"First, all of the weaknesses of LLMs are being addressed by exponential increases in compute"
Are we sure about that? Because I have not seen evidence on this. I kinda subscribe to LeCun's argument that LLMs cannot be "fixed" with more compute. But I am open to any counter-argument and evidence to the contrary.
"Second, the number of people who need to wield AI effectively for demand to skyrocket is decreasing."
I am not even sure what this means. What does "wield AI effectively" mean? What does the whole sentence mean?
"Third, the economic returns from using agents aren’t just impactful on the bottom line, but the top line as well."
What evidence do we have on AI helping companies generate more revenue organically because of its capabilities, not because of the demand for components? Google and Facebook, likely. The same article even said that consumers did not care about agents when it came to buying Apple products. Yes, we have heard Apple is behind in AI for months and guess what, more people are warmed up to the idea that the company might be smart all along.
"what has Apple done to inspire faith that they can do a better job?"
Or maybe because consumers don't care about agents or AI just because business executives, analysts or influencers like the author keep talking about agents and AI? If AI can make a product better, great! But should Apple focus on delivering the best products and experience possible whether or not it involves AI? Or should it deliver AI at all costs?
Anthropic found a product market fit, a clientele that finds value in paying for an LLM, whether it's real efficiency or just an excuse to cut costs. OpenAI must continue to ship models. That's how it keeps the music going. But a lot of consumers don't want to pay for something that doesn't justify the monthly cost that is just going to get more expensive.
In this case, he is entitled to liking or disliking the experience watching the game on VP. I read that several liked the experience, but he absolutely can have a different opinion. What I am talking about here is that he implied to know more about the product than Apple, which started the project way back in 2015 and has more information than he does. We can all like or dislike a product or opine on a strategy or a state of a business, but claiming to know more about a product, esp one like VP, than the company itself, looking outside in, is too much imo.
Surprise surprise, a long-time Apple hater who got a lot of things wrong on the company and who has not worked on hardware thinks he understands the hardware more than the company. I get that everyone can be disappointed with a product, but thinking he knows more about a product as sophisticated as Vision Pro than Apple is something else. I stopped reading his stuff a long time ago and it seems nothing changed
Are we sure about that? Because I have not seen evidence on this. I kinda subscribe to LeCun's argument that LLMs cannot be "fixed" with more compute. But I am open to any counter-argument and evidence to the contrary.
"Second, the number of people who need to wield AI effectively for demand to skyrocket is decreasing."
I am not even sure what this means. What does "wield AI effectively" mean? What does the whole sentence mean?
"Third, the economic returns from using agents aren’t just impactful on the bottom line, but the top line as well."
What evidence do we have on AI helping companies generate more revenue organically because of its capabilities, not because of the demand for components? Google and Facebook, likely. The same article even said that consumers did not care about agents when it came to buying Apple products. Yes, we have heard Apple is behind in AI for months and guess what, more people are warmed up to the idea that the company might be smart all along.
"what has Apple done to inspire faith that they can do a better job?"
Or maybe because consumers don't care about agents or AI just because business executives, analysts or influencers like the author keep talking about agents and AI? If AI can make a product better, great! But should Apple focus on delivering the best products and experience possible whether or not it involves AI? Or should it deliver AI at all costs?
Anthropic found a product market fit, a clientele that finds value in paying for an LLM, whether it's real efficiency or just an excuse to cut costs. OpenAI must continue to ship models. That's how it keeps the music going. But a lot of consumers don't want to pay for something that doesn't justify the monthly cost that is just going to get more expensive.