Knowledge › Area 1 · Understanding AI and data protection

Our company · The faces of AI

The people who build AI.

Who's actually behind the AI everyone's talking about right now? An ordered overview - from the pioneers of the 1950s to today's lab heads, safety researchers and critics.

Why this article

Artificial intelligence didn't come out of nowhere. It's the work of about a hundred people over seventy years - mathematicians, computer scientists, psychologists, philosophers. Whoever knows their names understands the field better.

This page isn't an encyclopedia. It's a map of the most important figures in four groups: the trailblazers, today's lab heads, the AI safety thinkers and the critical voices. Each with a short bio and what they stand for. So the next time you read one of these names, you can judge whether their voice fits your topic at all.

First group

The trailblazers.

Without them there would be no AI. Mathematicians, computer scientists, psychologists who laid the foundations between 1950 and 2010 - often decades before they became practical.

Theorist · Father of the idea

Alan Turing

1912 - 1954 · British

In 1950 he posed the question that drives the whole field to this day: "Can machines think?" His paper "Computing Machinery and Intelligence" defines the Turing test - a procedure to check whether a machine can hold a conversation so well that a human can't tell it apart from a person.

Cracked the German Enigma cipher in the Second World War, is regarded as the inventor of the modern computer. Died at 41 after state persecution for his homosexuality. Posthumously pardoned by the British Queen in 2013.

Coining the term · Dartmouth conference

John McCarthy

1927 - 2011 · American

In 1956 he coined the term "artificial intelligence" - at the legendary Dartmouth conference, regarded as the birth of the field. Developed the programming language LISP, which was the AI standard for four decades.

Stanford professor, Turing Award 1971. Was among the first to think about machine logic, knowledge representation and common-sense reasoning.

Sceptic · Symbolic AI

Marvin Minsky

1927 - 2016 · American

MIT professor, co-founder of the MIT AI Lab. For decades he was seen as the godfather of symbolic AI - the idea that intelligence consists of rules and symbols. In 1969 his book "Perceptrons" brought research into artificial neural networks to a standstill for 20 years.

His book "The Society of Mind" (1986) is still one of the most influential works on the question of how intelligence can arise from dumb individual parts.

First neural network

Frank Rosenblatt

1928 - 1971 · American

In 1958 he invented the perceptron - the first neural network capable of learning. The blueprint for everything that runs today under deep learning. He was publicly discredited by Minsky in the late 1960s, after which research funding dried up ("AI winter").

Only in the 1980s was his approach rehabilitated. Today his core idea is in every ChatGPT.

Turing Award 2018 · "Godfather of AI"

Geoffrey Hinton

b. 1947 · British-Canadian

In the 1980s he was one of the few who kept researching neural networks when hardly anyone believed in them. With the backpropagation algorithm he popularised the method every AI model learns with today. In 2012 his team won the ImageNet competition so decisively that it triggered the deep learning revolution.

In 2023 he quit his job at Google so he could publicly warn about AI risks. Turing Award 2018, Nobel Prize in Physics 2024.

Turing Award 2018 · Convolutional networks

Yann LeCun

b. 1960 · French

In the 1980s he developed convolutional neural networks (CNNs) - the method behind all image recognition today. First practical application: reading cheques at US banks. Today chief AI scientist at Meta (Facebook).

Unlike Hinton, a clear optimist; he sees no existential risks. He considers large language models a dead end - real intelligence needs world models, not more data.

Turing Award 2018 · Deep learning

Yoshua Bengio

b. 1964 · Canadian

Together with Hinton and LeCun he triggered the deep learning wave. Scientific director of the Mila institute in Montréal, one of the most cited computer scientists in the world.

Since 2023 he has joined the safety camp. He now leads the International AI Safety Report for the UN - the AI counterpart to the IPCC climate report.

Causality · Probability

Judea Pearl

b. 1936 · Israeli-American

In the 1980s he developed Bayesian networks - the mathematical foundation for machine reasoning under uncertainty. Turing Award 2011.

His book "The Book of Why" (2018) argues that current AI isn't yet intelligent precisely because it doesn't understand causality - only correlation. One of the most important books for grasping the limits of today's AI.

Second group

Today's lab heads.

The people who decide, in the world's big AI labs, which direction the research takes. Their tweets move share prices, their essays shape the public debate.

CEO OpenAI · ChatGPT

Sam Altman

b. 1985 · American

Co-founded OpenAI in 2015 (initially with Elon Musk, Reid Hoffman, Ilya Sutskever). In 2019 he restructured the company from a non-profit into a hybrid model in order to bring Microsoft on board as an investor.

Previously president of the startup incubator Y Combinator. Not primarily a scientist but a strategist - his talent is turning big vision into capital and attention. In November 2023 he was briefly fired by the board and brought back after five days - still one of the strangest tech dramas of the decade.

CEO Anthropic · Claude

Dario Amodei

b. 1983 · American

A physicist; in 2021 he left OpenAI to found Anthropic - together with his sister Daniela. The reason for the break: concerns that OpenAI was sacrificing safety for speed. Anthropic positions itself as the "safety-oriented" AI lab.

