Dear Reader,
In what is likely the largest publicly announced AI-related workforce strategy, IBM CEO Arvind Krishna recently provided some projections regarding the imminent disruption of the labor market.
An announcement like this is useful, as it provides us with some early insights about how and where large corporations will choose to employ the latest breed of powerful artificial intelligence (AI). Information like this has both investment implications, as well as useful information for us regarding future career choices…
Krishna’s plans are bold, and probably unpopular amongst certain IBM employees. He announced that the company plans to suspend hiring for any roles that it believes can be replaced by AI.
Broadly, Krishna’s target is back-office functions, and he specifically named human resource functions. He believes that many of the HR functions will be better served using AI technology, and he suggested that at least 30% of those back-office jobs will likely get replaced over the course of the next five years.
Given that IBM has about 26,000 back-office employees, a 30% reduction would amount to 7,800 jobs. That may sound like a lot, and it is. But within the context of IBM’s overall workforce of about 260,000 employees, it amounts to just 3% of IBM’s workforce.
But we should keep in mind that IBM is just one company. Medium to large enterprises, and hopefully bloated and wasteful governments, will employ the technology to improve both business operations as well as customer satisfaction. And those that don’t will become uncompetitive quickly.
Some obvious jobs that will be near-future targets for being replaced by AI are: customer service, administrative tasks, accounting and bookkeeping, marketing, order taking (i.e. at a restaurant), data entry, and data analysis jobs. These are just a few obvious categories.
Krishna has been tasked with modernizing IBM and getting the company back on a growth path after years of stagnation and a weakening competitive position. I’m not surprised that he’s willing to talk openly about his plans for putting AI “to work” for IBM.
But we can be sure that even though most executives aren’t making bold pronouncements like Krishna, they are not only thinking about it. They are developing their own near-future plans to do exactly the same.
The reality is that any technology that can reduce operating expenses, increase gross margins, improve customer satisfaction, and increase sales will be adopted. Quickly. And that means that Krishna is likely being too conservative with his projection. This shift will happen in less than five years and certainly be more than 30% of back-office functions.
We’ll be looking for AI companies that enable these kinds of workforce transformations, as well as those companies that are fast to adopt the technology giving them a competitive advantage.
As a perfect example, I wrote about Palantir on May 3. I highlighted how its employment of generative AI in its existing product offering was a smart strategy that would allow the company to quickly leverage its existing customer base for new sales.
Since then, Palantir’s stock has rallied about 35%, most of which happened in the last two days following Palantir’s earnings announcement. The news was clearly good. Palantir, already known as an AI/ML (machine learning) company, announced a profit for the quarter and a forecast that profits would continue through the rest of the year.
More telling, however, was a comment from the CEO of Palantir concerning the demand for its new AI product offering – “unprecedented.”
Artificial intelligence (AI) startup Hugging Face just released a new generative AI. It’s called HuggingChat. And it’s very much a competitor to OpenAI’s ChatGPT. But there are some key differences…
Hugging Face is backed by venture capital (VC) powerhouses Sequoia and Lux Capital. And unlike OpenAI, it’s pursuing an open-source model for HuggingChat. That means the code is open for anyone to use. It’s not proprietary like ChatGPT’s code is.
What’s more, organizations are free to modify the HuggingChat code to create their own applications based on the original generative AI. The key here is that any organization who does this does not have to pay licensing fees to Hugging Face. The primary cost to users of Hugging Face will be their own costs for computing power to further train and operate the HuggingChat generative AI.
This is important because the industry has been anxious to get alternatives to ChatGPT. For a few reasons.
To start with, as we said, ChatGPT is proprietary. No one can review the code or the data sets the AI was trained on. And OpenAI controls who may or may not license the tech.
Plus, Microsoft owns a controlling interest in OpenAI. That means Microsoft could throw its weight around if it ever wanted to. It already has, actually. Microsoft is choosing which companies it will allow access to ChatGPT and GPT-4.
