Hey there, supply enthusiasts!

Ever wondered why some companies still prefer old-school, in-house data centers in the age of the almighty cloud? And what’s the cost of building all those energy-guzzling data centers, anyway?

In labor news, maybe you’ve noticed that more workers are sticking around rather than jumping ship? We explore why the quit rate is dropping across America and unpack the rise of the boomerang employee.

Let's dive in.

Industry Highlights: The Cloud vs. In-House Battle Continues

The cloud has been gobbling up corporate computing like Pac-Man on a mission. Businesses dropped a whopping $230 billion globally on public cloud services last year, up from just under $100 billion in 2019. Amazon Web Services, Google Cloud, and Microsoft Azure are growing faster than small-town gossip (to use a Texas saying), with revenues soaring by over 30% annually. With their burgeoning AI tools, these tech giants seem poised to ring the death knell for ye olde on-premises data centers.

Or do they? According to The Economist, companies are still investing over $100 billion in their own hardware and software.

Many industrial players find on-premises computing has its perks, especially when it comes to handling data generated by their connected factories and products. For a start, time matters. Sure, the public cloud offers convenience and cost savings, thanks to its massive scale. But there's a catch. These hyperscale data centers are often far from where your data originates, which means transferring it back and forth can be comparatively snail-paced. And when you're creating digital twins of factories or tinkering with new products, real-time data analysis is suddenly crucial.

Data security and data sovereignty play a major part. Some companies prefer to keep their data close to their chest, especially when training AI models. Plus, more governments are getting touchy about data sovereignty, with countries including China and India passing laws to keep data within their borders.

Companies like Volkswagen, Caterpillar, and Fanuc are straddling the line between in-house and the cloud, using the best of both worlds to stay competitive. Companies that go down this route will need to consider their options - they can build and manage their own data centers, rent servers, or lease space in someone else's data center. The choice depends on data intensity and future growth projections. Other factors at play include the rising costs of power, land, materials, and labor; construction delays; and the hefty price tag of AI-capable servers. The cloud giants are also making moves, building data centers in new locations, and selling prefabricated data centers.

Meanwhile, the Financial Times warns we’re barreling fast towards a global data center crunch.

Last year, humanity generated a mind-boggling 100 trillion gigabytes of data, and that number will almost double by 2025. Building and expanding data centers means challenges like planning restrictions, the need for more electricity, and upgrades to national power grids.

Data centers are massive energy hogs, raising questions about sustainability and their impact on climate goals. Balancing data center growth, clean energy, and grid capacity isn't a walk in the park. These server-packed behemoths produce a ton of heat, requiring special cooling systems to keep them from overheating (the ideal temperature range is a toasty 64-80 degrees Fahrenheit). Data centers measure their efficiency with a metric called power usage effectiveness (PUE), aiming for a score of 1, which is perfect efficiency. Efficiency gains were impressive until 2018 but have since plateaued.

Now, let's talk geography. Data centers tend to cluster in certain regions due to factors like geography, climate, and access to skilled workers. Think Dublin, Amsterdam, Zurich, and Virginia. These hubs have something special – fiber routes from different continents converging in one place. But this clustering isn't evenly distributed worldwide, and it's causing issues.

Take Ireland, for example. It's home to 82 data centers with more in the pipeline. These data munchers consumed 19% of Ireland's energy in 2022, and that number is set to jump to 28% by 2031. Concerns about the grid's capacity have led to a halt on new connections to data centers in the Dublin area until 2028. Back home, Virginia boasts a whopping 27% of all US data center capacity, and these centers gulp down 20% of the state's electricity.

The good news is that the boffins at the tech giants are working on sustainable solutions. Google, for instance, plans to run on 24/7 carbon-free energy in all its data centers and campuses worldwide by 2030.

The Future of Work: Fewer American Workers Are Quitting

We’ve had the Great Resignation, followed by the Great Rehiring - are we now experiencing the Great Staying Put?

According to the Labor Department, the “quits rate” (resignations as a share of total employment) fell from its 3% peak in April 2022 to 2.3% in August 2023 - that’s the same rate we were experiencing in 2019, before anyone had even heard of COVID.

So, why are fewer people quitting their jobs? It’s tempting to think the slower rate is an indication of an economic downturn and that employees may be less confident they can find another position. But the Wall Street Journal believes people may be staying put for other reasons such as better pay, greater flexibility, and job satisfaction.

The fastest way to boost your salary during the Great Resignation was to make the jump to a better-paid position. Salaries rose across the board, but job switchers’ salaries climbed faster than those who stayed. Recently, however, the premium for switching jobs has cooled significantly.

Also, employees may be nervous about leaving what they consider a safe position in an environment where layoffs and discharges are increasing. Multiple big tech layoffs in 2023 including Amazon, Google, and Microsoft have made headlines and may be causing many to reconsider taking the leap. IT has been particularly hard hit, with the sector’s unemployment rate rising to 4.3%, well above the overall unemployment rate of 3.8%.

But what if workers are staying put because they are - dare I say it - happy? A survey published in May this year revealed U.S. workers’ job satisfaction is the highest it has ever been. Why? It may have taken time, but the positive impacts of increased flexibility, hybrid hours, and other benefits gained during the pandemic may be starting to translate into real satisfaction and higher retention. Those employers who are attempting to wind back the flexible working revolution should take note.

Signs are that jobs are getting harder to find. Data from Indeed.com shows job postings have dropped 15% since September 2022.

This may be one of the reasons why there are many more “boomerang employees” (people who quit, and then were enticed back to their former employer) than we assumed. HBR discovered that across industries, 28% of new hires were actually boomerang hires who had resigned within the last 36 months.

Why do boomerang employees boomerang? HBR found the main reason was that promotion and growth promises made by new employers did not live up to expectations. Predictably, there’s also a financial motivation, with boomerangs at the manager level often receiving whopping pay rises of up to 25% to return to their original roles. The power of social ties to their old colleagues shouldn’t be underestimated, either.

So, the next time a valuable staff member hands in their resignation, make sure you don’t burn any bridges. Give them a positive sendoff, make a note in your calendar, then check in with them at the one-year mark to make an offer of rehiring.

But be aware that the boomerang trend is a double-edged … well, boomerang. Employers should be wary of their new hires fleeing back to their original companies. Defend against this by making sure growth and promotion opportunities meet expectations, ramp up your onboarding experience, and use every means at your disposal to keep new hires satisfied.

AI Insights: OpenAI Gambling With Chips

The word from TechCrunch is that OpenAI is thinking about making its very own AI chips.

Currently, OpenAI relies on GPUs (those powerful graphics processing units) to whip up its AI magic. But the GPU supply chain is getting a serious workout thanks to the AI frenzy. Microsoft is sweating over a hardware shortage, and even Nvidia's primo AI chips are sold out until 2024.

GPUs cost a pretty penny. If ChatGPT became as popular as Google Search, it'd need an eye-popping $48.1 billion worth of GPUs upfront and $16 billion worth every year to keep running. The big question is whether OpenAI's investors, including Microsoft, are up for the gamble.

In other AI news…

• Lenovo earmarks $1.2 billion for AI research - the legacy tech behemoth has no intention of being left behind in the AI race. The move aligns with the industry trend as players big and small plow billions into this transformative tech.

• Google’s AI is making traffic lights smarter - Project Green Light harnesses the power of AI and data from its Maps app to streamline traffic light timings and slash wait times for motorists by 30%. The project has already cut emissions by 10% for 30 million vehicles monthly across a dozen cities worldwide.

The Supply Aside: What I’m Reading, Watching, Listening to, and Thinking About Re: Supply Chain, Work, and Beyond

📕 Read - Clear Thinking by Shane Parrish

I’m a sucker for any new book on behavioral science. Shane Parrish’s Clear Thinking explores the fact that in terms of decision-making, most of us run on auto-pilot, guided by behavioral defaults. This means that we’re simply reacting rather than thinking clearly in the moments that matter most. This book is about deploying our full cognitive ability, applying our reasoning and rationality, and making better decisions when it counts.

What Else I’m Reading

• Weight loss drugs could lead to lower airline costs: Airlines are always looking for ways to reduce weight. What if passengers weighed less? Airlines have the potential to save millions with the surge in popularity of weight loss drugs like Ozempic and Mounjaro.

• New AI tools boosting data efficiency for health care: Microsoft has introduced new AI tools for health care providers that combine data from disparate sources (think health records, lab systems, and medical devices), standardizes it, and makes it accessible in the same place.

• Utah sues Tiktok over impact to children: Won’t somebody please think of the children? Utah’s suit claims Tiktok harms children by deliberately mimicking the features of slot machines so that young consumers become hooked.

📺 Watch - Lost in La Mancha (2002)

A lot of productions have a behind-the-scenes crew filming the sort of standard footage you find in DVD extras menus. What made this one different was that it captured the unfolding of Terry Gilliam’s disastrous attempt to make his version of Don Quixote. From flash flooding to crew disagreements and the lead actor suffering from a herniated disc, Lost in La Mancha is a fascinating insight into how projects can rapidly unravel and fail. Watch for a bemused Johnny Depp’s visit to the set.

👂 Listen - How I Write with Marc Andreessen

David Perell’s How I Write is about being a writer in the age of the internet, and can be immensely helpful for finding what works best for your own writing. This episode featured Marc Andreessen, tech legend, founder of Netscape and the first web browser. But Andreessen isn’t just a tech-head - he’s also deeply immersed in history and culture and shows us how combining tech and liberal arts can be used to fuel your writing.

💡 Think - A Tragic Week

The tragic events in Israel this past week have deeply affected many, myself included. Having visited Israel and spent time on a kibbutz, the recent images and news from the region are heartrending. It's paramount to emphasize that no faith or political stance can ever justify harm to innocent women, children, and the elderly.

I've often shared my admiration for Michael Smerconish, an independent thinker, with a radio show on POTUS. This week, he engaged with callers from various backgrounds, offering diverse perspectives on the situation in Israel. I found it noteworthy how the views differed between older and younger generations, though the exact nature of these differences was multifaceted.

In these trying times, we must be discerning about our information sources. The spread of misinformation is rampant, and in an era where we often gravitate toward echo chambers on social media or news platforms, we must critically evaluate the authenticity of our sources. Recent criticisms from the EU directed at former Twitter underscore this point. 

Charts of the Week

Quote of the Week

Tweets of the Week

Finally...

Thanks so much for reading. I’d love to know what you think about this issue and how I can make it more useful to you.

If you have suggestions or topics you’re interested in seeing me address, shoot me an email at [email protected]!

Want more?

If you’d like to read more of my writing on the supply chain, entrepreneurship, or the future of work, check out my website.

For timely updates follow me on LinkedIn and Twitter!

Happy reading this weekend!

-- Naseem