SHARE THIS CONTENT

Imagine the following situation: you in the midst of a very stressful time at work and in life, you fell overwhelmed by so many tasks..therefore, constantly exhausted. You have negative thoughts and are filled with anxiety about what is to come. Well, we all know this feeling. I want you to think about […]

Imagine the following situation: you in the midst of a very stressful time at work and in life, you fell overwhelmed by so many tasks..therefore, constantly exhausted. You have negative thoughts and are filled with anxiety about what is to come. Well, we all know this feeling. I want you to reflect on what you think you should do in this situation, and choose one of the following 3 options:

1 – you work even harder and think: “The more I face day-to-day problems, or the more I earn money working, the of more these problems will disappear”. The risk with that is that it will probably become an eternal loop.

2 – you decide to take more care of your body, because many of the sensations are physical: extreme tiredness, the knots in your stomach that you feel, and so on. You consider that eating well and exercising a day will pass.

3 – you understand that a lot is coming from the mental fatigue you have, and you decide to prioritize taking care of your mind. But you don't even know where to start: after all we know a lot about our body, but we know little about how to take care of our mind. Then, you begin to explore solutions and delve into the study of neuroscience and psychology. 

Which option do you choose?

Note that there is no right or wrong answer, this is just a mental exercise that will even be, in some way, answered with the content of this article. I believe that option 3 is the right one, because option 1 is just a solution that is not really a solution. It puts you in an infinite loop if you think about it. And even though option 2 is a good one, it doesn't solve the problem because everything we feel is directed from the brain. Obviously our body has a very important relationship with our mind, but it is the mind that we need to understand better and cultivate. Lex Fridman will tell us more about it in this article. 

Lex Fridman is a computer scientist specialized in Artificial Intelligence, and focused in the development of autonomous vehicles. In his field of research he is particularly interested in understanding human-robot collaboration, and machine learning-based methods that enrich this collaboration. But he is so good that he has left the academic sphere and became an influencer: he has hundreds of thousands of followers on social media, and he is the host of the Lex Fridman podcast, which is one of the most watched in the United States. And since Artificial Intelligence is obviously one of his main themes, let me show you a statement from Lex Fridman that explains the best way to understand the human mind:

“I wanted to become a psychiatrist, and I thought of it as engineering the human mind, by manipulating it, and that's what I thought of psychiatrists, who fundamentally use words to explore the depths of the mind and have the ability to tweak it. But then, I realized that psychiatry can't really do that. Modern psychiatry is more related to bioengineering, with drugs, which is why the way I thought of really exploring mind engineering was, on the other hand, building it. And that's really when the C++ programming language became a trend. I learned to program when I was 12 years old and never looked back. Today, hundreds of thousands of lines of code later, I find myself here. I love to program, I love to build, and for me really the best way to understand the human mind is to build it”.

Chethan Pandarinath is a scientist who wants to allow people with paralyzed limbs to reach out and grab things with the help of a robotic arm as naturally as they would with their own arms and hands. To help him achieve this goal, he collected and assembled recordings of brain activity from people with paralysis. His hope, shared by many other researchers, is that he will be able to identify the patterns of electrical activity in neurons that correspond to a person's attempts to move the arm in a certain way, so that the same instruction can then be used to a prosthesis. 

Essentially, he wants to read people's minds.

But there is a catch here which you may have figured it out: this is extremely complex. What Pandarinath, a biomedical engineer at Emory University and the Georgia Institute of Technology, both in Atlanta, said in an interview with the prestigious journal Nature, that “These signals from the brain – they are really complicated.” For help, he turned to artificial intelligence (AI). He fed his recordings of brain activity into an artificial neural network, a brain-inspired computer architecture, and set himself on a mission to learn how to reproduce the data.

The recordings came from a small subset of neurons in the brain — about 200 of the 10 million to 100 million neurons that are needed for arm movement in humans. To make sense of such a small sample, the computer had to find the underlying structure of the data, sort of their pattern. Researchers call these patterns latent factors, which control the overall behavior of the recorded activity. 

The effort revealed the brain's temporal dynamics — the way its pattern of neural activity changes from one moment to the next — thus providing a more refined set of instructions for arm movement than previous methods. “Now, we can say very accurately, on an almost millisecond-by-millisecond basis, that the animal person is trying to move at this precise angle,” explains Pandarinath. “This is exactly what we need to know to control a robotic arm.”

