If you haven’t come across it already, read and marvel at this article by a Wired journalist who took a tour of a TSMC semiconductor fab.
“Every six months, just one of TSMC’s 13 foundries—the redoubtable Fab 18 in Tainan—carves and etches a quintillion transistors for Apple. In the form of these miniature masterpieces, which sit atop microchips, the semiconductor industry churns out more objects in a year than have ever been produced in all the other factories in all the other industries in the history of the world.” (h/t The Diff)
In the spirit of Feynman this superb blog post, by none other than Stephen Wolfram, gives a lucid explanation of what is going on under the hood of the latest tech phenomenon.
The short answer is “it’s maths”.
“But in the end, the remarkable thing is that all these operations—individually as simple as they are—can somehow together manage to do such a good “human-like” job of generating text. It has to be emphasized again that (at least so far as we know) there’s no “ultimate theoretical reason” why anything like this should work. And in fact, as we’ll discuss, I think we have to view this as a—potentially surprising—scientific discovery: that somehow in a neural net like ChatGPT’s it’s possible to capture the essence of what human brains manage to do in generating language.”
The argument that AI is unlikely to be a winner for the middle-ground companies.
Why? “was a feature not a product” – in other words value will either accrue to core AI platforms (e.g. Open AI) or to incumbent software tools with distribution who will just add AI features.
“Adobe will own the AI-based image editing market Office & Google Docs will own the AI-based writing market Salesforce will be the best AI-enabled CRM Shopify the best AI optimization and customer support Zoom the best AI meeting summaries … all with a few API calls“
“The most intriguing aspect of Wall Street’s behavior is that if we examine every period of industry decline over the past several decades, we see that stock prices reach their lowest point when the industry only starts to experience a fall in earnings, well before the trough of earnings. The stocks were already recovering and surging by the time layoffs and consolidation occurred.“
Semiconductor index (SOX) is up strongly since October lows and earnings cuts have only just started i.e. a typical pattern with the expected bottom of the cycle is Q2 2023.
Yet, as argued here, this cycle appears different – (1) days of inventory at record high which will take, despite a desire for supply chain resilience, more than two quarters to clear (2) there is an oversupply of certain process tech (3) channel stuffing has been a big feature.
A harsh, almost damning, write-up of Intel, pre-dating the dividend cut.
“In reality, Intel is not the giant of the industry. Intel’s total share of industry capacity is around 10%, they are not a giant who has stumbled, they are a niche player and have been for years.“
RISC-V, the royalty-free open-source instruction set architecture, is worth keeping an eye on especially as Arm convulses its way back to the stock market.
The December 2022 summit (and this great write-up) offered a deep feel of the status of RISC-V.
Bold statements abounded – “It’s really important that you get this. RISC-V is inevitable. RISC-V is going to have the best processors. And RISC-V is going to have the best ecosystem.”
“Direct-to-consumer brands moved 20–30% of their marketing dollars in Q4 2022 from Meta to Amazon due to the former’s declining performance metrics for ads. The shift occurred despite a recent reluctance from DTC brand to sell products on Amazon because of limited access to and ownership of sales and customer data—but now that brands are receiving only $2 back for every $1 spent on Meta ads (they used to get $8 back), they are more willing to work with the e-commerce giant. According to Advantage Unified Commerce, an estimated 75% of brands report customer acquisition is cheaper on Amazon than other media channels.”
