Photo: Maral Fard
Ariadna Font i Llitjós (Barcelona, 1974) is the current co-founder and CEO of Alinia AI. She graduated in Translation and Interpretation from Pompeu Fabra University and holds a PhD in Computational Linguistics from Carnegie Mellon University (Pittsburgh, USA), and has more than 20 years of experience in product development and managing multidisciplinary teams. She has worked at leading global technology companies, where she was the driving force behind Twitter’s algorithmic ethics area, as well as the engineering director of the machine learning platform. She has been heavily involved in diversity and inclusion (as a board member of Twitter Women and an active member of Twitter Alas, Twitter’s LatinX group). He has been living and working in the United States for more than 24 years. At Comunicació 21, Font emphasizes AI learning as a tool to improve the quality of work.
She trained as a translator and ended up becoming an authority in the technology sector in the United States. What motivated you to redirect your career?
Dr. Tony Badia gave me an appreciation for computational linguistics, whose mission is to train computers to design translation rules; That is, everything that includes the rules of machine translation, natural language processing, or artificial intelligence. I became so passionate about it that I ended up getting a master’s degree and a doctorate in the field.
Almost his first job was as a computational linguist.
After obtaining my doctorate, I worked in one of them spin off Which was based on creating an internal corporate search engine very similar to Google. At the time – and I’m talking about 20 years ago – that was a very innovative bet. Although I initially worked as a computational language expert, I touched on other keys that helped me learn and learn, especially regarding user experience. Training was essential at that time.
The technology sector is very masculine, especially in decision-making areas, above all, in the United States. As a woman, how have you been able to position yourself in management positions?
Until I got into the technology sector, I never thought about it, because during my studies we were mostly girls. Once I arrived in the United States, I noticed this. Women’s rise to the top in the world of work in the United States is extremely complex. As for leadership positions, you have to excel a lot, and you have to win them twice. It requires energy, dedication and, above all, laying the foundation for future generations of women.
In the United States, working in leadership positions is closely linked to the syndrome Workaholic. Did you suffer from it?
Yes, of course. I’ll tell you just one thing: My New Year’s wish is to maintain a balance between my professional and personal life. In the United States, there is a work culture that forces you to be constantly connected, and it is part of the DNA. The work never ends. So, if you don’t set boundaries for yourself, you’ll end up exhausted. I set boundaries, and that has helped me raise two daughters. If I was alone, I would work all day, including weekends. My daughters forced me to stop working and devote myself to being in the here and now. It has been the best medicine for me.
“X doesn’t have the mission that Twitter had. It has lost popularity, attention and, above all, the relationship among its users.
In 2019 he signed on Twitter (currently X). What task do they do?
Platform management Machine learning, which was composed of a team of one hundred professionals. And it was. Now, I realize that Twitter as a company was very small if we compare it to Facebook, Google or Microsoft, and instead, it had a very relevant social impact on a global level. For this I needed a specific field of algorithmic ethics.
because?
Because with the global impact that Twitter had at the time, it was necessary to have an internal team that took care of issues related to algorithmic fairness and biases. It was essential to be able to train the algorithms responsibly. So I gave them a suggestion: Create an algorithmic ethics area on Twitter built into the platform Machine learning. And they accepted that.
Photo: Maral Fard
What has the field of algorithmic ethics brought to Twitter as you have led it?
Since launching the field of ethics, we have set goals and metrics that we continually regulate. It was a company-wide job, because I brought everyone on board with a very comprehensive vision. Only then will it have an impact inside and outside the company.
However, his time at Twitter ends with the arrival of Elon Musk as brand owner.
Yes, this is correct. After months of uncertainty that we learned about through the press, the day came when we could not connect to a business conversation. I was never able to meet Elon Musk in person because I was working remotely from New York and Twitter’s offices are located in San Francisco. Most of my team went abroad. After months of hard work, it was all for nothing. What is more dangerous: Musk has removed all kinds of controls, and the platform has become unsustainable, with abuse, insults and insulting comments.
Which path do you expect X to take?
I don’t know, but X clearly doesn’t have the mission that Twitter had. Despite its flaws, Twitter was intended to be an open space for everyone to express their opinions respectfully and politely. The previous philosophy was a space for public service. With Elon Musk, the user became charged, and Twitter lost popularity, interest and, above all, the relationship between its users.
“ChatGPT has democratized AI”
Are you still a user?
I lost the habit of calling him daily. Currently, I can reach them when they send me a direct message, for the most part.
