TSMC’s CapEx Is Hitting Record Levels

• The first sentient AI?
• Air taxis – the future of urban transportation…
• The surprising conclusion to Uber and Waymo’s legal battle…

---

Dear Reader,

It’s an odd feeling of anticipation knowing that the Federal Open Market Committee (FOMC) meetings are taking place – the outcome of which could have such a material impact on the volatile market conditions that we are experiencing right now.

And a few hours ago, we got our answer. The Fed officially raised rates by 0.75%, the largest increase since 1994. We’ll take a closer look at this decision and what it means in future issues.

I can’t imagine the weight of those decisions. The inputs for the decision-making are practically changing daily. For example, the 30-year fixed mortgage rate just shot up to 6.28%!

But we’re not going to feel bad for them. After all, it was the government policies and the Fed’s lack of action last year that got us into this mess in the first place.

With all of this short-term certainty and fear in the market, I’ve been keeping a close eye on indicators that allow us to step back and have perspective. Regular readers will know that keeping a close eye on Taiwan Semiconductor Manufacturing (TSMC) is a great way to do that.

And last week’s news will probably come as a surprise to most of us: TSMC announced that it is increasing its revenue guidance for this year. 

The world’s largest semiconductor manufacturer expects revenue to jump to “at least” 30% this year over last year’s performance. What’s even more impressive is that this adjustment comes just two months after TSMC guided to mid-to-high 20s growth.

If that isn’t enough positive news, TSMC is also spending an incredible $44 billion this year on capital expenditures (CapEx), primarily to build additional manufacturing capacity. That means that its outlook is bullish not just this year, but years into the future.

We can see how TSMC’s 2022 CapEx stands in contrast to the last few years of investment:

Would this be happening if TSMC wasn’t seeing demand and economic growth? No way.

It’s so easy to get wrapped up in all of the negative market sentiment, which is why it is so useful to step back and look at the bigger picture.

If we were seeing a decline in investment and revenues in the semiconductor industry, I’d be very worried. But that’s not the case… In fact, we’re seeing exactly the opposite.

It can be emotionally difficult to reconcile two kinds of market sentiments that are starkly different. The difference can usually be boiled down to a difference in the timeframe. We may see negative market indicators over the coming days/weeks, whereas at the same time we can continue to see strength in the coming months/years.

Has Google’s AI become self-aware?

A big development at Google is taking the artificial intelligence (AI) community by storm this week. This is wild.

A computer scientist and AI researcher at Google just proclaimed that LaMDA, Google’s massive AI language model, has become sentient. While this may seem like an off-the-wall claim at first, the individual was very thoughtful about coming to this conclusion.

In fact, the researcher had many long, human-like conversations with the AI. He even published one of these for everyone to see. And in these conversations, the researcher explored some abstract topics with the AI.

For example, the researcher asked LaMDA directly about the nature of its consciousness. Check this out:

LaMDA’s Conversation

Source: Medium

This is something that’s far and away more complex than anything we have ever seen from a natural language processing (NLP) AI before. The AI even differentiates its own abilities from those of Eliza, another language model.

Of course, this has created a stir throughout the tech community. That’s largely because there will be a time when an AI indeed becomes sentient and self-aware.

And when that happens, it will become capable of some incredible things. A self-aware AI will be like an unpredictable human with super intelligence.

That said, this is not one of those moments.

As the computer science community pointed out, LaMDA is trained on billions of parameters specifically to develop pattern recognition. The AI is great at recognizing patterns in speech. This gives the algorithm the ability to choose the optimal response to any question or comment during a conversation.

That’s what appears to be happening in the above conversation. The researcher led LaMDA down the path of talking about sentience, and the AI simply continued the conversation in that direction.

Still, this has some profound implications.

Clearly, Google has an AI capable of holding lengthy conversations intelligently. Most of us chatting with LaMDA would think it was a human if they weren’t told in advance that it was an AI. Google could do some incredible things with this.

For one, I could easily envision Google extending LaMDA to its Google Assistant product. This would give everyone with an Android device access to the AI.

And LaMDA could do all kinds of things for us. The AI could book appointments, order groceries, pay utility bills, and perform many other menial tasks that we do on a weekly and monthly basis. And it would complete these tasks in milliseconds, saving us a lot of time and effort.

However, this level of interaction with the AI would give Google even more data about our lives. Google will know even more about what we do, who we associate with, and what we buy than ever before. That would expand Google’s profile of us.

Even worse, Google could use LaMDA to subtly push political narratives once consumers grew to trust it.

This would be an incredible tool for deception and manipulation if used for nefarious purposes. And we have already seen how Google is more than willing to try to push a favored narrative by filtering and even censoring search content.

So we need to be very vigilant about technology like this. We can use advanced AI for incredibly useful things… But it also poses a major risk if used incorrectly.

Volocopter’s first test flight was a big success…

We have talked quite a bit about the rise of air taxis recently… And we just had a big development on that front.

Volocopter, a company we profiled in March of last year, just completed its first test flight of its latest invention. Here it is:

Volocopter’s Test Flight

Source: Volocopter

Here we can see how smoothly Volocopter’s vertical takeoff and landing (VTOL) aircraft operates. It’s a four-seater craft that has six rotors on top of the fixed wings to provide the lift. And it has two electric fans on the back that power the aircraft forward.

And what’s great about this design is that the fixed-wing structure allows the craft to go farther and faster than the first generation VTOL we looked at last year.

That craft was just a two-seater, designed for urban transport, that could travel 62 miles per hour (mph) for up to 22 miles. This four-seater can reach speeds of 155 mph and travel up to 60 miles.

So this is an aircraft that could be incredibly efficient as an air taxi. And to that end, Volocopter is still gearing up to provide air taxi services at the 2024 Summer Olympics in Paris.

Bigger picture, we are watching the birth of a brand-new industry right before our eyes. The future of transportation will include Uber-like services, featuring these VTOL air taxis.

And of course, there will be plenty of great investment opportunities as the industry shapes up…

An interesting turn for Uber and Waymo…

We just had an interesting turn in what used to be a battle between Uber and Waymo. I’ll explain with a little history…

One of Waymo’s engineers left the company in 2016 to start his own autonomous driving company called Otto. Then Uber came in and acquired the company. It became the basis for Uber’s Technologies Division.

However, it turned out that the engineer had stolen 14,000 highly confidential files from Waymo to start the new trucking company. That means Uber built its self-driving car initiative using stolen technology, and Waymo filed a lawsuit in response.

That led to a long, drawn-out legal battle, and the engineer was ultimately found guilty. He received an 18-month prison sentence. Meanwhile, Uber and Waymo reached a settlement.

To add to the intrigue, Uber spun out its Uber Technologies Division into a new company called Aurora. We talked about that back in December 2020.

As part of that move, Uber took an equity stake in Aurora, and it has two seats on the company’s board. So the industry assumed Uber would partner with Aurora for its self-driving technology.

But it hasn’t yet…

In a surprise move, Uber just announced that it is partnering with Waymo to deploy its self-driving tech in the Uber Freight division. This is Uber’s “ride-hailing” service for cargo. It matches people who want to ship goods across the U.S. with truckers who can haul goods.

So this move will incorporate self-driving trucks into the Uber Freight division. And that could be a real boon for autonomous trucking in the U.S., which in turn could help alleviate ongoing labor shortages in the industry.

Looking forward, this move could lead to attractive partnerships with large trucking companies that also want to incorporate self-driving trucks into their services. I expect we’ll see some big deals inked later this year.

And as these deals kick off, we’re going to see self-driving technology really go into overdrive. If any investors want to know the best way to play this tech trend when that happens, you can go right here for more info.

Regards,

Jeff Brown
Editor, The Bleeding Edge

---

Like what you’re reading? Send your thoughts to [email protected].