Tech News of the Week: What’s a Minsky Moment and Why Are There Worries About One?

The mere mention of a “Minsky moment” — a sudden crash of markets and economies that are hooked on debt — is enough to send shudders through policy makers. The theory stems from the work of Hyman Minsky, a US economist who specialized in how excessive borrowing fuels financial instability. From time to time, booms in financial markets or sky-high debt levels around the world lead to renewed interest in Minsky’s theory or warnings from the International Monetary Fund, among others. US Treasury Secretary Janet Yellen once described his work as “required reading.”

1. What makes a Minsky moment?

The term refers to the end stage of a prolonged period of economic prosperity that hasencouraged investors to take on excessive risk, to the point where lending exceeds what borrowers can pay off. At that point, Minsky wrote, there’s an increase in “speculative and Ponzi finance.” When a destabilizing event as simple as an increase in interest rates occurs, investors can be forced to sell assets to raise money to repay loans. That in turn sends markets into a spiral amid a demand for cash. There have
been attempts to distinguish between a Minsky moment and a Minsky process that leads up to it.

2. Have there been Minsky moments?

Yes. In 1998, following the bursting of asset bubbles in Asia, Russia defaulted on its domestic debt and devalued the ruble. (It was during that crisis that Paul McCulley, then an economist at Pacific Investment Management Co., coined the term “Minsky moment.”) The global financial crisis of 2007-2008 is considered another Minsky moment, since it was caused by the implosion of the US subprime mortgage market.

Continue Reading: https://www.bnnbloomberg.ca/business/2024/08/14/whats-a-minsky-moment-and-why-are-there-worries-about-one/

In AI, Smaller, Cheaper Models Are Getting Big Attention

Miami-based Arcee is one of a growing number of companies redefining the conventional
wisdom in the tech industry that bigger is always better for AI. Meta, Google, and OpenAI
are all investing in more affordable alternatives to large language models. Rather than
trying to do everything ChatGPT can, Arcee’s software focuses on a more limited set of
day-to-day commercial tasks, requiring less data.

For a long time, tech giants like Google and startups such as OpenAI have been competing to create increasingly larger and more powerful artificial intelligence models using vast amounts of online data. Deployed in chatbots like ChatGPT, this technology can handle a wide range of complex queries, from drafting legal documents and planning travel itineraries to writing Shakespearean sonnets about ice cream.

Mark McQuade is taking a different approach. Arcee.AI, the startup he co-founded last year, helps companies train and deploy smaller, more specialized AI models. These small language models are becoming increasingly popular and powerful. Rather than trying to replicate everything ChatGPT can do, Arcee’s software focuses on handling specific commercial tasks—such as building a service that only fields tax-related questions—without requiring as much data. “For 99% of business use cases, you
probably don’t need to know who won an Olympic gold medal in 1968,” McQuade said.

Continue reading: https://www.bloomberg.com/news/articles/2024-08-08/move-over-llms-small-ai-models-are-the-next-big-thing

Resume written by Victor Ayala, Business Developer Amix Tech

More info@amixtechlab.com