Is there a winning strategy for smaller players in the Artificial Intelligence arms race?

How to compete with the big tech companies in the race for ‘AI supremacy’.

Some see Artificial Intelligence as just another new technology. Or an eternal promise that will not materialize for a long time. However, regardless of your beliefs, the reality of today is that ‘narrow’ AI’s are becoming economically significant. Billions of dollars are put into AI and machine learning research by countries and organizations around the world. What will be the effect of this arms race on the possibility of real Artificial ‘General’ Intelligence to emerge?

Google, not averse to some marketing publicity, last year claimed ‘Quantum computing supremacy’. How would you feel if Google, or Baidu, or ‘China’ in a few years would proclaim ‘AI supremacy’ with the first Artificial General Intelligence, working their way towards Super Intelligence? Even if you believe that the chances are low, you have to admit that the stakes are incredibly high. Any single entity being in control of AI is, in my opinion, a high risk (to say the least). No matter whether that entity is a Western / Eastern corporation or a democratic / autocratic government.

Over the last years countries and large tech companies are pouring in vast amounts of money in AI R&D in an increasing rate. Does this mean they will be the winners in this race? Is there a way for smaller actors in the space to keep up pace and even come out winning? I used to think that there was no way that these giants can be stopped, if they are really pulling their weight (which they are doing in an increasing rate!). But I’ve come to realize that there is an alternative way! A way that will give every person and every entity continued access to new and sophisticated AI services. This comes in the shape of a platform that is democratic, diversified, and designed for exponential growth of AI-services and -perhaps- even emerging general intelligence. Skip to the bottom to find out more on this platform, or read on for some more context.

Global investments in artificial Intelligence

To set the stage, let’s start off with a few numbers on global AI investments:

  • The US plans to double its non-defense AI budget from 1 billion in 2020 to 2 billion in 2021.
  • According to this article from MIT, China’s estimated expenses on non-defense AI in 2018 were already 5.7 billion
  • China has the ambition to become the world leader in AI, aiming for a 150-billion-dollar internal market in 2030.
  • Way behind China and the US, also other countries such as Russia, Canada, and most larger countries in the EU are investing heavily in AI (read more in this article)

Opposed to government spendings, the US still leads in corporate spendings, followed by the EU and China. Budgets are surging accross the board: in retail, banking, manufacturing, healthcare, etc. According to this article from IDC: “spending on AI systems will reach $97.9 billion in 2023, more than two and one half times the $37.5 billion that will be spent in 2019.” or, an anual growth rate of 28,4%.

All these staggering numbers add urgency to the question: what can you do as a smaller player in this AI-arms race? Is it possible to keep up and develop a meaningful AI strategy against these giants without breaking the bank?

Common AI strategies

Let’s have look at 3 different strategies:

  1. A first option is to leverage the AI research of others by applying ‘smart’ tools. There are countless options. in the top 3 largest usecases there are 2 that fit in perfectly in many common eBusiness strategies: On nr.1: ‘smart chatbots for customer service’, and on nr. 3: ‘sales automation and recommendations’. But the possibilities are endless. I’ve been quite surprised about the service that applies AI to assess job candidates in an online interview. Sounds a bit scary, but (if trained correctly) the algorithm will be less susceptible to human bias, giving better and perhaps fairer results.
  2. Another option is to mix and match technology or algorithms that are already available and make something out of it that will give you a unique advantage. Think about a translation service that uses ‘Speech to text’ functionality, translation services and image recognition to automate the creation of subtitles to a video. This requires less fundamental research and can be commercialized sooner. But it also brings less of a competitive advantage and involves dependencies to other entities in the market.
  3. The third, most ambitious and most complex option would be to compete head-on with the large companies and commit to R&D and experiment with new technologies. For the majority of the large organizations, let alone the SME’s this is not realistic or sustainable, unless…

…Unless there was a common platform where small and large stakeholders can collaborate by:

  • Contributing their AI services and be rewarded relative to their success.
  • Finding and using the contributions of others on a pay-per-use basis.
  • Combining different services for a complex usecase

With sufficient scale and adoption such a platform could give large, but also medium- and small companies a purpose and utility in the arms race to AI. It would boost all three AI-strategies above and make them reinforce each other. With -potentially- thousands of companies contributing and competing with each other on specialized or general-purpose AI components this platform could not only have its place between the corporate giants, it could even come out with a clear competitive advantage.

The limitations of large companies in innovation

There is yet another angle to look at the competition between large and small companies. Large and centrally organized entities do have advantages when it comes to executing on a strategy. Nevertheless, disrupting innovations often come from smaller organizations. Here are some reasons why larger organizations fail in real innovation:

1. Any centrally managed organization will put some limitations on how time spent, on what is the best way of doing things and what are the preferred outcomes. This is a strength and a necessity when working with limited resources, but this will also limit the variety of different approaches and hinder the discovery of unexpected positive outcomes.

2. By their nature big companies will prefer secrecy. This constrained openness, enhanced by the company’s culture will direct their efforts in a specific -narrow- direction. It will be hard for new employees to influence this general direction so this will be increased over time, unless a disruptive event occurs, such as a visible and important failure, or a change in management.

3. Although the large tech companies seem to have almost unlimited resources, they are in fact, limited. If in direct competition against a platform that is supported and seeded by thousands of organizations, each with their own drive, purpose, ánd specialized knowledge and experience, they may find themselves outnumbered (or ‘out-idea-d’ or ‘out-budgeted’)

Given all this, I would argue that, if a successful common AI platform would emerge on a short- or medium timeframe, the top players that feel confident and secure in their ability to lead the next revolution by their own secret R&D, are actually the ones that will be limited. Their strategy, culture and governance structures may not allow them to participate in a large, common structure. They will be alone against the masses.


Well, all fair and nice you might say, but this is highly theoretical. There are numerous reasons why such a platform will never work, starting with the question of design, development and most importantly: ownership. Who will be in charge of this platform that is designated to become of strategic value to so many organizations? There is no board that would be acceptable to all and still be effectively decisive or just plain realistic. Here comes the surprise: (You didn’t expect to get all this way just to find out there is no solution, right?)

This platform has already been designed, It has been developed for the major part, and the ownership dilemma has been solved. This might be your ticket to join the AI-arms race in a meaningful and profitable way:

To summarize: SingularityNET is a decentralized protocol that allows anyone to publish their AI algorithm and monetize it on a global scale. A governed marketplace makes all AI services findable and will push the most capable, usable and efficient implementations, while suppressing malicious or unethical algorithms. Not unlike the app store, but on top of a protocol that is owned by no-one and that anyone can use.

In its current state the platform is ready and most suitable for strategy 3 mentioned above: Publishing and monetizing new AI services. But as the number of services is growing, so will the utility of the platform for scenario 1 and 2: applying the offered services for your own business,

The most exciting part of the platform, the part that makes me believe that it can compete with the current forces in AI, is the designed capability of letting individual AI components communicate and collaborate. SingularityNET is designed around the capability to create dynamic, complex, multilevel hierarchies of collaborating algorithms. A robust networked system of collaborating parts, not unlike the internet, but based on networked AI’s instead of information.

The internet of AI

With a bit of imagination we can see this platform as the third stage in the network economy:

Stage 1: Publishing and consuming content on the internet
I remember the exciting promise of the early days: “Every person or organization can become a publisher of digital content”. This principle gave rise to a whole new ecosystem of small and large organizations, some of which evolved into the tech giants we all know today like Amazon, Google and Alibaba.

Stage 2: Publishing and monetizing apps in the app store
The mobile phone became the new battle field for digital services. The app store enabled anyone to publish Apps, giving rise to new initiatives such as Uber, AirB&B, Spotify and Netflix.

Stage 3: Publishing and monetizing AI services on a decentralized platform
SingularityNET offers every individual and every organization the opportunity to publish AI services. This platform is more comparable to the internet than to the app store, because a) it is owned by no-one and b) it offers completely new network effects. On the internet, hyperlinks enable the reader to navigate from one article to another. On SingularityNET, algorithms will be not only be able to interconnect but also incorporate and build on each other’s capabilities. When you are accessing a service on SNET 2 years from now, you may have no idea how many subservices are called and applied to deliver the final output. This will undoubtedly open up the way for new business models that will be hard to imagine today and may give birth to a whole new category of tech giants. And this revolution has barely started yet!

Rising complexity?
Going back to the question in the title of this article: “Is there a winning strategy for smaller players in the Artificial Intelligence arms race?”
You may argue that creating and publishing a new AI is in a completely different leage than just adding some content on a page or even developing an app. But remember that in the early days of the internet, publishing a page on the ‘world wide web’ was just as exotic to most as the development of an AI algorithm is today. The same goes for apps; Let yourself (or your children :-) be inspired by these teenagers that became rich in the app industry. How old was Mark Zuckerberg when he founded Facebook?
Also the platform for AI services will be democratized. Pretty quickly you can leverage the services created by others for your own purposes (Strategy 2 and 3 above) Eventually tools will surface that make development of new AI’s on the platform easier. But for the moment AI development capabilities are required to publish new services on the network. However, this ‘competitive advantage’ is well in reach for businesses of any size.

Time to ask yourself: What will be your place in this new revolution? Do you see the opportunities? Can you come up with a winning strategy?

This is just a small introduction. If you are interested in more background on SingulartityNET and its surrounding ecosystem of platforms, please read my other articles on the subject:

Of course much more information is available on




Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

AutoML will pave the way for the dispersion of AI

Google Launches New Pixel Phone with ML Front and Center

Machine Learning in a Nutshell

Explore the benefits which MNCs are getting from AI/ML and emphasizing the enhancement of AI…

Bioprocessing 4.0 Will Require More Than Analytics

The Five Senses of Robotics

Artificial Intelligence is the tool eCommerce needs in COVID times

Artificial Intelligence 101 How to get started — SKILL MONKS

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Jan Horlings

Jan Horlings

More from Medium

Dystopian or utopian future

UN human rights chief to visit Xinjiang in May

Awaken Dream Synergy

The Ethical Machine Podcast