LLMs can't do frontier science because they can't invent new language, argues Ha

<table> <tr><td> <a href="https://www.reddit.com/r/OpenAI/comments/1wdde0p/llms_cant_do_frontier_science_because_they_cant/"> <img alt="LLMs can't do frontier science because they

<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1sxffva/llms_are_a_dead_end/"> <img alt="LLMs are a Dead End"

<table> <tr><td> <a href="https://www.reddit.com/r/ChatGPT/comments/1ts1bcc/hot_take_llms_cant_do_statistics_not_really/"> <img alt="Hot take: LLMs can’t do statistics. Not

<!-- SC_OFF --><div class="md"><p>Something I keep noticing is that LLMs can give really strong, coherent answers in a single prompt, but as soon as the conversation goes on for a

<!-- SC_OFF --><div class="md"><p>I asked Gemini to give me all the days that have &quot;d&quot; in them. It returned -</p> <p>Monday, Wednesday, Thursday, Sunday<br />

<!-- SC_OFF --><div class="md"><p>I asked gpt about this, and it said yes, and gave some good examples of how some python code might look like in a new language, where the

<!-- SC_OFF --><div class="md"><p>Every &quot;AI agent&quot; that controls a computer today does this: generate text → parse it → execute → serialize result back to text → repeat.

<!-- SC_OFF --><div class="md"><p>I’ve been saying this for quite some time now and this paper that came out recently really puts it clearly</p> <p><a