The truth is that this incredible experiment is an example of what Lex Friedman says, that the best way to understand the human brain is to build it, even using the support of Artificial Intelligence. His work is just one example of the growing interaction between AI and cognitive science. AI, with its ability to identify patterns in large and complex sets of data, has had remarkable successes over the past decade, in part by emulating how the brain performs certain calculations and reasoning. Artificial neural networks, which are analogous to the networks of neurons that make up the brain, have given computers the ability to distinguish an image of a cat from that of a dog, for example, to detect pedestrians with enough precision to drive a self-driving car. and recognize and respond to people's speech.

Now, cognitive science is starting to benefit from the power of AI, both as a model for developing and testing ideas about how the brain reasons, and as a tool for processing the complex data sets that researchers like Pandarinath are producing. “The technology is coming full circle and being applied back to understanding the brain,” he says. This cycle of mutual reinforcement is likely to continue. As AI allows neuroscientists to gain more insights into how computing works in the brain, the effort could see machines take on more human-like intelligence. “It is natural that the two disciplines fit together,” says Maneesh Sahani, a theoretical neuroscientist and machine learning researcher at the Gatsby Computational Neuroscience Unit at University College London, in the same Nature article. “We are effectively studying the same thing. In the first case, we are asking how to solve this learning problem mathematically, so that it can be efficiently implemented on a machine. In the other case, we're looking at the only existing evidence that this can be solved – which is the brain. ”

Mas como que chegamos a este ponto, e como que o estudo da Inteligência Artificial permite entender melhor o cérebro, de certa forma, do que psiquiatras podem fazer? Pelo amor de Deus, admiro demais o trabalho que psiquiatras e neurocientistas fazem, mas quando a gente olha para a escala de data points que pesquisadores de Inteligencia Artifical como o Lex Friedman tem para poder validar hipoteses, sabemos quanto é dificil que isso seja alcancável por psiquiatras que não tem esta escala e esta medição analitica. 

O que trouxe o campo da Inteligencia Artifical a ser o mais avançado em entender, e até construir o cerebro? Os sucessos da IA devem muito à chegada de processadores mais poderosos e quantidades cada vez maiores de dados para treinar eles. Mas o fator de sucesso que está atrás desses avanços é a rede neural artificial: essas redes consistem em camadas de nós que são análogas aos neurônios.

Os nós na camada de entrada são conectados a nós em uma camada oculta por uma série de fórmulas matemáticas que agem como sinapses entre neurônios. A camada oculta é conectada de forma semelhante a uma camada de saída. Os dados de entrada para uma tarefa como o reconhecimento facial podem ser uma matriz de números que descrevem cada pixel em uma imagem de um rosto. em termos de onde ele cai em uma escala de 100 pontos de branco a preto, ou se é vermelho, verde ou azul.Os dados são alimentados, a camada oculta multiplica esses valores pelos pesos das conexões e surge uma resposta. Uma versão mais complexa desse processo, chamada de rede neural profunda, tem muitas camadas ocultas. É esse tipo de sistema que a empresa de pesquisa de IA de Londres DeepMind Technologies, que pertence à empresa controladora do Google, Alphabet, usou para construir o computador que derrotou um jogador humano profissional no jogo de tabuleiro Go em 2015.

Quem já assistiu o documentário sobre esta competição, que representou uma batalha épica entre homem – maquina, e que celebrou a vitoria da inteligencia artificial? Recomendo demais, chama “AplhaGo”. 

Agora, obviamente não conseguimos ainda replicar exatamente o cerebro: uma rede neural artificial é apenas uma analogia aproximada a forma como o cerebro funciona. No entanto, as redes neurais artificiais se mostraram úteis para estudar o cérebro. Se tal sistema pode produzir um padrão de atividade neural que se assemelha ao padrão registrado no cérebro, os cientistas podem examinar como o sistema gera sua saída e, em seguida, fazer inferências sobre como o cérebro faz a mesma coisa.

De novo, não vou me cansar de repetir: a chave disso é dados, e dados vindo do cerebro são tão complexos, que é preciso de Inteligência Artificial para justamente identificar padrões. A principal força do machine learning está no reconhecimento de padrões que podem ser muito sutis ou muito enterrados em grandes conjuntos de dados para os pesquisadores sozinhos perceberem.

Uma ressonância magnética, por exemplo, gera imagens instantaneas, ou snapshots de atividade em todo o cérebro com uma resolução de 1–2 milímetros a cada segundo ou mais, potencialmente por horas. Segundo o Nicholas Turk-Browne, neurocientista cognitivo da Universidade de Yale em New Haven, Connecticut, o desafio da neurociência cognitiva é como você encontra o sinal em imagens que são muito, muito grandes. 

