493 lines
20 KiB
HTML
493 lines
20 KiB
HTML
<?xml version="1.0" encoding="UTF-8"?>
|
|
<html newsdate="2021-12-09">
|
|
<version>1</version>
|
|
|
|
<head>
|
|
<title>20 Years FSFE: Interview with Vincent Lequertier on AI</title>
|
|
</head>
|
|
|
|
<body>
|
|
|
|
<h1>20 Years FSFE: Interview with Vincent Lequertier on AI</h1>
|
|
|
|
<p>
|
|
In our sixth birthday publication we are interviewing Vincent
|
|
Lequertier about crucial aspects of artificial intelligence, such
|
|
as its transparency, its connection to Open Science, and questions
|
|
of copyright. Vincent also recommends further readings and responds
|
|
to 20 Years FSFE.
|
|
</p>
|
|
|
|
<p>
|
|
A PhD candidate at the Claude Bernard university in Lyon who
|
|
researches artificial intelligence for healthcare, Vincent supports
|
|
software freedom and volunteers for the FSFE in his free time. He
|
|
has been a part of the <a
|
|
href="https://wiki.fsfe.org/Teams/System-Hackers">System
|
|
Hackers</a>, the team responsible for the technical infrastructure
|
|
of the FSFE, for many years. His contribution was valuable in
|
|
setting the foundation for the for the good state that the FSFE's
|
|
System Hackers team is today. Vincent is also a member of the FSFE's
|
|
General Assembly, and participates in the 'Public Money? Public
|
|
Code!' campaign. In our interview, Vincent shares his thoughts
|
|
answering questions about the current state of AI and its future
|
|
implications.
|
|
</p>
|
|
|
|
<h3 id="interview-with-Nico-Rikken">Interview with Vincent Lequertier</h3>
|
|
|
|
<p>
|
|
<strong> FSFE: You are deeply involved in the field of artificial
|
|
intelligence. How would you explain to a 10-year-old what AI
|
|
is?</strong>
|
|
</p>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
<strong>Vincent Lequertier:</strong> A few years ago I was a
|
|
speaker at a local radio station, and sometimes I was responsible
|
|
for mixing the audio. At the station, there were several inputs: the
|
|
mics of the radio speakers (mine included), the music, the jingles,
|
|
and so on. And then there was the output broadcast to the radio
|
|
listeners. Between the inputs and output there was the mixing
|
|
table, with its uncountable knobs and sliders. I needed to adjust
|
|
the knobs and sliders so that the inputs were well mixed together,
|
|
thus producing an output that sounded nice to the listeners. At the
|
|
time of writing, an AI works just like that. It automatically adjusts
|
|
the numerous parameters of a digital, virtual mixing table.
|
|
Once put through it, the inputs produce a satisfying output
|
|
according to a predefined definition of success (that the sound was
|
|
nice in this analogy).
|
|
</p>
|
|
</div>
|
|
|
|
<figure>
|
|
<img src="https://pics.fsfe.org/uploads/medium/9161b78da1d86f828a793d2fbf000dfd.jpg"/>
|
|
<figcaption>A PhD candidate at the Claude Bernard university in Lyon who
|
|
researches artificial intelligence for healthcare, Vincent supports
|
|
Software Freedom and volunteers for the FSFE in his free time.</figcaption>
|
|
</figure>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>You are advocating for accessible and transparent AI.
|
|
According to your research, what would you say are the necessary
|
|
requirements to make sure that programs using artificial
|
|
intelligence are accessible and transparent?</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
Reusing AIs makes sense because they are costly to develop and
|
|
train, both in terms of human and computer resources. Additionally,
|
|
training AI models demands a lot of data which are particularly hard
|
|
to obtain and work with. Therefore, being able to reuse an AI is
|
|
important, as it saves time and potentially scarce resources.
|
|
Moreover, making an AI available to others fosters innovation by
|
|
facilitating collaboration. I think a fundamental requirement for AI
|
|
accessibility is Free Software, because AIs licensed as Free Software
|
|
(also known as Open Source) are inherently accessible. Other
|
|
requirements can be Open Standards and Open Data. AI models should
|
|
therefore be published and freely accessible.
