What Does Unlikely AI Do? Pushing the Boundaries of Artificial Intelligence


Artificial intelligence (AI) has made tremendous strides in recent years, with systems able to beat humans at complex games like chess and Go, generate human-like conversations, and even create art and music. However, most current AI systems are narrowly focused on specific tasks and lack the flexibility and general intelligence of human cognition.

Recent Released: BloombergGPT: Enhancing Financial NLP Capabilities with AI

Enter Unlikely AI, a London-based startup aiming to challenge assumptions about what AI can and cannot do. Founded by AI pioneer William Tunstall-Pedoe, Unlikely AI has a bold vision to develop AI that can master a broad range of capabilities previously thought exceedingly difficult or even impossible for machines.

In this article, we’ll explore Unlikely AI’s origins, mission, technology approach, potential applications, funding and growth. We’ll also look at some frequently asked questions about this intriguing new company seeking to push the boundaries of artificial intelligence.

Origins and Mission

Unlikely AI was founded in 2017 by William Tunstall-Pedoe, an experienced AI researcher and entrepreneur. Tunstall-Pedoe previously created Evi, an AI-powered virtual assistant acquired by Amazon in 2012 that would go on to help form the basis of Amazon Alexa.

With Unlikely AI, Tunstall-Pedoe aims to tackle more expansive AI capabilities. The company’s name reflects its mission to achieve artificial general intelligence by tackling AI challenges commonly deemed “unlikely” to be solved anytime soon.

On its website, Unlikely AI states its goal is to “create a world where highly intelligent automated systems enable humanity to flourish and benefit us all.” The startup aims to build AI that integrates broad capabilities, context and common sense – allowing systems to learn complex new skills rapidly, communicate at a human level and dynamically adapt to new environments.

Funding and Investors

To pursue its ambitious mission, Unlikely AI has raised $22.1 million in funding over two rounds.

In April 2020, the company announced $13.1 million in seed funding led by 8VC and Amadeus Capital Partners. First Round Capital, Homebrew, #Angels and Update Capital also participated.

In August 2022, Unlikely AI raised a $9 million extension to its seed round. New investors included Breyer Capital, UCL Technology Fund and AME Cloud Ventures. Existing investors 8VC, Amadeus Capital Partners and First Round Capital increased their investments as well.

This funding will help Unlikely AI recruit high-level AI talent and accelerate its technical work on artificial general intelligence. The vote of confidence from prominent venture capital firms highlights the excitement surrounding Unlikely AI’s unconventional approach.

Leadership Team

Unlikely AI was founded by William Tunstall-Pedoe, who serves as the company’s CEO. Tunstall-Pedoe studied Computer Science at the University of Cambridge. After graduating, he founded two startups focused on AI and semantic technology before launching Evi Technologies in 2007.

Tunstall-Pedoe has assembled an accomplished team of AI experts at Unlikely AI:

  • Serkan Piantino, CTO – Former Director of Facebook AI Research
  • Shubho Sengupta, Chief Scientist – Former research scientist at DeepMind
  • Kendall Lowrey, PhD, VP of Engineering – Former Lead Software Engineer at Illumina

This experienced founding team combines deep AI research expertise with a successful track record of building companies and shipping products.

Unique Approach to Artificial General Intelligence

So how does Unlikely AI plan to achieve its goal of developing unlikely AI capabilities? The company is taking a distinctive approach to pursuing artificial general intelligence.

Most AI today utilizes deep neural networks that excel at pattern recognition in narrowly defined domains like image classification or speech recognition. However, these systems struggle with complex reasoning, planning and knowledge integration across different disciplines.

Unlikely AI aims to develop AI that has a deeper understanding of concepts, causes, contexts and how the world works. The company is focusing on integration across vision, language, robotics and planning to achieve more human-like general intelligence.

Unlikely AI believes mastering unlikely capabilities like common sense reasoning, social intelligence and generalist learning will pave the way towards broader AI progress. The company is exploring innovative techniques like causal models, compositional generalization and transfer learning to make progress on these difficult challenges.

Potential Applications

Unlikely AI’s artificial general intelligence approach could enable a wide range of applications that enhance human capabilities and empower new technology:

  • Personal assistants – AI assistants that can carry on free-flowing, informative conversations on any topic and complete sophisticated tasks.
  • Medical diagnosis – AI that can integrate disjointed patient data into unified models for more accurate diagnosis and treatment.
  • Scientific discovery – AI research partners that can connect insights across scientific disciplines to accelerate discoveries.
  • Education – AI tutors that can adaptively teach students of all ages across diverse subjects.
  • Robotics – Robots able to safely operate in unpredictable real-world environments and learn new skills.
  • Creative arts – AI capable of general creativity and originality across media like writing, visual art, and music.

