HexnMarcusWolfe
HexnMarcusWolfe on Hacker News
By short circuiting the exploratory process via AI to find a solution. Some of the questions that would normally be asked during the struggle of said process will also be circumvented.
Questions that would lead to other discoveries, ideas, and perhaps entire fields of study.
At the end of the day AI is still a machine and (assuming maximal alignment) it's only going to give a response to what you ask it.
Even if you give it a directive: Keep an eye out for inferences that can be made from within the solutions reasoning that would prove useful or interesting.
That's not going to cut it.
Google is slightly undervalued, Misanthropic is massively overvalued.
Let's talk timelines.
2015: Google deployed it's first TPU (tensor processing unit)... For context, openAI in 2026 is only now producing their first TPU chip. Even being generous saying it was only INT8 the same as Nvidia 20-series (2018) that's still 3 years ahead.
2017: The landmark research paper detailing "transformers" in LLM's, titled: Attention is all you need, was published... by 8 engineers working at Google's Deep Mind.
In my mind there's a few possibilities.
1. Google really were blindsided and through lack of leadership / vision were surpassed.
2. Google has already war gamed this out. They know the limits of stochastic tech with "AI" being mostly marketing hype (same as crypto / NFT's), which is why they've limited their liability by investing 10x less than most others in the space.
3. Or they knew AI is going to have some success, but stuck to principles. Google the search engine started out by running distributed (max redundancy) on absolutely dogshit hardware. Yet through clever software optimizations + minimal CAPEX / OPEX, they were able to provide the best service and out compete everyone (Alta vista, yahoo, etc.). Same thing this time around. I have a feeling they (and anyone else not caught up in religious / marketing hype e.g. Deep Seek) is looking at the OPEX and trying to figure out how to optimize.
The amount of problems it's had from compromised extensions is ridiculous.
I said ~20% because that's the minimum required to bring it to commercial market while being competitive. If they go more, great, bonus.
Tho' something interesting. Those with large enough arrays can use their excess to produce nitrogen fertilizer via cold plasma.
This would either allow them to store value in that (for sale to farmers / national stockpile) which may give better yields then storing via batteries and selling back to the grid. Or they could use it to grow their own produce making them self reliant and less dependent on supermarket chains and by extension their dependency on fossil fuel use in their distribution networks.
Personally i'm waiting for the next iterations of Kesterite cells.
If we can get single cells to crack ~20% efficiency. With a tandem cell architecture, it's going to become the new standard overnight.
Durability of silicon cells or better, but orders of magnitude $ cheaper to produce, and without the need to use toxic agents as in perovskite cells.
We're movin' fast.
For over a decade we were stuck at ~12% efficiency. And in the last 4 years (probably with AI help) we've cracked 15% efficiency
https://interestingengineering.com/energy/chinese-researcher...
Few more years, and with a bit of luck, we'll all be around to see solar panels that cost 1/10th of what they do now.
Beyond just saying "China copied it", show us the logs of a tens if not hundreds-of-gigabytes worth of AI models being downloaded? And if that indeed happened, where are the prosecutions for US companies who through lack of security handed over the equivalent of digital nuclear bomb blueprints to the Chinese... no one's in prison yet?
As far as i can tell the only thing that can be claimed is that Chinese AI are saturating US AI's requests and attempting to train their own adversarially.
But if that's the case, China is already at the point where they can implement such adversarial training on their own internally. They don't need the US.
Clearly the US platforms / software side is unprofitable, and the hardware side is engaging in obscene business practices like channel stuffing and profiteering. In combination with huge overvaluation, this has created a house of cards.
In context of that card house, this article sounds like something written by someone who doesn't have a clue about AI, but has a vested interest in keeping the US share market afloat... like a conventional economist that subscribes to neoliberal dogma.