The race to build Artificial General Intelligence (AGI) may be running into an unexpected and interesting problem, and that is, humans simply can’t type fast enough.
Alexander Embiricos, who leads product development for Codex, OpenAI’s coding agent, has said that one of the biggest barriers to achieving AGI today is human speed. Speaking on ‘Lenny’s Podcast’, Embiricos pointed out that the way humans interact with AI, through typing prompts and manually checking outputs, is slowing everything down.
He described the “current underappreciated limiting factor” to AGI as “human typing speed” and “human multi-tasking speed on writing prompts.”
ALSO READ | What is Google Disco? A new AI browser that can revolutionise web surfing
In simple terms, AI systems can already do a lot, but humans are becoming the bottleneck in telling them what to do and reviewing their work.
AGI refers to a future form of AI that can reason and think at the same level as, or better than, humans. It is the long-term goal that most major AI companies are chasing.
Embiricos explained that even if an AI agent can observe and assist with work, progress slows if humans still need to validate everything.
“You can have an agent watch all the work you're doing, but if you don't have the agent also validating its work, then you're still bottlenecked on, like, can you go review all that code?” he said.
ALSO READ | The ChatGPT 5.2 update: Here is the list of new abilities OpenAI’s chatbot brings with the new GPT-5.2 model
Embiricos shares the solution
According to him, the solution lies in reducing how much humans need to type, prompt, and supervise. “If we can rebuild systems to let the agent be default useful, we'll start unlocking hockey sticks,” Embiricos said, referring to rapid, exponential productivity growth.
He added that there is no one-size-fits-all solution. Different industries and tasks will need different levels of automation. However, he expects visible changes soon.
“Starting next year, we're going to see early adopters starting to hockey stick their productivity,” he said, adding that larger companies will follow in the years after.
Embiricos believes AGI will emerge somewhere between early proactive gains and fully automated systems at scale.
