Bastien Liétard

💬 they/them 🐈 Cat person, French PhD. student 📚 Computational Linguistics & Semantics at @Inria_Lille.

[Bastien Liétard @BastienLtrd]

📣 Missed my talk at @LChangeWorkshop ? (Well probably because I should've advertised it...) No worries then, here is a rundown !

"Towards an Onomasiological Study of Lexical Semantic Change Through the Induction of Concepts"

Paper link: https://t.co/iJfQoDmoz3

[François Chollet @fchollet]

I believe that a clear understanding of intelligence at the level of fundamental principles is not just possible, but necessary for the development of AGI.

[Justus Mattern @ ICML @MatternJustus]

The only piece of advice I give to undergrads that want to get into research is to cold email PhD students with a good track record.
Most undergrads are bottlenecked by research ideas whereas good PhD students have way too many ideas that they cannot execute. If you can code reasonably well, PhD students will be more than happy to work with you

[François Chollet @fchollet]

I kinda miss when "AI Twitter" was folks doing AI research or app development, posting under their own identities. Slightly more intellectual depth than anime avatars whose main AI-related qualification is paying for a ChatGPT subscription.

[Mikel Artetxe @artetxem]

I also have my own opinions, but I rarely do anything to defend them besides some occasional Twitter rants. When people like me have disagreements about some direction that ACL has taken, it is also on us to be more proactive and make our voice heard where it matters. 6/n

[Mikel Artetxe @artetxem]

Some people with strong opinions (legitimately) use all means at their hand to influence the community. They publish position papers, give opinionated talks, start hot debates, and serve for executive positions. 4/n

[Subbarao Kambhampati (కంభంపాటి సుబ్బారావు) @rao2z]

So I was present in person at @emilymbender's #ACL2024 thought provoking presidential address.
As the past president of @realaaai, I did feel compelled to say something in Q&A. I actually got up from my seat intending to grab the mic, but decided to sit down thinking that it may not be appropriate as I am, after all, an outsider attending his first ever ACL.

I did have an opportunity to chat with Emily at length after the talk.

Let me share my (current) thoughts.. This is a long post with nuance that I hope is befitting the original talk.

Of course, as the past president of @realaaai, my first impulse was to push back. Despite what Emily is clearly worried about--which is about hyped-up-corporatised-AI with seemingly little more than obligatory lip service to (under the "responsible #AI" umbrella) to societal impacts--the title does, at first glance, insinuate indirect guilt to #AI the scientific field--something I--as someone who worked in the area for nearly 40 years--and many others--would be disappointed with.

I am sure there are people across all of #AI research who share Emily's conerns, and I was justifiably worried that they will be turned off by the syntactically exclusionary tone of the title (something that Lucas @soldni--who I didn't know before--brought up with his thoughtful question..).

There is also the concern that none of the academic conferences/communities are really controlled top-down in anyway. I am sure there are papers in ACL--including this one, that I can personally vouch for--that already run afoul of Emily's concerns.

Finally, there is the issue that all of us need allies and ACL can hardly afford to turn off the folks who consider themselves #AI researchers--myself included--who do share Emily's (well founded, IMHO) concerns.

Having thought about this more though, I am however persuaded by Emily's point (that she made in response to someone's question)--that ACL community should first make a statement about what characterizes them, before focusing on building bridges to the allies. This, IMHO, is a legitimate thoughful exercise for an elected president.

I do hope those intended allies see what I eventually saw--that it really is a statement of concern about corporatization of the science of AI and its impact/ramifications on the computational study of human language, than about #AI the scientific discipline.

I often end my "general interest" #AI talks with Fred Friendly's quote "Not to make up your mind, but to open it. To make the agony of decision-making so intense that you can escape only by thinking." On that benchmark, I have no doubt that Emily's address succeeded.

[The Data Therapist was @ CoNLL/ ACL2026 @yuvalmarton]

The more you care about the linguistic side, the more CL’ly you are. The more you care about the engineering, the more NLPer you are. Yes, to me it’s graded, there is no clear-cut borderline.

[Christopher Manning @chrmanning]

I agree with much of both @emilymbender’s #ACL2024 presidential talk and @yoavgo’s rejoinder, but I want to comment on just one aspect where I disagree with both: the definition and domain of CL vs NLP. 🧵👇

[Bastien Liétard @BastienLtrd]

@asayeed @zehavoc French food does not need improvements (expect at being less meat-based only) ;))

[Bastien Liétard @BastienLtrd]

Btw @LChangeWorkshop was an absolute banger and I feel very lucky working on this topic, all of you should get into lexical semantics and language variations just to have the opportunity to meet these wonderful people !

[William N. Havard @WilliamNHavard]

"ACL is historically and should remain: A venue for people who care about the language in language technology" @emilymbender. This is SO true. How many times have I seen or heard people saying they didn't really care about language and/or linguistics.

[Bastien Liétard @BastienLtrd]

Of course, this method can be improved in many ways ! This paper is a proof-of-concept of a methodology we find worth the look, as it brings an interesting new light on NLP studies of Lexical Semantic Change !

Thank you for your reading and go take a look at the paper !

[Bastien Liétard @BastienLtrd]

📣 Missed my talk at @LChangeWorkshop ? (Well probably because I should've advertised it...) No worries then, here is a rundown !

"Towards an Onomasiological Study of Lexical Semantic Change Through the Induction of Concepts"

Paper link: https://t.co/iJfQoDmoz3

[Bastien Liétard @BastienLtrd]

Finally, we leverage our method to explore onomasiological aspects of change and classify concept clusters depending on whether their naming expended, shrinked or remained identical. You can find discussions on some examples in the paper.