|
|
</p>
|
|
|
|
<p>
|
|
Transparency in AI is the ability to understand and interpret the
|
|
output coming from it. Although given the complexity of today's AI
|
|
systems transparency can be hard to obtain, it is an important
|
|
characteristic as it fosters trust. <strong>Being able to understand why a
|
|
given output was produced, and what part contributed the most to it,
|
|
increases confidence in the model and makes it easier to debug.</strong>
|
|
Moreover, understanding the role played by each input can help
|
|
data-driven policy making. For example, in healthcare, understanding
|
|
the most important factors impacting the quality of patients' care
|
|
for a disease can validate or change healthcare practices. Free
|
|
Software is a key part of transparency because it allows everyone to
|
|
use the AI and analyze its predictions to better understand them.
|
|
</p>
|
|
</div>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>How can we make sure that inequalities in our current
|
|
societies do not pass on to AI data training? How can we assure
|
|
that AI results are fair?</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
As AI is really good at magnifying existing inequalities found in the
|
|
data used for its training, fairness issues will creep into AI.
|
|
Detecting those issues in the dataset and in the AI's output is
|
|
therefore critical. However, simply removing data that might be a
|
|
source for unfairness (e.g. a training dataset variable that is not
|
|
representative of the data used once the model is put in production)
|
|
may not always work, because these data might be correlated to other
|
|
attributes in the dataset which would need to be removed as well.
|
|
Completely removing any potential inequality may therefore remove a
|
|
lot of data from the dataset, potentially limiting the ability
|
|
of the AI to properly address the problem it has been designed to
|
|
solve. Inequalities therefore come from badly constructed datasets,
|
|
and advanced methods are required to circumvent them.
|
|
</p>
|
|
|
|
<blockquote>
|
|
Data related to COVID-19 are public, and <a href="https://covidtracker.fr/">the most popular website</a> to
|
|
visualize these data as well as other tools <a href="https://github.com/CovidTrackerFr">are Free Software</a>.
|
|
</blockquote>
|
|
|
|
|
|
<p>
|
|
To detect fairness issues, a definition of fairness must be decided
|
|
upon. For example, fairness may be defined as whether pairs of
|
|
similar individuals get similar predictions (individual fairness), or
|
|
it may be defined as whether predictions are similar across a
|
|
majority and minority group according to some characteristics (group
|
|
fairness). This fairness measure may be computed once the AI has
|
|
been trained to identify potential unfairness, or may be computed
|
|
during the AI training so that it can take the notion of fairness
|
|
into account when it adjusts its parameters.
|
|
</p>
|
|
|
|
<p>
|
|
Free Software is also important here, as it <strong>allows everyone to check
|
|
for fairness issues, whether by inspecting the source code or by
|
|
running the AI directly and analysing its predictions.</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<figure>
|
|
<img src="https://pics.fsfe.org/uploads/medium/8d10355a4334cefde7f2f21d6f44daaa.jpg"/>
|
|
<figcaption>
|
|
Vincent Lequertier presents crucial points about AI during an FSFE Community meeting in Bolzano. Italy, 2019.
|
|
</figcaption>
|
|
</figure>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
Your research focuses on healthcare, a field that has universally
|
|
raised the question of supporting Open Science. To what extent are
|
|
health metrics and biometrics open? Is artificial intelligence for
|
|
healthcare a big and globally collaborative aim or independent and
|
|
competitive?