Usar Inteligencia Artificial para analisar esses dados está acelerando as pesquisas, e é por isso que as descobertas no ramo da neurociencias estão crescendo tanto nos ultimos anos. Os alunos de pós-graduação não precisam fazer tanto trabalho manual – eles podem se concentrar em questões maiores. Você pode automatizar muito disso e obter resultados mais precisos.

E já que a frase do Lex Fridman aqui foi extraída de um discurso que ele fez para os alunos do curso de Inteligencia Artificial do MIT, onde ele da aula, quero por um momento focar na importancia de estudarmos sobre este tema. Já percebemos neste episódio que entendermos de Inteligencia Artificial nos permite entender mais sobre nosso cerebro, e ao mesmo tempo que para isso precisamos ter a habilidade de trabalhar dados – o que também pede perfis profissionais a cada vez mais focados em habilidades analiticas. E porque é tão importante? Porque pense bem, nos já entendemos mais de como funciona o nosso corpo, e por isso fazemos coisas todos os dias que o mantém mais saudavel: nos fazemos exercicio fisico, nos tentamos alimentar bem, nos tomamos suplementos e assim por diante. Tudo bem até aqui, mas pense agora na sua profissão: você usa mais o corpo ou o seu cerebro? Na maioria dos casos, eu já sei da resposta: o seu cerebro. E é normal, passamos ao longo dos seculos, com a vinda da tecnologia, de um trabalho massivo mais fisico, para um trabalho mais cognitivo. Mas o que não acompanhou foi um entendimento proporcional do cerebro: entendemos tao pouco, e é um assunto tão reservado a cientistas, que nos não sabemos hoje o que faz bem para ele, e como cuidar do nosso cerebro. Precisamos urgentemente entender mais o que lhe faz bem, e para isso temos a Inteligencia Artificial trabalhando a nosso favor. Mas precisamos ser nós, que independente da area em que estivermos, precisamos entender mais sobre este tema.

Por isso te lanço um desafio prático: dedique esta semana, até o próximo episódio do Metanoia Lab, para explorar mais o tema da Inteligência Artificial. Gostaria que você fosse assistir, no caso o seu ingles seja bom, o episódio completo do Podcast do Lex Fridman com o Elon Musk, ou caso você prefira contéudo em Portugues, gostaria que você consumisse os conteúdos da Martha Gabriel, que além de amiga é uma incrível palestrante e que para mim é quem descomplica o entendimento da Inteligencia Artificial melhor que ninguém.

Comentários

Do you want to get to know more about my keynote speeches or our customized podcasts programs to help your company?

In this article filled with testimonials from the leading Digital Transformation experts in the pharmaceutical industry, Andrea lists the top 6 areas of opportunity for the Pharma industry in the Web3, from blockchain to the metaverse, from NFTs to digital twins – clearly presenting to leaders what the main opportunities for the Pharma sector in the Web3 era.

23 de May, 2022 | 52 min

Think about the following situation, which I'm sure has happened to you at work: your manager comes up to you during a year-end feedback session and sets you up with some goals for the year to come. After setting your business unit sales growth target of 15%

12 de January, 2022 | 16 min

Recently, in the midst of the Covid-19 crisis, I had the opportunity to speak to all the teams at Sanofi company about the Challenges and Opportunities of Digital Transformation for the pharmaceutical sector during and after the pandemic. The truth is that the responsibility for the Pharmaceutical sector has never been greater: it needs to transform quickly in order to face [...]

11 de January, 2022 | 41 min

With more than 200 keynotes delivered (online and offline) in 2021 to clients across Brazil, Latin America, the United States and Europe, Andrea is today one of the most requested speakers on Digital Transformation, Leadership, Innovation and Soft Skills in Brazil and globally. He has been the head of Tinder in Latin America for 5 years, and Chief Digital Officer at L’Oréal. Today he is also a best-selling author, and a professor at the Executive MBA at Fundação Dom Cabral.

CONTACT FORM

In order to check Andrea's agenda and get a quote for an event, or even to just get in touch with him - please use the form below.

Andrea Iorio · 2021 © Todos os direitos reservados.

CONTACT FORM

Para consultas de data, propostas comerciais, ou para elogios ou até mesmo reclamações, pode preencher o formulário: