How to save the world and forward your career in 5 easy steps | Women in NLP Talks

Key Takeaways

hello everybody and thank you very much for coming to my talk my name is sabine now i'm a fourth year phd student at the university of edinburgh and i'm also an organizer with queer and ai and queern ai is a volunteer-run organization that represents lgbtq plus people within the ai community my research is on entailment detection and entailment detection um means the things that we assume to be true when we read a certain sentence for example if we read the sentence john bought an apple we assume that john owns the apple and um this is because of the way the concepts of buying and owning are defined in english and how our society defines these things but um today i will not talk about my research and my phd at all but instead i will talk about all the things that i did during my phd that aren't exactly research related but that nevertheless i think helped me with my career or my phd status ship in nlp but first of all i want to start you off with a story so at edinburgh in like um the nlp department there we have this tradition of uh pizza fridays or we used to when uh we were still doing work in person we would get together on a friday evening at around five and somebody from like either students or faculty would give a nice short talk or presentation about some non-academic topic like some hobby or research interest that they had that was like not primary nlp and one time a professor gave a talk about imposter syndrome and i will like not name any names here but like she is an incredibly accomplished researcher one like all the best paper awards has brilliant students and she talked about how she since the beginning of her career in nlp always felt like she didn't belong like her success was just a fluke just something that um yeah she got lucky with and she one day will be found out that she's actually not as good as anybody thinks which is exactly what imposter syndrome is about like this feeling of not belonging of being an imposter more or less and this talk was very touching how she said how she like dealt with his feelings and how we as students can deal with them too because this is something that can happen to you when you're surrounded like by brilliant people and after the talk there was like a q a part and one guy raised his hand and said something along the lines yeah this is more of a comment than a question this was a very nice talk but i never had a problem with this at all and everybody was like oh okay this is not a universal experience but we wish it were um except for like this being like a bit tone deaf to say what this instance brought up for me was this feeling of yeah it's easy to feel like you belong in a place when you're surrounded by people who look like you who talk like you who have the same background as you and who yeah you can see yourself represented in all levels of hierarchy and this is not the case for many people especially many women or queer people so let's talk a bit about that every one of us this is a picture of me in my workplace or on the way to there so each of us has um holds different identities or different words that we can describe ourselves as for example i'm a woman i'm queer i'm a second generation immigrant i'm white i'm cisgender able-bodied middle-class neurotypical and so on all of these identities in some situations award us with power or privilege for example because i'm white i'm less likely to be randomly searched by the police but some of these identities that we have can also be used against us to discriminate against us for example as a woman in a male dominated workspace i'm more likely to be harassed and to be discriminated against and i really don't want to sugarcoat how much it sucks to hold a marginalized identity in a hostile work environment or even in a work environment where this is just a minority position to be in like it comes with constant like second guessing and double standards that just your colleagues and your supervisors are not aware of because they're not aware of the barriers that you face in your life but on the other hand why this really sucks it also provides us with a certain insight that our peers don't have sorry i clicked the wrong button so we have a certain insight of where systems don't work or where systems go wrong because we face these barriers and these failures of systems all the time and then we have like these other parts of our identities where we hold a certain power for example by being in natural language processing or by working in a tech company where we our position and our knowledge awards us a certain power where people just believe us when we talk to them for example about computer science because we have a computer science degree they'll be like oh yeah that's somebody who i can trust on that and what i want to talk about this this talk or what like i want to be the guiding light of what i'll talk about now is how we can combine our insights that we have from like facing up barriers with the power that comes with being an nlp in like whatever position and um first of all i want to put a huge disclaimer here that what i wanted what i don't want this to be is like a plato year for people with marginalized identities to try harder or to do more there's absolutely no reason why you shouldn't deserve the same kind of respect and appreciation than your peers and just holding a marginalized identity in a hostile space is a huge effort and you're doing absolutely amazing just showing up so when i talk about all these things that i will be talking about next like mentorship or outreach this is not saying lean in and you have to like try harder to be accepted but rather you are enough and everybody should see this so in the next um half an hour so i want to talk about things that i've done during my phd that are yeah really inspired by this idea of like how can i use my insight as a queer person as a woman as a second generation immigrant and like other experience that i have of like systems being broken and the power that i have as a computer scientist as a phd student to combine these things to move things forward i want to talk about outreach that i've done mentoring that i've done i want to talk about grassroot organizations in nlp and lastly i will talk about a bit a bit about what to do when the thing that you want to exist doesn't exist how to start a thing from scratch so first of all outreach outreach can be anything where you talk about your expertise in your research or in your work to an audience that is like not your colleagues or other like researchers i for example did stand-up comedy about like the dangers of ai i paired up with an artist in a project where artists and scientists got together to visualize their research using artistic means and i also spoke to python programmers and data scientists about my research how is this saving the world how is you doing outreach good for everyone well first of all you can use this platform to raise awareness for important issues for example talking in form of like a comedy set about the dangers of ai can like help people to um be better informed citizens and be aware of like for example the discriminations that people face from facial recognition software and that those things are not like great impartial system but actually really biased and dangerous secondly every time you stand up in front of an audience as a person with a marginalized identity for example people can see you in a position of competence and power in a sense so when i stand up on a stand of comedy stage and i say like i'm i'm a phd student in computer science and you listen to me talk about ai for like 15 minutes or so somebody in that audience sees me as a woman and a computer scientist and clocks this as oh yeah like women are actually good at sciency stuff and like this can help like wear down barriers that exist in people's mind just like one bit at a time how is doing outreach good for your career well first of all it really helps your confidence if you like go up on the stage and see that nothing terrible happens this can help you like find the same kind of like center of competence when you talk to your colleagues or your supervisor also it helps you to learn to explain your research to different audiences and to like do this work of thinking okay what do the people who listen to me know and what are words or metaphors or other help for things that i can use to get my point across and that really helps you when again if you talk with your colleagues if you talk with your supervisors if you do a scientific presentation as well so i've collected some resources um for example uh the place where i did the stand-up comedy was bright club and they exist in edinburgh glasgow and like other cities in the uk then pi data and like other meetups are a great place to start if you wanna like talk to an audience of programmers um where like your research can meet kind of like an application level then um the arts project was um that's like an edinburgh institution the asco's lab but maybe there are like similar things in other cities so like that might be a place to start and then two things that i haven't done myself but that are also kind of outreach that you could do if you're um still a student you can join the three-minute thesis competition couldn't find a good a good homepage for that but maybe your university has like some kind of representation for that and the other one is the pint of science which is like a uk-based science festival where scientists go to pubs and like do kind of uh science education for like a broad audience so next on i want to talk about mentorship mentorship is what we when we team up with somebody who is either more advanced along a career line that we want to take or less advanced in the career line and like meet up with them or like um once every two weeks or once a month or like on some regular schedule and put in dedicated time to talk about career goals and obstacles to reaching these goals um i have done both uh these mentorships where i was like the the younger partner so to say like baby yoda or and where i was older yoda by uh mentoring an undergrad student and mentorship can be really helpful to just navigate this problem of not having role models or not having representation in your immediate surrounding so you might be like the only woman in your research group or you might be the only queer person you know at the university but you can find a mentor that is maybe like not locally there but that you can still like talk to and share experiences with and so this is good because it lifts up lifts up others who have maybe who share maybe marginalized identities with you or even not like i mean it is just a good way to share the experiences and the resources that you have another important point is that especially if you have holder marginalized identities these mentorship relationships could work as whisper networks to flag abusive [Music] supervisors or um abusive workplace situations and sometimes like this is the only way how this information can be passed down because institutional oversight is just lacking so i think it is important for example to uh have um hire a person for example um somewhere in your in your organization that like can maybe point out like okay this person has a history of harassment so like stay clear of them so maybe that seems like really obvious now but like it is good for your career as well because like you this networking and um sharing of resources of course helps you to navigate kind of the non-scientific side of your work like you can be really good at like the coding or the experiments and so on but like working in an environment that is not necessarily supportive of you will put up barriers that have nothing to do with this and so just having somebody you could talk to and you can work through these problems can be um a huge benefit it also um trains you in problem solving for example if you have like mnt or like an undergrad student that that you talked regularly it it also is great to put yourself in their mind space and to figure out okay what are good next steps what are good questions to ask to help them kind of figure out their problems and again it like is really good for your communication skills as well to even if it's like kind of talking about research on like a very abstracted level it still helps you to um yeah get in the habit get the ideas across resources for finding a mentor or finding a mentee can be sometimes a bit tricky so if you're still at university your careers office at your university might be a place to start then sometimes university might have like stem societies or computer science societies or like other kind of informal student clubs that organize these kind of things so if you're already out of university you can you might be able to reach out to an alumni network for like some university of binet or for example i found a mentor through like a scholarship that i received in my master's and then like they had kind of an alumni network that was running a mentorship program but often these things are not available to you if you're like not in a university environment you can also like really reach out to your friends and colleagues for example somebody you know who like graduated a year or two earlier you can reach out and be like hey would you be up to talking with me about like my work stuff every other week or you could reach out to colleagues who are like higher up the career ladder but you still feel like you have um a good way of talking to each other so i think when it comes to mentorship it is not even the most important part that like the person matches you like in expertise or or seniority but rather that you have like a good way of talking to each other and figuring out problems together and lastly another good source to find mentors are affinity groups and that's what i want to talk about next affinity groups or grassroots organizations and nlp are volunteer run organizations like queern ai or black and ai who um like stem from the desire of people to like work on better representation and tackle problems that are um really obvious in in the nlp community but like not obvious to the people higher up but only to those who are affected and kind of banding together to um fight discrimination in that way so what you see on the right here is um a session on diversity and inclusion that uh i've chaired at knuckle which where we brought together people from black and ai and mazakane and latinx and ai to discuss these things and i can really strongly advise to join a grassroot organization because it gives us more power and important issues for example one problem that trans people in academia face a lot is that academic journals are very restrictive about changing your names in a publication so if you sometimes change your name during the course of your life this might mean that all of your publication that you published before your name change are lost and you cannot like prove that you have actually a track record in publishing and due to advocating and like really reaching out to the publishers as an organization rather than like single people clearly i was able to exert the power that was needed to make um large journal publishing companies much more accessible for people who want to change their names and to to be better in this sense another thing is that if you are part of an organization like queern ai it is easier for you to access resources because it is easier to reach out to companies as for example an organization saying okay hi we're queer and ai we want to do a workshop can you like give a sponsorship money in exchange for like you advertising there and as i said it's a great source of role models or mentorship because in like the queen and members survey that we did we found out that 80 of the members of queern ai don't have role models in their immediate environment and having like this worldwide network of queer a.i researchers actually allows you to find somebody who has like similar interests to have like a similar identity and team up with them to be your mentor and like to just find people to collaborate with or just like build this personal network of knowledge and contacts that you can't find in your immediate environment so this again can spark off good opportunities for your career for example queern ai has a mailing list where people can share um job postings and where people share like summer schools or other projects that they're working on and this way you can kind of profit of kind of this informal knowledge of having for example a person at a company that you want intern at and that can help you figure out um what the requirements are that you have to fill um we also for example queen eye does a lot of things like um we have a dedicated undergrad mentorship program we have a program for financial support for grad school application when you for example need to get a language certificate or you need to pay administrative costs to apply to university and we also have a great series of conference conference socials where we also sponsor like conference attendance for people um i also want to borrow this amazing slide from shakir muhammad's uh keynote talk at knuckle 2021 where he put up on the map the um different affinity groups or grassroot organizations and and machine learning and ai that existed at the moment like east european machine learning or southeast asian machine learning mazakhana in africa and deep learning in dalbar hipu and here at the edge queern ai and black and ai who represent people all over the world and i strongly encourage people to join these um i collected a few here and you can also find them on the queen ai website like ostemp for example is kind of our mothership organization for queer people in the sciences and engineering and there's black and ai latinx and ai women in machine learning widening nlp which is like again kind of an over arching organization um for people with marginalized identities in nlp than black women nai and disability and ai for for example um and for queern ai i can say that la allies are very welcome so like even if you don't um like take the boxes in terms of identity if you're for example interested in low research machine translation then masakana is still like very much welcoming your contribution and also queern ai is inviting allies because we need to fight this fight together and lastly i want to talk about what to do if the thing that you need doesn't exist so what if there isn't an affinity group that like covers the identity that you identify with what if there isn't a meet up in your city that um covers your specific like research interest or a specific societal problem that that you want to tackle so i'm gonna take a sip of water so when i moved to edinburgh i was really struggling with the winters there because i have a seasonal affective disorder and um i would get really sad in november and december and this was also compounded by the fact that um i just moved to a new city i didn't know anybody there and i was like feeling really down and lonely and i found out that like making art with people and community really helped me with that and that it gave me something to look forward to and something that i could really identify with and at my university there were funds for organizing social events for people who work in computer science so a friend of mine and i applied for this funds and we bought a bunch of art supplies and for over a year before the lockdown happened um every week we would just put out all the art supplies and like put out these cool protective like plastic things over the table and it was really like low barrier entry like just people could come by and sit down with us and do whatever they wanted with our art supplies so it wasn't like taught or