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
Well, it depends! Because of security and privacy, access to
|
|
individualized healthcare metrics is often restricted and each study
|
|
using them must be approved by a ethical committee. However,
|
|
aggregated statistics may be widely available. For example, the
|
|
website data.gouv.fr has a <a href="https://www.data.gouv.fr/fr/pages/donnees-sante/">section
|
|
dedicated to healthcare</a>. Also, the data related to COVID-19 are
|
|
public, and <a href="https://covidtracker.fr/">the most popular
|
|
website</a> to visualize these data as well as other tools <a href="https://github.com/CovidTrackerFr">are Free Software</a>.
|
|
</p>
|
|
|
|
<blockquote>
|
|
The openness and collaborative aspects of research on AI will
|
|
improve, partly because scientific journals encourage researchers to
|
|
share all the research materials, including source code, and also
|
|
because funding institutions can also ask them to do so. [...] I also think that the
|
|
line of reasoning around our "<a href="https://publiccode.eu/">Public Money? Public
|
|
Code!</a>" campaign applies for AI research.
|
|
</blockquote>
|
|
|
|
<p>
|
|
However, it should be noted that data without enough granularity can
|
|
reduce the AI's performance in healthcare, as, just like humans,
|
|
an AI application needs to have detailed information, especially if
|
|
the goal of the AI is to make predictions at the individual level.
|
|
Because healthcare outcomes are so dependent on context, prediction
|
|
abilities depend on specific healthcare situations. More open data
|
|
and more Free Software (i.e. Open Science) make it easier to
|
|
collaborate. A shared dataset released under a Free Software licence creates a "playground" where AI
|
|
models can be easily compared and where we can create benchmarking
|
|
tasks, such as hospital length of stay prediction. Without a proper
|
|
benchmarking task, finding methodological improvements is harder. An
|
|
example of an open dataset for healthcare is <a href="https://physionet.org/content/mimiciii/">MIMIC</a>.
|
|
Also, a lot of papers about AI research are freely available on arxiv.org. I
|
|
think the openness and collaborative aspects of research on AI will
|
|
improve, partly because scientific journals encourage researchers to
|
|
share all the research materials, including source code, and also
|
|
because funding institutions can also ask them do do so. For example
|
|
the Horizon 2020 program of the European Union <a href="https://ec.europa.eu/programmes/horizon2020/en/h2020-section/open-science-open-access">values
|
|
Open Science</a>.
|
|
</p>
|
|
|
|
<p>
|
|
I also think that the line of reasoning around our "<a href="https://publiccode.eu/">Public Money? Public Code!</a>"
|
|
campaign applies for AI research.
|
|
</p>
|
|
</div>
|
|
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
A common expectation for the future of AI is that it can have abrupt
|
|
economical and societal impact by making many job positions
|
|
redundant. Do you see this as a possibility for the upcoming years?
|
|
If so, is there any practice that could alleviate these
|
|
consequences? Would Free Software be one?
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
I think AI has come a long way in the last ten years. It is more and
|
|
more able to organize and structure information. The fields which
|
|
have made the most impressive progress are natural language
|
|
processing (i.e. tasks involving text such as sentiment analysis) and
|
|
computer vision (i.e. tasks involving images such as image
|
|
classification). In natural language processing, deep learning models
|
|
can semantically understand words and documents as well as the
|
|
relationships between them. So I think the jobs where AI will be able
|
|
to assist us (I consider things only from a technical point of view
|
|
here) are the jobs dealing with a lot of structured information that
|
|
needs to be understood, processed, and memorized, as AI is becoming
|
|
better at this than us. For example, AI-based software has shown
|
|
good results in assisting in <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6268174/">radiology</a>,
|
|
legal document analysis, and programming (see next question). So
|
|
it's possible that AI makes people more efficient, which would
|
|
reduce the amount of human work required. However, this work would
|
|
require skills where AI does not work well at the time of writing,
|
|
such as creativity or emphatic and thoughtful communication.
|
|
</p>
|
|
|
|
<blockquote>
|
|
The jobs where AI will be able to assist us (I consider things only
|
|
from a technical point of view here) are the jobs dealing with a lot
|
|
of structured information that needs to be understood, processed, and
|
|
memorized, as AI is becoming better at this than us. For example,
|
|
AI-based software has shown good results in assisting in <a
|
|
href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6268174/">radiology</a>,
|
|
legal document analysis, and programming.
