The Biggest AI Project in the World!

You work in project teams right? Maybe you lead a department? Or, perhaps you run a business?

If any of these are true you will probably have heard the words AI mentioned a thousand times. When you hear about AI, what thoughts come to mind? Maybe large language models that can help with research, productivity tool plug-ins that can make your presentations look more interesting, or perhaps even an app that turns your boring image into a much more interesting and aesthetic looking Anime pic.

All of these use cases for AI can be useful and are increasing in capability and value at what seems like an exponential level of pace. But what if I said there is an even more powerful and potentially useful form of AI that could potentially revolutionize the way we all live our lives forever?

I believe there is a wave of new AI use cases emerging that fall into the category of Embodied AI, which could change how we live and work, to the point where our roles may differ and new careers will emerge.

This post is the beginning of the journey to unpack embodied AI, and an opportunity to start thinking about how leaders and project workers can think ahead and embrace a future where coworkers may increasingly be made up of robots, rather than humans.

You found global talent. Deel’s here to help you onboard them

Deel’s simplified a whole planet’s worth of information. It’s time you got your hands on our international compliance handbook where you’ll learn about:

• Attracting global talent

• Labor laws to consider when hiring

• Processing international payroll on time

• Staying compliant with employment & tax laws abroad

With 150+ countries right at your fingertips, growing your team with Deel is easier than ever.

Get the guide

Embodied AI refers to artificial intelligence systems that are integrated into physical, or sometimes virtual bodies, allowing them to interact with the world around them. This takes the AI away from a screen where we’re currently most familiar with engaging with it, and into the real world to interact with humans like never before. So what are the examples?

1. Home robots - such as automated vacuum cleaners.

2. Industrial robots - such as factory use robots that perform physical tasks.

3. Drones - typically flying robots used for deliveries and surveillance.

4. Smart home devices - including smart speakers.

5. Autonomous vehicles - providing taxi services without a driver.

Embodied AI usually involves robots of some kind, and whilst we have seen many prototypes of robots over the years that have been able to perform interesting tasks, we are yet to see many bots completing high value useful tasks that could solve physical problems otherwise completed by humans.

Benefits

There are theoretically many benefits to embodied AI, even if you only consider the value to businesses. Some benefits could include:

• Training Efficiencies. Unlike training humans, training an embodied AI products could be done centrally and all at once, removing learning curves. For example if you have a fleet of humanoid robots in a factory and one is replaced by a new robot that has never been to the factory, it can arrive pre-set with all the necessary skills to be as effective as the others immediately. Or if one robot learns a new way to save 3 minutes from a task whilst working in a factory, all robots in that fleet could be dowloaded with the same knowledge all at once, compounding the efficiencies to the business.

• Longer working hours. The ability for an embodied AI product to work more than the average human working week automatically enables increased productivity and output, which could lead to reduced operating costs for businesses.

• Employment rights. Businesses would not need to consider worker’s rights for robots, such as unions or holidays or weekends off.

• Undesirable jobs. Robots are potentially most likely to be used for uninspiring, repetitive or dangerous jobs, which can typically be challenging for employers to fill or retain with human labor.

Market opportunity

A McKinsey study has suggested that the global robotics market could exceed $300 billion by 2032.

Other business executives state that the size of the overall AI-driven humanoid robot opportunity could match or exceed the GDP of each country they operate in. The US GDP in 2024 by comparison is $29 Trillion.

My personal view is that the opportunity is more likely to be closer to the latter statement, rather than the former, due to the fact that cheaper labor in the form of AI robots, will create higher demand for additional goods and services at a lower price, which itself could increase global productive output.

I also tend to believe that adoption, at least in a commercial setting will start sooner than we may expect, but will take many years before embodied AI is ubiquitous within our societies.

One key player taking embodied AI very seriously is NVDIA.