anything it was just like grab some paint and and paint with us and it was pretty nice like it is established a group of like 10 to 15 people who were kind of regulars who would would we would always see and yeah it was it was a good thing to do so starting things from scratch it's good because your needs might be shared by others so although it was something that i wanted to do that i enjoyed they were like the 10 those 10 15 people who would show up and also really enjoy it but like they would not have uh had the time or the energy or or the incentive to start that by themselves so for because i put something out there in the world others profited from it as well and it was great for again getting to know new people like i talked to like people from areas of computer science who i haven't known before or like administrative workers and which was really nice to get to know these people and last but not least it was fun so i mean you can't work if you're totally burned out so it's also good to integrate that in your career somewhere so in terms of resources where can you turn to if you want to start something from scratch because what you need in the world doesn't exist yet you can lots of things can be done without the need of extra funds like if you just want to start a reading group about a certain topic or if you want to like have a meet up for people who have like a shared identity with you sometimes just reaching people via a mailing list for example it's like your institution has a mailing list for people in computer science or phd students reaching out by this mailing list reaching out via meetup which is a home page or reaching out via other social media can already be enough to be like hey let's meet up in a cafe and talk there or let's meet up at like some other public space sometimes the thing that you want to put up needs funding because you need to say pay materials and you want to pay rent for like a room where you can put up a presentation if you are at your university sometimes like your grad school office is a good place to reach out to or like whatever other office is in charge of like social events at your school then if it's something that it has to do with like a marginalized identity sometimes you can reach out to local non-profits that have something to do with this marginalized identity another thing that can work for things in computer science is company sponsorship because often companies are really interested in recruiting people or like selling their software products to other people in computer sciences or in like the nlp industry and um for example like the big five like apple or google whatever they have specific campus outreach groups that you can like write to or like you can just team up with like some local company that you know to say like hey can you sponsor this event that we think of and like you can advertise at this event and last but not least again affinity groups is a great place to get funding from if you have like a specific idea that fits with queer and ai or like any other group those groups have funds from company sponsorship and they are really like happy to put things up like for example this diversity inclusion panel was something that just like started as an idea where i said oh yeah we should do this and then like we collected the people and we found a platform to host this so this again might might be a good way to put the thing that you want to see in the world into existence so i want to close here with again this is not a to-do list this is not the things that you need to somehow achieve to be accepted in your community because you are already absolutely amazing for holding marginalized identities in a hostile environment and the last thing that i want to say is like together we can make great things happen because we have unique insights and we have unique powers and putting those things together i hope will make the world better thank you very much thank you so much for your talk it's very uh very important and it's full of very useful information for me as a minority in ai and computing science and i think for everyone else as well uh so now if anyone has a question please write it in the chat thank you it was fun putting this talk together so yeah okay until people start asking questions i i will ask some of my questions uh so you you mentioned that uh women in academia and in computing science they are looked at as they are not as competent uh and i myself came across uh incidents when i say a comment or something and um a lecturer will smile like you know this smile that yeah what you're saying is stupid and that will come back to me making me feel even worse and i thought i need to to take an action like if someone does that i should i don't react because that will help me later in my day and later in my life but i don't really know what kind of actions should i take when such situation happens do you do you have any suggestions well yeah like it's really difficult and i mean like the hard answer is the one to you don't need to be in a space where you you're treated like this like i've before i come to edinburgh i worked in like a situation where my colleagues basically threw me under the bus because i was the lowest in hierarchy and i was also the only woman there and like they kind of set me up on a project to fail and that was like it really broke my heart so much like i was so pissed off and yeah i mean i could have called them out on it i could have like i mean i talked to kind of the higher up boss about this but it was ignored and so on and i think like the bottom line for me was like i should have left the situation earlier because there are places that treat you better and like sometimes you end up if it's like your phd supervisor or if it's a job that you like really depend on where you feel like okay i have i have no alternative but yeah it's like really hard to say i don't know i'd say to myself in retrospect i should have left earlier but um of course you can only say these things in retrospect so yeah i i i don't i don't have like a short term solution so to say like okay if somebody like gives you a mean comment then uh that's the thing you need to do but it's it's more like on the grand scheme of things if these things pile up i think you need to be like clear to yourself like okay how much more of this can i take and i think like this is something that i also learned during my phd and so on to like have a better sense of like okay this is this is the level of that i can take before i walk away so yeah i don't know yeah i i know it's hard the problem is those subtle things really contribute to my self-force i think everyone's self-worth [Music] and if we leave them to happen more more often eventually we will not feel like we are strong enough to stand up for it or to leave the place or something uh yeah yeah it is it's something uh something difficult and uh i have to say that the complaints um strategy doesn't always work like you can't complain but nothing will happen yeah most likely everything will stay the same um so i think just need to care for our mental health uh we just should be aware of that um yeah yeah i mean like caring for your mental health is so important and like another thing that i kind of learned during phd is also like just to find other areas that like are not necessarily your job or so if where you can like feel this sense of like ownership or competence if you can't always get that in other places so but yeah i think i think that's also like it's a dangerous thing to say like you have to endure in a toxic workplace just take care of your mental health like it's i think it's really both things it's it's like being clear about yourself okay how much can i take and then like also making sure that you don't burn yourself out yeah yeah okay waiting for more questions i will ask another question so you mentioned how important mentorship is but it's not easy to find a mentor yes it's definitely yeah it is it's not easy to find a mentor like one thing that i i did is really just reach out to my friends like i have like one of them like two of them are here like when it where uh like i um just reached out to like people that i i know in person for like kind of pure peer-to-peer mentorship where i said like hey would you be up to like meet once a week or meet in the morning before we start our work day to just like talk about what we plan to do or like would you be up to like meet with me once a week and we talk about like work related stuff so kind of yeah if there are no formalized networks for mentorship then you have to kind of like go go this weird way and like create these things for yourself by asking friends then but it's it's difficult like what if you don't have friends that like kind of qualify for this where you'd say like okay this would be a person where who i think has more experience like i don't know like again affinity groups like queern ai or um or like the other ones they have also kind of more or less formalized mentorship programs where you can team up with people so that's one other thing that i can like recommend maybe like reaching out to their slack channels or their um uh mailing lists or for example like attending um a social event that they're hosting and kind of like getting to know people that way um yeah like it's like the mentorship question is really difficult because like some universities will have kind of a formalized way to like just assign you a mentor if you like sign up for the program but like it's hard to to conjure that up from from nothingness yeah and how would you promote yourself as a mentor um well like the the one like the one man mentee relationship that i had was like because it was kind of set up by the women in stem society at my university where they just like kind of you had a questionnaire where you put in the things that you are competent in and then they just like matched uh undergrad student for you but um if you want to be a mentor like i mean this it can be a peer-to-peer thing right where you just like say in your friend's circle like hey would you be out for talking about you your career and then can be like it's like a two-sided thing where like one time around it's you like taking on the role of the mentor and kind of asking the question walking through the problems and then the other time it's like your friend doing that for you and then this way it can be like both sided thing yeah i don't know like then again alumni networks are another thing where where this can happen if you like there there are some things i don't know um how like different universities handle this right like so i only know from my personal experience at like some universities have like an alumni i don't know program where where you can do mentorship as well okay and um you talked earlier in your presentation that you joined activities like bright club to get some confidence talking on stage but now this experience is really hard to to have because of the lockdown yeah so um do you have uh i don't know suggestions for alternatives well i mean one thing we're doing here right now right like i mean just putting up like kind of a virtual space to talk to is is like when like lots of like kind of these talk series have shifted online so that's also kind of a benefit if you're like not living in a place where these things these things happen or if you're like not you know at a university and so on it's now kind of easier because you have the access of like just going to meet up and like joining a meetup for example or um i think a similar thing about public speaking is toastmasters i've never joined that but they also have like groups that are organized over meetups and that is also kind of like a public speaking thing that's like not really like computer science or nlp focus at all but that is like really just for people to practice their public speaking skills and are you worried that when things go back to normal if they go back to normal it will be more difficult again to stand on a stage and face real audience yeah yeah like i mean i have no idea what it will be like when it comes back to normal like i i yeah i think i have kind of a mental image of like oh when we're all like meeting a person again it will be so and so but i think i'm just like projecting from my own experience and it will of course be different but yeah oh um somebody poses can you talk more about your experience in the bright group so yeah i can laura posted that um in the chat i think i can i can say something about that so the way i i came to do this was just that like a friend of mine told me that she did it and it was fun and i was like oh yeah i could try this um and it is organized in edinburgh quite informally there it's just like two or three people who have like this connection with the comedy club there where they like put this on they do kind of the administrative thing and you meet up with them and whoever else signed up for the next rotation of it and you do like a few sessions where they just teach you how to structure um a set like a comedy set so this was really helpful for them for me to get like what is the structure what are kind of the the goal posts that i need to have in mind what does like how does funny even