NYTimes: “Like a globe-spanning tornado that touches down with little predictability, deep economic anxieties are leaving a trail of political turmoil and violence across poor and rich countries alike…The causes, context and conditions underlying these disruptions vary widely from country to country. But a common thread is clear: rising inequality, diminished purchasing power and growing anxiety that the next generation will be worse off than this one. The result is that citizens in many countries who face a grim economic outlook have lost faith in the ability of their governments to cope — and are striking back. The backlash has often targeted liberal democracy and democratic capitalism, with populist movements springing up on both the left and right. “An economic malaise and a political malaise are feeding each other,” said Nouriel Roubini, an economist at New York University.”
Andrej Karpathy: “We’re entering a new computing paradigm with large language models acting like CPUs, using tokens instead of bytes, and having a context window instead of RAM. This is the Large Language Model OS (LMOS).” [via Haider]
Business Standard: “The talent shortage or skills gap has been a persistent issue in the industry. A marathon general election and extreme heat have only exacerbated the problem. The shortage of skilled labour appears to have hit India’s engineering and capital goods firms the hardest. Industry executives attribute the shortage to a mix of India’s growing order book and increasing demand, while other factors impact the supply end. An executive who does not wish to be named says: “Not all labour is skilled, and those we skill, we tend to lose to markets like West Asia due to better pay. Preference for an air-conditioned work environment is another factor impacting labour supply for engineering firms.”” A possible solution from Niranjan Rajadhyaksha.
CNN: “Governments owe an unprecedented $91 trillion, an amount almost equal to the size of the global economy and one that will ultimately exact a heavy toll on their populations. Debt burdens have grown so large — in part because of the cost of the pandemic — that they now pose a growing threat to living standards even in rich economies, including the United States. Yet, in a year of elections around the world, politicians are largely ignoring the problem, unwilling to level with voters about the tax increases and spending cuts needed to tackle the deluge of borrowing. In some cases, they’re even making profligate promises that could at the very least jack up inflation again and could even trigger a new financial crisis.”
FT: “No economic model would have predicted stocks would be at all-time highs and credit spreads would be very narrow after the Federal Reserve raised its benchmark interest rate by 5.25 percentage points since early 2022. Yet, that is exactly what has happened. The Fed seems ready to declare victory in its fight against inflation, but the outperformance of highly speculative investments suggests that even such a sharp increase in interest rates hasn’t been a big enough mop to soak up the excess liquidity sloshing around the financial markets.”
That bit about Large Language Models acting as an Operating System by Andrej Karpathy is originally from this video by Karpathy – https://youtu.be/zjkBMFhNj_g?feature=shared&t=2535( start at 42:15). It is an interesting take, and that entire video by Karpathy( Intro to LLMs) is worth a watch even if one is well versed with LLMs.