Not content with owning the datacenter and gaming GPU markets, NVDIA has been preparing to be the go-to provider of neural engine hardware and associated software for embodied AI, in particular for autonomous vehicles and humanoid robots.

In an interview with YouTube channel Tiff in Tech, NVDIA founder and CEO, Jensen Huang stated that whilst we already have a language model for screen-based interfaces, but we need a language model for the world around us, and the world needs robots as we don’t have enough labour to fulfil current market demands.

Whilst some are concerned that robots may come for our jobs, many are more concerned that we have a global labour shortage and that AI-driven humanoid robots can eventually help to fill this shortage.

This suggests that whilst fears of workers being replaced by robots is valid, such events may not be a reality in the near term due to the current labour gap that has not been filled by humans.

Who else is waking up to embodied AI?

Aside from NVDIA, there is a slew of highly funded global industry leaders investing potentially billions of dollars into the embodied AI opportunity.

Alphabet:

The parent company of Google also owns Waymo, a software company that provides self-driving taxi services in various cities within the US. The cars use various sensors including Lidar and radar and are gradually expanding their footprint.

Meta: 

It has been rumored that the Facebook and Instagram parent company is secretly working on a humanoid robot, which is expected to use Meta’s Llama AI models to form its neural engine.

Apple:

• The iPhone maker was rumored to have been working on self-driving vehicles for multiple years only to scrap it after spending billions of dollars in research. Apple has not commented and certainly has not confirmed that it has ever been working on such a vehicle.

• Latest rumours from people in the know have stated that Apple is currently working on a humanoid robot for home use, and is entering in the R&D stage, meaning that if this were true, we may not see an Apple home bot anytime soon.

• Personally I believe that if anyone can produce a compelling home-friendly humanoid bot, it would be Apple.

Tesla: 

• The electric vehicle and distributed energy company has also been working on a humanoid robot known as Optimus, with an expectation that multiple thousands of them will be built in 2025 for internal use, followed by significant increase in production towards the end of 2026, which could mark the start of use outside of their business.

• In addition, Tesla has been developing self-driving technology for multiple years under the guise of Full Self-Driving (Supervised). This service allows users in the US to have a more advanced driving system akin to Waymo, that still requires owners to be at the wheel and supervise the vehicle.

• Reports have shown significant improvements since Tesla’s move to a neural engine based solution that allows the car to use AI inference to navigate through roads and use a mixture of training data and real-world data to navigate and park.

• Tesla has announced that the first launch of an unsupervised version with paid self-driving cars will start in June 2025 in Texas.

In addition to the above, there is also talk of Amazon and Open AI embarking on developing their own humanoid robots, which leads me to believe that the opportunity in this space is so significant, that for the first time to my knowledge, all major tech players are making notable strides to capture a slice of the market.

This is unprecedented, as most of these companies play in different commercial spaces with some overlap, but never have so many began to operate in the same arena.

Startups and Research Companies

Boston Dynamics

The Hyundai-owned robotics company is not really a start up. They started out as a robotics research firm and have been producing highly agile pre-programmed robots for many years, however, as AI advancements are now making it possible to consider mass-produced robots for productive tasks, Boston Dynamics has started to shift towards the development of lower cost robots with AI brains to compete in this space.

Figure

Startup, Figure AI has managed to court investment and partnerships with significant players such as NVDIA, Open AI and BMW, developing their robot Figure 1, and upgraded Figure 2 in the space of 2-3 years, which is already causing them to be seen as a formidable player.

Apptronik

The team at this start up has experience building over 10 robots including NASA’s Valkyrie robot, and are now embarking on their own Apollo robot geared towards commercial use such as in factories where lifting and carrying is required.

There are many more startups typically based in the US or China, but all seeking to get a head start and capture what could be the largest value unlock in the labour market since the internet.

I have merely touched on this subject and will be sure to return and see how this form of AI is developing, and what it could mean for jobs, in particular for those in leadership and are delivering innovation.

More from CLP