work how do i set up a story that i tell so that people will laugh at some point so that that was really cool and then you can like um yeah start thinking about like okay what is it about my research that i care about that i want to put on the stage so people listen to it and take away some kind of message and um it was like a really rewarding experience so um it's it's whenever like for me it was the hardest part was kind of writing the set and like setting everything up and being like okay doesn't make sense and you need to like have some people as a sounding board where you can say like okay is this funny is this funny is this funny because like you totally don't know that yourself what what like somebody else would would consider a funny thing about computer science but yeah like just like getting up on the stage and being there for those seven minutes that you have it's it's really cool because um yeah it's really a confidence booster because like you literally have a room full of people laughing at you at your jokes not if you laugh at your jokes so it's really confidence building i can i can highly recommend it yeah you're welcome okay until we get uh more questions i have one more question to ask um so um i don't know if you have um if you like from your practice or you're joining uh different uh affiliations like queer in ai sometimes people start those kind of affiliations but they don't really continue like they come together excited and they start setting up activities and then within a few months it just fades away um do you do you have any advice how to keep it up yeah like it's it's definitely difficult like i've uh had like different experiences with that like for example with the art space that i ha have put up it's like it had definitely like phases where like at the beginning we had like 20 or 30 people and then we were like down to five people and so on so like there is like something like a seasonal flow but what i found kind of helped is consistency so that you like just you do it every week or every month or like whatever your date is you advertise it every time and you're just like there the same time the same place and sometimes two people show up sometimes 50 people show up but like through this consistency people just know that they can rely on it and and then they'll come back hopefully but also to just like keep keep on advertising um and find for example different channels to advertise it as well like when the art space kind of died down we like reached out more to social media we set up like uh an instagram account and we like did did like these like little paper airplanes that i had on the slides that you can like throw around with like the dates and times of it and so on so yeah it it it is a tricky question like like i definitely experience it as well like with reading groups that i've like been part of that like it kind of people are around for the first three chapters and then kind of flattens out so i think like yeah what what helps there is also having like a core of people that you can kind of rely on like those two or three friends that will just always show up and that's that's also kind of good but yeah it's i think i think it's it's a definitely difficult question of like what keeps those things sustainable like for example queerney i is just huge now but i think what keeps it going and what like kept its rise from like when it was created four years ago is that like there are a handful of people who are really into it and who are really there all the time and who make sure that like we have a social at every like nlp conference and like yeah i think again having more than just one or two people but having three or four that are kind of like a hardcore who are really motivated keep helps keeps these things going like i mean i would not have been able to do the art space if it wasn't like for my friend that i like did this with and like we would just show show up for it together and go and eat pizza afterwards and like if if two other people came it was okay because at least i got to hang out with my friend so that that's i think if i can give any advice that that would be advice oh yeah uh okay google slide links i will do that so i'm a res i'm researching nlp now and i'm new to this could you briefly say wrote map kind of thing resources so um what exactly do you mean by by roadmap kind of thing like can you can you specify this um because i think there are just many ways to go about things and my my way of getting to where i am now was quite non-linear and random like it looks now from from the perspective now it looks like it all makes sense but it was actually like kind of a zigzaggy thing and i i don't have like a clear trajectory of like oh yeah in two years i want to be the ceo of my own startup award okay so i mean they are so you say you're an undergrad so you definitely get like taking machine learning courses taking nlp learning courses is a great place to start then doing internships at uh companies that do nlp stuff is also great like um i had kind of a part-time job during my master's for example where i got hands-on experience with coding nlp applications while getting kind of the theoretical background from from being a class which was really good but also super stressful like i don't recommend people to have part-time jobs when they go to university but like sometimes that's what you have to do um [Music] so yeah like just take the courses that your university offers that there are like great things online like for example the the stanford nlp uh course like machine learning for natural language processing all of that is on youtube like if you just google stanford machine learning for nlp course like their material is pretty good and then yeah i would recommend doing internships and figuring out what exactly like is the part of nlp that interests you most what what is the like area that you want to go into or like again affinity groups um masahane for example um they uh do low resource machine translation or like low research nlp applications and um they have a really cool slack channel and um weekly meetups where they discuss the projects that they're in and that's like a thing if you want to do um some kind of coding work in your free time just like as kind of a cool project to show off for example in your portfolio when you apply to places um oh yeah cool yeah that's that's again that's a really good place to start because people there are incredibly nice and and like enthusiastic and they're really happy about like anybody who can help them with things so that's yeah a huge recommendation to them as well