“`html

This Dot-Com Survivor Thinks AI Boom Feels More Like 1997 Than 1999 — And Why Holding Cash Still Makes Sense

If you were around during the dot-com boom, you probably remember the rollercoaster of emotions — the excitement, the FOMO, and that nagging feeling that “this time it’s different.” Well, seeing how AI investments are booming right now, it’s starting to feel a lot like those early internet days, especially around 1997, not the crazy frenzy right before the crash in 1999. And that’s why, even with the hype and sky-high stock prices, I still recommend keeping more cash on hand than usual.

Let me be clear — the AI build-out is very real. Companies are throwing billions at Nvidia chips, cloud infrastructure, and startups promising all kinds of generative AI magic. The big players like Microsoft, Google, and Amazon are in a fierce race to dominate. The S&P 500’s recent gains? Mostly driven by a handful of AI “winners.” But here’s the catch: most teams are still struggling to actually deploy AI at scale. Getting from a flashy demo to real, bottom-line impact takes time, effort, and often more trial and error than most expect.

Think back to 1997. The internet was clearly a game-changer, but the killer apps weren’t in place yet. Amazon was just an online bookstore, Google wasn’t even a household name, and Pets.com was still two years away from both its IPO and its spectacular crash. Investors were curious and excited, but the infrastructure wasn’t quite ready, and business models were still pretty speculative.

That’s exactly where we are with AI today. The potential is massive, but so are the unknowns. For every company genuinely using large language models to boost productivity, there are five more just adding “AI-powered” to their pitch decks without much substance behind it. Integrating AI into existing systems is slow, costly, and often underwhelming at first. The risk? Investors get caught up in hype, expecting instant dominance — just like with the internet bubble.

Yet, the market keeps climbing. Take Nvidia, for example — its valuation multiples look a lot like what Cisco or Intel had back in 1999. Maybe that’s justified, but we’ve all seen how quickly enthusiasm can turn sour when reality sets in.

So, what’s the smart move? Here’s what I’ve learned over the years:

1. Hold more cash than usual.

Yeah, that might sound dull, but cash is your safety net. It’s the “dry powder” that gives you options when the music stops. This isn’t about trying to time the market — that’s a dangerous game — but about respecting the cycles and not getting swept away by every shiny new thing. Investors who kept some cash during past bubbles had the advantage when bargains popped up after corrections.

2. Focus on fundamentals.

Look for companies with real cash flow, not just flashy promises. During the dot-com bubble, too many investors ignored basic valuation rules and paid the price with years of dead money. Today’s best AI bets have strong competitive moats — think solid data advantages, wide distribution, and the patience to handle setbacks.

3. Don’t put all your eggs in the “AI” basket.

That means owning some steady, old-school businesses — industrials, healthcare, or even short-term bonds. Overconcentration in any hot theme can blow up your portfolio when sentiment shifts. It’s tempting to ride the wave when your AI picks are soaring, but nothing goes up forever, and exit doors get crowded fast.

When does this advice not apply?

If you’re truly in it for the long haul — like decades, not just a few years — then ride out the ups and downs. You’ll likely face paper losses, but if you never sell in panic, you could come out ahead. The issue is most people can’t stomach a 40% drop and end up selling at the worst moment.

Also, if you’re in the AI space yourself or have a deep insider edge, you might spot opportunities before others do. In that case, leaning into your knowledge makes sense. But for most folks, chasing the hottest AI stocks without that edge is a fast track to frustration.

Liquidity matters more than you think

One key lesson from past crashes: when everyone piles into the same trade, liquidity can evaporate in a panic. We saw this in the dot-com bust and again in the 2008 financial crisis. Holding cash isn’t just about missing out on some gains — it’s insurance against getting stuck when the market turns.

Of course, cash means you might underperform during a strong rally. If AI really is the next big thing, you could miss some upside. But from experience, missing the last 10% of a rally is way less painful than suffering through the first 30% of a crash.

What’s different this time around?

AI has actually come further along the adoption curve than the internet was in 1997. The infrastructure’s stronger, millions of users are onboard, and the business cases for automation and workflow improvements are clearer. But markets are still pricing in perfection. The reality of AI at scale is messy — with lots of trial and error, regulatory hurdles, privacy concerns, and technological limits that could slow things down.

Here’s what I’m keeping an eye on:

• How fast enterprises move from pilot AI projects to real adoption
• Whether the cost of compute power stays affordable or spikes
• If AI actually starts moving the productivity needle in the broader economy

If these signals improve, I’ll get more optimistic. But for now, the smart play is to keep some cash, stay diversified, and avoid chasing every “next big thing.”

At the end of the day, every cycle feels unique, but the emotions don’t change — greed, fear, hope. The investors who survive and come out ahead are the ones who remember that. I got through the dot-com crash by holding cash, focusing on fundamentals, and tuning out the noise. That’s a lesson worth holding onto as we ride this AI wave — whether it’s 1997, 1999, or a story all its own.

“`