Are Twitter trolls inevitable?
Yes, because there are people who just want to say bad things. However, everything has limits. Setting boundaries was part of our job. One of the goals we had was to reduce the number of trolls. Now with Musk there are so many other things that he not only avoids, but seems to want to cultivate trolls.
In the United States, is there a platform being developed that could have the same function and influence as Twitter?
There are many projects trying to launch to fill the void that Twitter might leave if it loses momentum. However, none of these projects have taken hold yet.
Photo: Maral Fard
Why is the algorithm so powerful?
Because everything we see on a digital screen is guided and selected by an algorithm. Language models, for example, are algorithms. And they are powerful because, although we don’t always realize it, they powerfully influence many of the things we do every day: from choosing which doctor, restaurant, or movie to watch on the weekend; These algorithms are called “recommendation”. Those who think they control their destiny online don’t know how apps and web pages work behind the scenes. And let’s not even mention the users of apps like TikTok!
Can you explain what or what a recommendation algorithm is? Machine learning?
An algorithm based on neural networks is a type of algorithm Machine learning Inspired by the structure and functioning of the human brain. Artificial neural networks consist of artificial neurons organized in layers that process input data, perform calculations, and produce output. Specifically, an algorithm is a set of mathematical formulas that are used to train a model Machine learningWhich can be either neural networks or of another type. During training, the model is exposed to a very large set of data, and the result is weights that are applied to the different neural layers. Once the model is trained, once it receives new data, there is an inference process that calculates what is whatproduction,based on the learned weights. Algorithm and data are two sides of the same coin, and it is difficult to separate them. For a long time, we have lived between algorithms and data. We are in a state of constant determination, even if we think this is not the case.
She is currently the co-founder and CEO of Alinia AI. What is this based on? start?
that it start Making generative AI more accessible and secure for businesses. At Alinia, we work to ensure generative AI systems align with company values and policies, expected behaviors, and even regulatory requirements. Personally, I’m interested in the intersection between research, product, design, AI, ethics, and society, and I work to ensure that technology creators understand not only the business context of their products and services, but also the social impact and ethical implications of doing so. New technologies.
“Companies have seen the potential it has and one of their goals is how to use these technologies to improve productivity or customer service.”
Why should companies invest within a framework of ethical governance?
As a result of the boom in generative AI, companies have realized its potential and one of its goals has been how to use these technologies to improve productivity or customer service. However, it is essential that companies do this in a responsible and reliable manner.
With AI, are there more risks or more opportunities?
Both There are as many risks as there are opportunities. AI is a tool, and if you know how to use it well and there are good intentions behind it, it clearly has more opportunities than risks. This does not mean that there are no risks. For this reason, it is a priority to set boundaries with AI from a business, ethical and legal point of view.
Who sets the boundaries for artificial intelligence?
We as a society have to decide what we want our relationship with artificial intelligence to be. We should not leave it to chance, or let it be decided by the richest people on the planet or the four most powerful and influential governments in the world. Therefore, it is necessary to understand the technology. If we do this, the conversation will be more productive, and we can decide how we interact and relate to this technology.
Photo: Maral Fard
Is there opposition or disdain because artificial intelligence generates a response to ignorance for some?
Ignorance breeds uncertainty and even fear. Artificial intelligence has risks and dangers, but it is the people behind the technology who decide its purpose and how to use it. Anyone who does not want to interact with artificial intelligence is making a mistake, because if he does not understand it, he will not benefit from it. For this reason, those of us who are part of today’s society must make the effort to understand what AI can bring and determine the paradigms we should delegate to this technology. There are things that humans don’t want to do, and this technology – artificial intelligence – works very well.
that?
Repetitive jobs that tire us but do not enrich us. For example, processing thousands and thousands of articles to finish preparing an abstract, can give us ideas for a book title… In this sense, ChatGPT has contributed to the democratization of artificial intelligence. It’s a mistake to think that ChatGPT has an answer for everything; Instead, it can help you address questions and problems. In this regard, I encourage everyone to use it.
With ChatGPT, is journalism and the role of journalists at risk?
It’s a question almost every journalist asks me, but it should be asked by everyone. Do we want a society without journalists? If you ask me, I’m clear: impossible! AI will not replace journalists, but it is a technology that will help journalists improve the quality of their work and, above all, do it much faster. Therefore, journalists who do not use AI will be at a disadvantage.
85
“Infuriatingly humble social media buff. Twitter advocate. Writer. Internet nerd.”