than encouraged. For instance, since January the NSF hasn't had a personal robotics liaison at seedfund.nsf.gov. I don't see this executive order reversing this ebb tide, let alone spurring a flood of activity.