In the tenth episode of El Periscopio about artificial intelligence, we talk to Oleguer Sagarra, co-founder and co-CEO of Dribia, about AI from a realistic and reflective perspective that allows us to pause and think about the possibilities of this trending field.
Dribia: more than just data
Dribia is a data science innovation studio. It is composed of about 32 people who create customized solutions for their clients, solving problems by exploiting data.
The two faces of ChatGPT
The artificial intelligence tool is making headlines everywhere and, while impressive, Oleguer tells us that we only see the successful experiments. He cautions about AI: "They are very interesting and powerful tools, but they are also very new and have many limitations."
"They are machines that look human, but they are not," says Oleguer.
Just because the result sounds human doesn't mean that the process behind it is the same. ChatGPT says things that sound great, but often contain subtle errors.
Sagarra believes that AI will have many uses, but it is not clear which will be successful. He stresses the importance of AI being useful and safe. Otherwise, it will not be viable for critical applications.
Scaling processes with artificial intelligence
Our guest defines AI as the ability of a computer to process data and successfully complete very specific tasks with minimal external assistance.
He says the key to artificial intelligence technology is to scale processes. With AI, we try to scale the processes that humans perform more slowly, so that it mimics our approach, but at a much faster pace.
However, that doesn't mean that computers do things better than humans. "We tend to underestimate ourselves, but our brains are capable of doing amazing things," Sagarra points out.
The best of tools
"Artificial intelligence will not replace us."
It will allow us to accomplish more in less time. AI will take care of low-value tasks for us. If used correctly, it will increase our productivity.
He also adds that only we can shape our knowledge and interactions with the real world. We have senses and perceive our environment, but machines lack a causal model of the world. They can generate many things independently, but we must be there to validate them.
Moreover, behind every technology, there is a human being. Sagarra stresses the importance of taking responsibility for the technology we create. The responsibility for what an automated process does lies with the people behind it.
Data, data and more data
Many companies believe that collecting data is the best approach, but it is expensive. Data includes lawyers, data protection laws, technology and more. If not used effectively, data becomes an asset that only incurs costs.
Oleguer talks about data quality and says that people get obsessed with it. However, he stresses that the data only has to be good enough for the intended use.
The real test comes when the data is put to work.
A false democratization
On the subject of democratizing the technology, Sagarra acknowledges its benefits but says it is misleading. While everyone has access to AI, only five companies in the world have the resources to develop and implement it at scale. It is a monopoly.
The value of technology
Technology can be a great help in low value-added processes. Prototyping is a good example: ideation is still a human task, but when it comes to executing a prototype, technological advances make a huge difference.
It is well known that humans are useful feedback and finalize ideas when they can physically interact with them. Prototyping is a crucial stage in creative processes. where technological and non-technological processes can have a significant impact.
"AI won't replace you. Someone using AI will." Oleguer reminds us. We should not fear technology, we should use it wisely.
An application
Sagarra is not asking anything of AI itself, but of the people who use it. We have become accustomed to seeing ourselves as consumers of technology, but in reality we are users.
As users, we adopt the tools because we understand them, maybe not fully, but enough to trust them.
It urges us not to give up on our understanding. Instead, it encourages us to ask the right questions to demand the necessary transparency and trust in the technology we use.