|
|
</blockquote>
|
|
|
|
<p>
|
|
If AI is bound to get better, and will at some point have the
|
|
capacity to completely automate some work, transparency and fairness
|
|
can only become more and more important. Although not sufficient,
|
|
Free Software is a big part of what helps putting strong
|
|
safeguards in place.
|
|
</p>
|
|
<p>
|
|
However, I don't think it's up to the scientific community to design
|
|
policies around employment. Putting together a proof of concept or
|
|
finding a novel theory that could automate some work is not a reason
|
|
for implanting it in everyday lives. In the past years, the EU has
|
|
already had to deal with AI applications that are impressive
|
|
technically but raise ethical concerns. For example, the Clearview AI
|
|
facial recognition platform has been judged illegal in some EU
|
|
countries, and citizens have the right to opt out from this
|
|
technology. The next few years will be important with regard to AI
|
|
ethical concerns, and the upcoming EU Artificial Intelligence Act
|
|
might play a big role in it.
|
|
</p>
|
|
<p>
|
|
And finally, although I'm not a historian, I think that over the last
|
|
centuries we have made tremendous technological progress and
|
|
society has always evolved along with it. Thinking about the past
|
|
challenges of technological improvements would help us to
|
|
understand whether they would be different this time around, and how to
|
|
deal with them as best as we can.
|
|
</p>
|
|
</div>
|
|
|
|
<figure>
|
|
<img src="https://pics.fsfe.org/uploads/medium/97bcc3b73d413a4eae7143025fb6323f.jpg"/>
|
|
<figcaption>FSFE Community Meeting during Rencontres mondiales du logiciel libre conference in Strasbourg. France, 2018.</figcaption>
|
|
</figure>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
What legal issues do you think will be raised regarding AI in the
|
|
next ten years? Would it be issues of ownership or responsibility?
|
|
For example, we are already seeing ethical and technical aspects of
|
|
the AI ownership in Github's Copilot. We
|
|
are interested to know what the upcoming crucial questions are,
|
|
according to you.
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
Issues around ownership and responsibility will be very important,
|
|
and Copilot is a prime example of that, where the fundamental
|
|
question is whether AI creations can be considered as novel ideas,
|
|
and, if they do, whether they are copyrightable on their own.
|
|
Specifically on Copilot, the fact that a code completion tool may
|
|
yield <a href="https://twitter.com/mitsuhiko/status/1410886329924194309">straight
|
|
copies of licensed work</a> can be problematic, as, at the time of
|
|
writing, the AI does not know the licence under which the source of
|
|
autocompleted code is released, and how the licence should be
|
|
respected. For example, to the best of my knowledge, it is not clear
|
|
whether code autocompleted by Copilot originally released under the
|
|
GNU Public Licence makes the rest of the project a derivative work.
|
|
Being able to freely use source code often comes with obligations
|
|
that need to be fulfilled, regardless of whether it is accessed by an
|
|
AI or a human being. Our <a href="https://reuse.software/">REUSE
|
|
project</a>, which aims to make it easier to programmatically
|
|
understand how a project and its diverse components are licensed, may
|
|
help building licensing-aware programming tools. The same legal
|
|
troubles apply to other models able to generate content, in domains
|
|
such as in painting or in music production.
|
|
</p>
|
|
|
|
<p>
|
|
Another legal issue is with patents, where <strong>the question of
|
|
whether an AI can be a patent author is still undecided</strong>. In
|
|
the UK and EU, a patent whose inventor was an AI was rejected because
|
|
they considered that AI does not have a legal personality and cannot
|
|
have a legal right over its output. But a couple of months ago, <a
|
|
href="https://theconversation.com/in-a-world-first-south-africa-grants-patent-to-an-artificial-intelligence-system-165623">the
|
|
first patent which lists AI as the inventor was approved</a>.
|
|
</p>
|
|
</div>
|
|
|
|
<blockquote>
|
|
The fundamental question is whether AI creations can be considered as
|
|
novel ideas, and, if they do, whether they are copyrightable on their
|
|
own.
|
|
</blockquote>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
Is there any book about artificial intelligence you would like to recommend to our readers?
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
|
|
<p>
|
|
I can't recommend "Genesis" from Bernard Beckett enough. It is a
|
|
small novel showing a philosophical debate around the questions of
|
|
what it means to be human and of whether machines can have
|
|
consciousness. The classic "I, Robot" from Isaac Asimov also raises
|
|
many questions that make a lot of sense today (it was published in
|
|
1950!) If we are building autonomous robots with some freedom of
|
|
action, what safeguards must we put in place? The book is really
|
|
about how to ensure AI works as intended.
|
|
</p>
|
|
|
|
</div>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
You have been a part of the FSFE for several years. What is an
|
|
important thing that you learnt from this experience?
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
I learnt that Free Software can be viewed from a lot of different
|
|
angles and is not only a technical topic. This translates into the
|
|
diversity and breadth of our community. This diversity is a huge
|
|
strength.
|
|
</p>
|
|
</div>
|
|
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>
|
|
And what is a story that still makes you smile when you remember it?
|
|
</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
My first FOSDEM in 2019. I met some awesome people from our
|
|
community. That was really heartwarming.
|
|
</p>
|
|
</div>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>As a last question, what do you wish the FSFE for the next 20 years?</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<div class="answer">
|
|
<p>
|
|
I wish the FSFE will be able to tackle the challenges ahead. The next
|
|
years will be full of innovations that will make technology even more
|
|
ubiquitous in our lives. I hope we will be able to keep spreading the
|
|
word about Free Software and the values behind it.
|
|
</p>
|
|
</div>
|
|
|
|
<blockquote>
|
|
Being able to freely use source code often comes with obligations
|
|
that need to be fulfilled, regardless of whether it is accessed by an
|
|
AI or a human being. Our <a href="https://reuse.software/">REUSE
|
|
project</a>, which aims to make it easier to programmatically
|
|
understand how a project and its diverse components are licensed, may
|
|
help building licensing-aware programming tools.
|
|
</blockquote>
|
|
|
|
<div class="question">
|
|
<p>
|
|
<strong>FSFE: Thank you very much!</strong>
|
|
</p>
|
|
</div>
|
|
|
|
<h3>About "20 Years FSFE"</h3>
|
|
|
|
<p>
|
|
In 2021 the Free Software Foundation Europe turns 20. This means
|
|
two decades of <a href="/about/about.html">empowering users to
|
|
control technology</a>.
|
|
</p>
|
|
<figure>
|
|
<img src="https://pics.fsfe.org/uploads/medium/53aeebdafa0a0c83324dcafc1e4bef1c.png" alt="Banner with FSFE 20 years. FSFE since 2001" />
|
|
</figure>
|
|
<p>
|
|
Turning 20 is a time when we like to take a breath and to look back
|
|
on the road we have come, to reflect the milestones we have passed,
|
|
the successes we have achieved, the stories we have written, and
|
|
the moments that brought us together and that we will always
|
|
joyfully remember. In 2021 we want to give momentum to the
|
|
FSFE and even more to our pan-European community, the community
|
|
that has formed and always will form the shoulders that our
|
|
movement relies on.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>20 Years FSFE is meant to be a celebration of everyone who
|
|
has accompanied us in the past or still does. Thank you for your
|
|
place in the structure of the FSFE today and for setting the
|
|
foundation for the next decades of software freedom to
|
|
come.</strong>
|
|
</p>
|
|
|
|
</body>
|
|
|
|
<tags>
|
|
<tag key="front-page"/>
|
|
<tag key="interview">Interview</tag>
|
|
<tag key="AI">AI</tag>
|
|
<tag key="fsfe20">20 Years FSFE</tag>
|
|
</tags>
|
|
|
|
|
|
<discussion href="https://community.fsfe.org/t/769"/>
|
|
<image url="https://pics.fsfe.org/uploads/medium/8d10355a4334cefde7f2f21d6f44daaa.jpg"/>
|
|
|
|
</html>
|