SALN #10 – Two strategies to ride out the AI avalanche.

Reading time: 7 minutes

It’s happening, and it’s happening at breakneck speed.  

Robots are eroding repetitive manual labor, leaving less skilled workers unemployed.

Most of us have seen car factories full of robots.  

Many people predicted this.  

But this astounds me. 

AI is replacing knowledge workers. The jobs of copywriters, accountants, business administrators, and even consultants are at stake. This is happening faster than we can come up with a global standard for EV charging plugs.  

This is an avalanche, not a glacial movement.  

We have no idea how big it will get. 

Today, we witness the avalanche breaking free at the mountain top 

The last few months have only been the start of a new revolution.  

In the next five years, we’ll find out how quickly this avalanche accelerates and how large it becomes. Here are a few facts to help you understand the dynamic. 

Fact 1: No technology in human history has seen such rapid adoption. Nothing. ChatGPT grew to 1 million users in 5 days and to 100 million users in 2 months. 

OpenAI may not be able to dominate the AI industry. However, this technology will completely dominate all other industries. 

Overwhelmed by this avalanche, all major technology companies (Google, Microsoft, SalesForce, and SAP) are working to incorporate AI into their products. Soon, all office-based jobs will be AI-assisted in some way. 

Fact 2: Computing power is increasing at an exponential rate. There is no indication that Moore’s Law will run out of technical possibilities anytime soon.

Even if silicon-based chip production reaches physical levels, the next push to increase computing power may be in sight: Google CEO Sundar Pichai recently announced a breakthrough in error correction for quantum computers, removing a major impediment to the practical application of this technology. 

Fact 3: Faster recursive iteration cycles for technologies and startups that create them are more efficient and shorter. “Recursive iterations took years, just like iPhone one to iPhone two,” says venture capitalist Chamath Palihapitiya. “App store took two years. Years—maybe months for really aggressive and disruptive things—measured everything. These groundbreaking innovations now take days and weeks. Startups today: Why create an MVP with a ten- or fifty-person company? Three or four could do that. Because all of our capital allocation models were based on a different technology. We now need a couple hundred thousand dollars instead of 10 or 20 million in the first round and 100 and 500 million in the second. Midjourney and others can now bootstrap to massive size with little capital.“ 

Fact 4: The exponential growth of AI chats powered by Large Language Models continues. As a market for access to training data emerges, so will price discovery for large data sets used in Language Models. Soon, we’ll hear about multibillion-dollar data deals for platforms like GitHub, StackOverflow, Quora, and possibly even Amazon ratings. The learning rate will accelerate as soon as these datasurfaces for LLM training and cast in contracts and license agreements. 

AI is a greater revolution than the internet or mobile phones. 

The avalanche has just left the mountain’s summit. As we get closer, we’ll be able to see how significant the change will be for economies, society, and the way we live in the future.  

Many AI experts believe that human-level artificial intelligence will be developed within the next few decades, and some believe it will be available much sooner.

Suppose we build machines that outperform us at any task or process we encounter daily. Using this technology, our abilities are catapulted to previously unattainable heights. We can overcome obstacles and expand even further.  

However, there is still some skepticism. Not everything in life is a task, and not everything can be done by a computer. 

However, the foundation for the impact of AI is the digitization of the business world. 

The way we work is changing profoundly 

This decade will be defined by AI’s expansion into all sectors and aspects of human existence, including business, finance, transportation, and medicine, as well as science, research, and policymaking. In the future, we may not be able to make major decisions in business or in our personal lives without the assistance of AI. 

As a result, some tasks and jobs will be eliminated, including complex jobs such as copywriting. These jobs are entirely computer-based, with the keyboard serving as the primary interface. The avalanche will prioritize these and will take down some well-paying knowledge jobs.  

Most tasks and jobs will be enhanced. Early research indicates such productivity gains, a trend that could easily be multiplied 10 to 50 times in the future due to the speed and accessibility of information. 

New positions will be created, such as Machine Learning Engineer, Prompt Engineer, and AI Strategist. I see them as the AI’s builder, teacher, or shepherd, developing, discovering, growing, and deploying capabilities while mitigating the AI’s risks and errors. 

