From what I can tell there's are mostly two options: Either AI is and will be useless or it's severely undersupplied. People, even those deeply technical, where AI has the most impact right now, still widely argue about if AI even offers any value. Adoption is far from anything that is plausible, if (not when) it became clear that it does.
If you land on "does not", given the investments so far, commercial entities would obviously be overvalued already and any investment goes to 0 over time.
But if we land on "does", how could Nvidia not be anything other than undervalued right now? No matter what frontier model: I can look at my screen, LLM generated characters visibly appearing in chunks, depending on the model after initially waiting for 10-20 seconds, for even benign queries — because that is the best we can do right now. And that's while most people still argue if AI will actually do anything and humanity at large does not really use it, neither personally nor societally.
If AI does in fact do something valuable and that something gets better, everyone will want it and there will be demand for lots of chips.
From what I can tell there's are mostly two options: Either AI is and will be useless or it's severely undersupplied. People, even those deeply technical, where AI has the most impact right now, still widely argue about if AI even offers any value. Adoption is far from anything that is plausible, if (not when) it became clear that it does.
If you land on "does not", given the investments so far, commercial entities would obviously be overvalued already and any investment goes to 0 over time.
But if we land on "does", how could Nvidia not be anything other than undervalued right now? No matter what frontier model: I can look at my screen, LLM generated characters visibly appearing in chunks, depending on the model after initially waiting for 10-20 seconds, for even benign queries — because that is the best we can do right now. And that's while most people still argue if AI will actually do anything and humanity at large does not really use it, neither personally nor societally.
If AI does in fact do something valuable and that something gets better, everyone will want it and there will be demand for lots of chips.