His essay "Machines of Loving Grace" (2024) is the most detailed positive vision of the future any AI head has ever published. Long a backstage figure, he has been publicly present since 2024.

CEO Google DeepMind · Nobel Prize

Demis Hassabis

b. 1976 · British

Chess prodigy, game designer, neuroscientist. Founded DeepMind in 2010, sold it to Google in 2014. His lab delivered AlphaGo (victory over the Go world champion), AlphaFold (solving the protein-folding problem) and many other breakthroughs.

Nobel Prize in Chemistry 2024 for AlphaFold. Regarded as the "most scientific" of the lab heads - his explicit goal is to use AI to solve science itself, not to build chatbots.

Co-founder OpenAI · returned 2025

Ilya Sutskever

b. 1986 · Russian-Israeli-Canadian

A Hinton student, co-founder of OpenAI, for years its chief scientist. Regarded as one of the most influential minds behind ChatGPT. In November 2023 he was central to the brief ousting of Altman - over safety concerns.

He left OpenAI in 2024 and founded Safe Superintelligence Inc. (SSI) - a lab with a single goal: to build safe superintelligence, with no commercial distractions.

Teacher · Explainer · Tesla alumnus

Andrej Karpathy

b. 1986 · Slovak-Canadian

One of the founding members of OpenAI, later head of Tesla's Autopilot development, back at OpenAI in 2023, gone again in 2024. He currently works independently in education - his YouTube channel with explainer videos on language models is the most-used learning resource for AI engineers worldwide.

In 2025 he founded the learning platform Eureka Labs. He regularly coins terms the field then adopts - for example "Software 2.0" for AI-based programs.

Ex-CTO OpenAI · Thinking Machines

Mira Murati

b. 1988 · Albanian-American

Was CTO of OpenAI throughout the entire ChatGPT era. In November 2023 she served for a few hours as interim CEO when Altman was removed. She left OpenAI in 2024.

In 2025 she founded Thinking Machines Lab - with almost 800 million dollars of seed funding, the largest first round in the industry's history. Focus: AI that works together with people rather than replacing them.

CEO Microsoft AI · DeepMind alumnus

Mustafa Suleyman

b. 1984 · British

Co-founder of DeepMind (2010), later co-founder of Inflection AI (the chatbot "Pi"). In 2024 he and his team were "poached" by Microsoft - Inflection effectively dissolved. He now leads Microsoft's AI division.

His book "The Coming Wave" (2023) is one of the most important books for strategically situating the next ten years - arguing that AI and synthetic biology together will shift all existing balances of power.

CEO xAI · Grok · Tesla · X

Elon Musk

b. 1971 · South African-American

Co-founded OpenAI (2015) and left in 2018 after a falling-out. In 2023 he founded xAI and builds the chatbot Grok, which is integrated into X (formerly Twitter). Since 2024 he has been fighting OpenAI in several court cases.

His position is contradictory: in 2018 he was still a loud warner against AI risks, today he builds his own models as fast as possible. He acts on the field mainly as a political and economic amplifier, less as a scientist.

Visionary · Singularity

Ray Kurzweil

b. 1948 · American

Inventor, futurist, Director of Engineering at Google since 2012. In 2005, in "The Singularity Is Near", he predicted that AI would reach human intelligence by 2029 and far surpass it by 2045 (the singularity).

Long mocked as an outsider, since 2023 many regard him as a seer. In 2024 his update book "The Singularity Is Nearer". He sticks to his original forecasts unchanged.

DeepSeek · China

Liang Wenfeng

b. 1985 · Chinese

Founder and CEO of DeepSeek - the Chinese lab that made headlines worldwide in January 2025 with the R1 model. R1 was reportedly trained for a fraction of the cost of Western models and challenged the assumption that only hyperscalers can build AI.

Earlier career in the quant hedge fund world. He stands for the non-Western side of AI development, which the West long underestimated.

Third group

The safety researchers.

A distinct group of thinkers isn't concerned with making AI better, but with making it safe. Their tools are philosophy, mathematics and risk theory.

Berkeley · Design principle

Stuart Russell

b. 1962 · British

Berkeley professor, author of the standard textbook "Artificial Intelligence: A Modern Approach" - used in 1,500 universities worldwide.

His book "Human Compatible" (2019) proposes building AI from the ground up anew: machines should not pursue their own goals, but should seek to fulfil human preferences under uncertainty. One of the few to sketch a constructive way out of the alignment problem.

Oxford · Existential risk

Nick Bostrom

b. 1973 · Swedish

Philosopher, founder of the Future of Humanity Institute (Oxford, 2005 to 2024). His book "Superintelligence" (2014) was the first to bring the risk discourse into the mainstream - with readers such as Bill Gates and Elon Musk.

In 2024, with "Deep Utopia", he switched sides: what is meaningful in a world where AI solves every problem? A question hardly anyone asks today - but one that could soon become important.

MIRI · Doomer

Eliezer Yudkowsky

b. 1979 · American

Self-taught, founder of the Machine Intelligence Research Institute (MIRI). From 2000 he was among the first to warn about AI risks, long mocked as an outsider.

In 2025, with Nate Soares, he published "If Anyone Builds It, Everyone Dies". He holds the most extreme position: current AI development will, with high probability, lead to the extinction of humanity. His proposal: an international halt to all large AI training runs.

Oxford · Longtermism

Toby Ord

b. 1979 · Australian

Philosopher at Oxford, adviser to governments and the WHO. His book "The Precipice" (2020) is the most sober quantitative analysis of existential risks to humanity.

He estimates the overall risk of a catastrophe this century at one in six - of which one in ten is from misaligned AI alone. More than all other risks (nuclear war, climate change, pandemics) combined.

Louisville · Control theory

Roman Yampolskiy

b. 1979 · American

Computer scientist, University of Louisville. He argues that controlling a superintelligence is fundamentally impossible - comparable to a perpetual motion machine. His book "AI: Unexplainable, Unpredictable, Uncontrollable" (2024).

He estimates the risk of an AI catastrophe at 99.9 percent over the next 100 years. The most extreme publicly held position - but taken seriously in academic discussion.

Center for Humane Technology

Tristan Harris

b. 1984 · American

Ex-Google ethicist; in 2013 he coined the term "Time Well Spent" - the critique of social media's attention design. In 2018 he founded the Center for Humane Technology.

Since 2023 he has focused mainly on AI. His talk "The AI Dilemma" (2023; Netflix documentary "The Social Dilemma") brought the safety topic to a mass audience.

Fourth group

The critical voices.

A fourth group is neither builder nor safety researcher, but observer. Social scientists, philosophers, linguists who analyse the effects on society, language and power.

Historian · Sapiens, Nexus

Yuval Harari

b. 1976 · Israeli

Historian at the Hebrew University of Jerusalem. World-famous for "Sapiens" (2011). His new book "Nexus" (2024) is about the history of human information networks - from the Stone-Age rumour to AI.

His central thesis: AI is the first "alien intelligence" that hacks the operating system of civilisation - not through extinction, but through erosion of the shared space of reality.

Harvard · Surveillance capitalism

Shoshana Zuboff

b. 1951 · American

Social scientist, Harvard. Her book "The Age of Surveillance Capitalism" (2019) was the first to systematically describe the business model of Google and Facebook.

She argues that AI isn't something new but the continuation of the same model with better tools. Training on copyrighted data is simply theft, and "innovation" a sedative against regulation.

Linguist · "stochastic parrots"

Emily Bender

b. 1973 · American

Computational linguist at the University of Washington. Co-author of the most influential critical paper on language models: "On the Dangers of Stochastic Parrots" (2021) - which cost Google employees like Timnit Gebru their jobs when it was published.

Her thesis: language models are not intelligent beings but stochastic parrots - they predict probable word sequences without understanding anything. The anthropomorphisation is the real problem.

Algorithmic Justice League

Joy Buolamwini

b. 1989 · Canadian-American

Computer scientist and artist, MIT doctorate. In 2018, in the study "Gender Shades", she demonstrated that facial recognition AIs make drastically more errors with dark-skinned women than with light-skinned men.

With this she sparked the debate about algorithmic discrimination. Founder of the Algorithmic Justice League, book "Unmasking AI" (2023).

Sceptic · Critic of the pioneers

Gary Marcus

b. 1970 · American

Psychologist and AI entrepreneur, formerly an NYU professor. Regarded as the loudest sceptic of large language models - in his Substack Marcus on AI he documents daily where current models fail.

His thesis: language models will never reach real intelligence, because they lack world models, causality and common-sense reasoning. His demand: hybrid AI - symbolic logic and neural networks combined.

Mathematician · Weapons of Math Destruction

Cathy O'Neil

b. 1972 · American

Mathematician, ex-hedge-fund quant. In 2016, with "Weapons of Math Destruction", she started the debate about algorithmic harm with the wider public - credit scoring, applicant selection, policing algorithms.

Her thesis: algorithms are not objective. They encode the biases of their builders and scale them - and it hits the weakest members of society the hardest.

What these four groups have in common

AI isn't built by one company. It's negotiated by a small, interconnected world of people - who often know one another, have often worked against one another, often had the same doctoral supervisor.

Whoever reads a name from this list in a headline should keep two things in mind: Which group are they in? And: What self-interest do they have in this statement? Sam Altman warning about AI risks isn't the same as Eliezer Yudkowsky saying the same thing - even though the sentences sound alike. Whoever knows the field hears the differences.

That's our job as wendwerk: to hear and situate these voices for you - so you don't have to subscribe to every newsletter and read every book yourself to know where the next sensible step for your business lies.

If you want to go deeper

Knowing the faces is one thing. Knowing what they mean for your business is another.

How these voices line up into larger camps - read Where we're heading. If you feel overwhelmed by the variety, read AI paralysis - why you don't get started. If you want to know what AI can already do today, read What AI makes possible.

Let's talk ← Back to the knowledge base