Unlikely AI aims to make these once-improbable applications a reality by understanding complexity, uncertainty and ambiguity as well as humans. Its generalist AI approach could enable systems to learn how to do almost anything.

Company Growth

Since its launch in 2017, Unlikely AI has seen rapid growth. The company now has over 50 employees and continues hiring leading AI researchers, engineers, and operations personnel.

Unlikely AI is headquartered in London, with additional offices in Cambridge, England and Berkeley, California. In 2021, the startup opened its first US office in the San Francisco Bay Area to tap into the region’s deep talent pool in artificial intelligence.

The company has a distributed workforce across Europe, allowing it to recruit staff unconstrained by physical location. Unlikely AI aims to triple its team size in the coming year to accelerate progress.

While still early-stage, Unlikely AI is quickly building up its capabilities and resources. The startup will continue growing its team and refining its technology to achieve unlikely breakthroughs in AI.

Current Open Positions

Unlikely AI is actively recruiting to grow its team and is looking for exceptional talent across engineering, research science, design, operations and more.

Some of the positions currently open include:

  • Senior Machine Learning Researcher
  • Senior Backend Engineer
  • Senior Full-stack Engineer
  • Software Engineer, Platform
  • Product Designer
  • Recruiting Coordinator
  • Executive Assistant to the CEO

The company provides competitive compensation including salary, equity and benefits. Unlikely AI also touts its engaging culture focused on creativity, wellness and personal growth.

Those interested in working on cutting-edge AI are encouraged to browse Unlikely AI’s open positions and apply to help realize the company’s ambitious vision.


Who is the founder of Unlikely AI?

Unlikely AI was founded by William Tunstall-Pedoe, an experienced British AI entrepreneur and researcher. He previously founded the AI company Evi Technologies which was acquired by Amazon in 2012.

How much funding has Unlikely AI raised?

Unlikely AI has raised $22.1 million in seed funding to date over two rounds. Investors include 8VC, Amadeus Capital Partners, First Round Capital and Breyer Capital.

What is Unlikely AI’s technical approach?

Unlikely AI aims to achieve artificial general intelligence by focusing on broader capabilities like reasoning, conceptual understanding, and generalist learning. The company believes mastering “unlikely” AI challenges will unlock more expansive progress.

What AI capabilities is Unlikely AI working on?

Unlikely AI is working towards AI that exhibits common sense reasoning, social intelligence, generalist learning, creative thinking, and the ability to dynamically adapt to new environments.

Where is Unlikely AI located?

Unlikely AI is headquartered in London, England and has additional offices in Cambridge, England and Berkeley, California. It has a distributed workforce across Europe.

Is Unlikely AI hiring?

Yes, Unlikely AI is rapidly hiring across engineering, research, design and operations positions. Interested candidates can view and apply for open roles on the company’s careers page.


In summary, Unlikely AI aspires to push artificial intelligence into uncharted territory by tackling capabilities previously thought improbable or impossible. The company is taking an unconventional approach focused on integrative general intelligence rather than narrow, siloed tasks.

Backed by strong funding and an experienced founding team, Unlikely AI aims to achieve unlikely breakthroughs by combining reasoning, imagination and adaptability. The startup’s bold vision could unlock AI that helps humanity flourish through broad applications enhancing science, creativity, productivity and more.

While significant challenges remain, Unlikely AI represents an exciting path forward for artificial intelligence research. The company’s out-of-the-box thinking and technical rigor have the potential to profoundly expand what AI can achieve. By hiring top talent across disciplines, Unlikely AI is well-positioned to turn audacious goals into reality.

The coming years promise to be an intriguing ride as Unlikely AI works to fulfill its mission of developing AI that does the unimaginable. Will the company achieve artificial general intelligence and transform technology through unlikely advances? The AI community watches eagerly as Unlikely AI strides into uncharted territory.

Table 1: Key Facts about Unlikely AI

Founding Year2017
FounderWilliam Tunstall-Pedoe
HeadquartersLondon, England
Funding Raised$22.1 million
Investors8VC, Amadeus Capital Partners, First Round Capital, Breyer Capital
Key TechnologiesCausal models, compositional generalization, transfer learning
Focus AreasCommon sense reasoning, generalist learning, social intelligence
EmployeesOver 50
Jobs AvailableEngineering, Research Science, Design, Operations, and more

Leave a Comment

%d bloggers like this: