Profile picture
Alex does #Movember @adrechsel
, 31 tweets, 22 min read Read on Twitter
Translation Director-General Rytis Martikonis kicks off #TEF2018 here in Brussels. You can follow along on YouTube and @Slidoapp; just use our hashtag.
I‘ll be representing @EUInterpreters, talking about how interpreters use data & technology. (Come say hi!) Hoping to add to the momentum of the upcoming launch of the Knowledge Center. @CathyPe1971 @jhsaseta #1nt #scickci
Our moderator is @AnaGuerberof - kudos and good luck trying to tame hundreds of curious translators over two days! 😉 #xl8
Ana acknowledges the @EUInterpreters working for us today (English, French - but no German? 😱) and she briefly mentions “human parity” of machine interpreting. Food for thought… #1nt
Braving Brussels traffic today, our keynote speaker has arrived: @NozhaBoujemaa, research director at Inria. Her topic: “Trusted impact of #AI on the language industry” - Data isn’t worth much without algorithms to make sense of it! #TEF2018 (More info: project.inria.fr/nozhaboujemaa/)
While Nozha mentions data *visualisation* as one of the five pillars for #DataScience, a quick shout-out to @drawnalism who cover #TEF2018 visually.
This is, from @NozhaBoujemaa, is key: “Data-driven” does not equal “objective”! Algorithms need to be transparent by design, input data must be chosen carefully (complete, correct, diverse, up to date…). #AI #Bias
Really chuffed to see this all-female panel at #TEF2018: @AnaGuerberof, @lspecia, @Knowldgillusion and @AlisonBLowndes discussing how #AI transforms society and the #xl8 industry.
Just now realising I’ve been using the alternative/wrong hashtag. #2018TEF —> This also works over on Sli.do! You can add your questions there.
“Data is oil. And just like oil, data is crude.” - nice bit from @Knowldgillusion #2018TEF
I disagree with @Knowldgillusion when she says that for every piece of data that is stored, a decision has been made. Today still, so much data is stored simply because it is possible and cheap: “We’ll figure out what to do with it later.” #2018TEF #eudatap
Potentially heretical question: The technical level and knowledge at #2018TEF is 💯. Same at #1nt events?

/cc @ProfessorOlsen
I’m just going to put this here…
To clarify, @Knowldgillusion, I think what you describe is the ideal. But the decisions are still often badly thought through and not deliberate. (Many examples over on @internetofshit)
As we resume #2018TEF, here’s the excellent @drawnalismMatt at work! #sketchnote /cc @saurau Matt Buck at his easel, drawing.A sketchnote on safe AI and deep learningA sketchnote for the panel on AI & societyA sketchnote: Data and translation - a pretty picture?
This should be an interesting panel (or should I say #Manel?): A Neural MT reality check, with Dieter Rummel, @johntins, Bruno Pouliquen (@WIPO) and Matt Post. #xl8

Live stream:
(Follow-up from panel 1: If you want to learn about, and play around with, AI and Machine Learning, check these resources from @GoogleAI:

ai.google/education/

experiments.withgoogle.com/collection/ai

/cc @AlexandraSoska #2018TEF)
It’s the end of the road for statistical #MT. For neural MT, there are so many roads opening up all the time. (@johntins) #xl9 #2018TEF
“We need clean data and smart people for #NMT,” says Bruno Pouliquen. Less is more, as long as it’s quality data.
Tough love for current #CAT tools from Bruno Pouliquen. (WIPO language services use three different ones.) A good CAT tool is yet to be invented… #2018TEF #xl8
Love this question over on Slido: “Will parity be reached when the machine is as subjective/fallible as a human translator?” #2018TEF /cc @ProfessorOlsen
And another Slido question from “an interested participant”: “What is real progress: Helping humans or overriding them?” 🤔

Share your question here: app.sli.do/event/apmn47dy…
« Je pense en français, je parle en anglais. Et après, un interprète reformule tout. » - un beau moment là, de Bruno Pouliquen. #2018TEF
Look, it’s @erik_hansson! On the panel discussing the translator’s perspective of working with data. #2018TEF
The @erik_hansson conundrum: I did my research and found the right term. Why do revisors, agencies and clients insist on the outdated/worse term from the CAT tool? #xl8 #2018TEF
🚨 Depends-on-the-context™ klaxon! 🚨
The problem is not technology. It’s how people use it, what they expect it to do. —> Spot on, @lnunesvieira! #xl8 #1nt #terptech #2018TEF
If you’d like to know more about how Gruffydd Jones and the National Welsh Assembly work with Machine Translation, look at this: assembly.wales/en/abthome/abo… #2018TEF #xl8
#TIL that “Fog computing" is a thing. #2018TEF #CloudIsSoPassé Fog computing[1] or fog networking, also known as fogging,[2][3] is an architecture that uses edge devices to carry out a substantial amount of computation, storage, communication locally and routed over the internet backbone.
My favourite comment today: “Let’s not get into #blockchain!” 😂 #2018TEF
TFW you ask a panel of EU representatives: “Who wants to answer the #Brexit question?” #2018TEF
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Alex does #Movember
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!