The world is buzzing over the capabilities, potential, and in some cases, legal implications, of artificial intelligence, and for good reason.
In a very real sense, this feels like the hype that preceded the very first personal computers, before they became an inseparable part of the human experience.
Like late ‘80s/early ‘90s PCs, AI is traversing some growing pains – some quirky applications, tech foibles, and even some legal challenges. But make no mistake – AI is here, continuously evolving to suit our needs and (what I hope will be) humanity’s interests.
At CES 2024, the consumer world made its bold AI gambit, showing off a mountain of intriguing tech demos and kitsch. CEA President Gary Shapiro has noted that AI “tools and platforms can solve big global problems and create valuable knowledge,” while also acknowledging the risk of “deep fakes, political manipulation, and fraud aimed at vulnerable populations.”
Keep in mind that Shapiro leads a trade organization representing a $505 billion industry, much of which depends on monetizing an idea that, by its nature, downplays humans. So if even Shapiro is sounding the alarm – a faint one, mind you – you know it’s a serious issue.
A couple months ago, I had Temple professor Subodha Kumar on our PSDcast to discuss the alarming AI privacy concerns and whether it’s possible to reseal AI’s Pandora’s Box (spoiler alert – it’s not, but that’s not the whole story).
In February, I explained why the historical benchmark for AI, the Turing Test, has become mostly irrelevant. And in this month’s Final Thought, I detail how and why AI has entered a minefield of litigation and legal challenges. AI might be evolving at an expeditious pace – possibly independent of human influence – but the courts still have the ability to impede its development, if only temporary.
We’ve also just wrapped up the Applied Power Electronics Conference (APEC), and fittingly, the March issue focuses on Power Supplies.
One of this month’s contributed pieces, by GAIA Converter, discusses “Optimising Distributed Power Architectures with Programmable PoL Converter Current Limits.”
GAIA’s Jean-Loup Guédon notes how “Even while accounting for DC power distribution, the disparity between input and output voltages gets larger, with supply rails often subject to wide variations, while the outputs, sometimes down to around 1V, must be ever-more accurate and disturbance-free.
Designers often use an intermediate DC bus (at typically 12V, 24V or 28V), generated by an Intermediate Bus Converter (IBC), followed by typically non-isolated high-efficiency Point of Load (PoL) converters, and Guédon ends up explaining why PoLs with controlled inrush, trimmable current limit and trimmable output voltage are “the ideal future-proof solution.”
Enjoy the March issue!
Best Regards,
Jason Lomberg
North American Editor, PSD