By way of example, Microsoft won’t allow Alphabet (Google) access to ChatGPT/GPT-4. It’s a competitor. And after all, Microsoft is doing whatever it can to steal web browsing business away from Google. Its entire strategy is built around OpenAI’s technology, and it has already paid at least $13 billion to control it.
So I expect HuggingChat will gain strong adoption quickly. The code’s transparent, there are no strings attached, and the AI can do basically anything that ChatGPT can. It’s not as advanced as OpenAI yet, but that’s a problem that can be fixed quickly with additional capital and workforce. If I had to make a guess, I’d say that HuggingChat is about three months behind ChatGPT.
So I fully expect Hugging Face to leverage this product launch into another major VC raise. That would give the company the capital it needs to train HuggingChat and launch a newer, smarter version later as early as the third quarter of this year.
What we’re seeing is a natural evolution in the industry. It was never going to be just ChatGPT. The potential of this technology is already deeply understood, as is the ability to monetize the technology. That’s why we’re already witnessing a veritable arms race to dominate the generative AI space in the years ahead.
Big news from autonomous driving startup Cruise. The company just made two huge announcements.
As a reminder, Cruise is the self-driving company that was partially acquired by General Motors (GM) in 2016. And in February of last year, Cruise launched a new, autonomous ride-hailing program in San Francisco.
At first, Cruise only had approval to run between the hours of 10 p.m. and 5 a.m. local time – when the streets were less crowded. But Cruise just announced that it has approval to operate the robotaxi service in San Francisco 24/7. This is a huge milestone.
And that’s not all…
Cruise also announced that it is gearing up to launch a robotaxi service in Dallas, Texas. That will be its next city.
This is a huge move. Dallas is a major metropolitan area that’s been growing fast. In fact, as of 2021, Dallas-Fort Worth was the fastest-growing metropolitan area in the country.
And when I look at Dallas – it should be far easier for the self-driving taxis to handle than San Francisco. Dallas is far more spread out. Its streets are flat and simple compared to San Francisco. And the weather is generally better compared to San Francisco without any of the fog.
So I see this as a great choice for Cruise’s next robotaxi service.
For readers in Dallas, this new service should be ready for launch in the months ahead. If anyone decides to test it out, I would love to hear about it.
Back in December, a private Japanese company called iSpace set out to land a spacecraft on the Moon. They were the third private company to attempt this challenging task.
iSpace’s moon lander is called Hakuto-R. Here’s an artist’s rendering of what it would look like on the lunar surface:
Source: iSpace
And unfortunately, an artist’s rendering is all we’ll see of Hakuto-R on the moon. That’s because the vessel crash-landed on the lunar surface.
The lander did manage to get into the Moon’s orbit. That alone is a big victory. From there, it began making its way down to the lunar surface… but then mission control in Tokyo lost contact with the craft.
The post-mortem revealed that Hakuto-R ran out of fuel needed to slow itself down on its descent to the Moon’s surface. Most likely the lander was shattered upon impact.
This demonstrates just how difficult it is to land a craft on the Moon. Because there’s very little gravity and no atmosphere, it requires a lot of fuel to slow down the spacecraft for a controlled descent and landing.
So iSpace became the third private company that’s failed to pull this off. The last two were back in 2019.
That said, this isn’t surprising. We should expect a fair amount of trial and error when it comes to space exploration.
Most don’t know this, but even NASA had failed lunar missions before Apollo 11. In 1958, NASA sent a probe – Pioneer 1 – on a mission to achieve lunar orbit. But the probe failed to achieve orbit and was eventually destroyed upon entering Earth’s atmosphere.
But what’s exciting is that these early attempts are a precursor of what’s to come. Private companies will not only be launching spacecraft to orbit, they’ll be exploring our solar system for minerals and resources that will ultimately become critical for a continued human presence in space.
And thanks to SpaceX and its radically lower launch costs to orbit, this new space economy is accessible to large and small companies alike.
Regards,
Jeff Brown
Editor, The Bleeding Edge
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