Position yourself for that avalanche   

This, however, is a wake-up call.  

You may not be directly affected right now. However, you may begin to position yourself for that possibility.  

For example, becoming a pilot or a truck driver today is the equivalent of becoming a sail ship captain in 1850. In the future, there will be no hand-steered transportation. Most transportation will be fully automated. Humans will be limited to nostalgia and leisurely activities when driving a car or flying an airplane. 

We don’t know when the transition to fully automated transportation will occur, but we know it will happen. 

AI is created through a learning process 

Machine learning is the most important component of developing an AI. 

In Machine Learning, we figured out how to collect, organize, and process massive amounts of data. Then, using statistical techniques, we created programs that detect data patterns. These patterns are programmable in models such as ChatGPT’s Large Language Model (LLM): the model “learns” or acquires a new algorithm, that can solve tasks. These models later respond to complex tasks guided by prompts in so-called chats. 

The human and the machine learning process are conceptually quite similar. 

But humans have no chance of learning faster than a machine because AI works literally at lightspeed and 24/7. 

But we learn differently and different things. 

  1. Human perception is much broader and more diverse than the sensors of a machine, a car, or a computer. Machines cannot smell, feel texture and rhythm, understand a specific social context, or even read emotions. Consequently, the data sets for the AI will remain incomplete for a long time, and the AI will fail on tasks requiring deeper feeling. 
  2. Humans learn over time judgement and acquire wisdom. A test case for autonomous driving is the “Trolley Dilemma,” where the vehicle needs to decide whether to run into a group of pedestrians to avoid a lethal accident. To overcome this dilemma, judgement is required that allows for error and grace. On the other hand, an AI creating an error is broken and should never be deployed for real world tasks. 
  3. Plus, we humans have a drive to self-directed adaptation, and a general willingness to learn. A human that loses his job is free to choose a different occupation and learn the required skill. Make a robot jobless and in most cases, the robot gets disassembled. 

There is much more that we humans can do better than a machine, like grasping context, dreaming and being creative, or being artistic and innovative. 

On the contrary, every task in which unique data is processed in a specific task environment will be completed by AI sooner, safer, and at a lower cost than a human. After any potential lack of digitization is overcome, most employees working in business processes will be either enhanced or replaced by AI. 

This is terrifying.  

On the other hand, you do not have to let this frighten you in any way. 

Here are two alternatives for repositioning yourself 

  1. Expand your skills as a designer, cook, craftsman, or builder if you are more of a sensory person. These are all examples of people who create things. You also can work as a teacher, trainer, instructor, safety advisor, counselor, football coach, or music teacher to pass on your knowledge to others. Utilize AI solutions for your accounting, advertising, scheduling, and content creation but focus on developing your craft and perceptual skills. 
  2. Studying artificial intelligence is a promising idea if you are someone who thinks a lot or if your job involves working with computers. Embrace the difficulty, and you will develop. Become an architect of artificial intelligence, an educator, or a shepherd. Boost your productivity while also enhancing your personal brand and the value you contribute. 

In five years, we will be able to assess the magnitude of this change.  

AI does not necessarily pose a threat to us, but it can compensate for one of humanity’s major shortcomings: the constraint in logical reasoning in solving complex problems. We will be able to eliminate tedious tasks and solve century-old issues in science, business, society, health and many more. 

Our slope will be powered by the avalanche. We’ll have a good time on the ride this is going to be amazing. 

Before you go: 

The Shanghai Motorshow last week served as a potent illustration of the capabilities of Chinese OEM and the direction of the automobile industry towards EV and AV. Bloomberg published a piece on autonomous driving this week. In a word, China’s government places a lot of focus on the development of an autonomous vehicle regulatory framework. As a result, various Chinese businesses release their AV devices, including DriveGPT. I enquired of the article’s author, Anjani Trivedi, if she believed that Tesla’s AV approach had reached a dead end. “I believe it’ll be a troubled path forward. However, China could help it.” 

The scales just seem to be tipping toward the world’s largest auto market. 

Verwandte Artikel