"Hi everyone, I would like to fix a translation in the audio / subtitle that better captures what President Zelenskyy was saying. I will delete this post once we find a good translation. The President said:"
"UPDATE #3 (FINAL): I went with 5: "We cracked down hard on everyone" to make intended meaning absolutely clear."
except 'slap on the wrist' is still there and will be "fixed" after everyone who wanted to hear the interview already listened.
GP maybe found one nit to pick too vigorously. However, their first point of asking to conduct the interview in Russian is such a big deal to me that I’ll probably not be listening to Lex anymore. Either it was a completely tone-deaf request from someone excited to practice his rusty Russian, or it was a political play to help frame Zelensky as unreasonable. Neither are good looks.
I’m reminded of an interview William Buckley gave Malcolm X in which the first question was “but what is your real name?” Its a request he knew 100% would be denied, doesn’t pertain to any important issues, and frames his interviewee as uncooperative from the first question.
Now I’m not saying Lex has any responsibility to give Zelensky softball questions. On the other hand he is just some schmuck* in a suit interviewing the commander in chief of a country at war and asking if they could use the language of their aggressors. I don’t like it
* ok maybe I’m biased against lex a little already…
Because of the war, speaking Russian has become political. Many Ukrainians who spoke Russian as their first language now refuse to ever speak it, at least publicly.
It is an odd request though considering Zelenskyy speaks very good English and Lex normally does all his podcasts in English.
Because Milei didn't want to use English. Which is fine. He could have done the Zelenskyy interview in English/Ukrainian. That would have been normal. Again, the Russian language is very politically charged in Ukraine right now, for obvious reasons. Which is why Zelenskyy doesn't typically use it in public despite it being his native tongue...
Lex seems to use his interviews as an excuse to listen to his own voice, he just looooves talking over people and is generally insufferable. I wish all those interviews were with someone less in love with themselves.
How is changing "we punish" into "slap on the wrist" while using AI to make it seem like Zelensky said it himself, and on top of that moving this single reply to the very beginning to set the tone _not_ a big deal? Its the book definition of spreading fake AI content.
There is no "slap on the wrist" expression in russian language, there is however "схватить вора за руку" meaning catch the thief red-handed while word for word it goes "grab thief by the hand". When you say in russian 'beat their hands' it means punishment, when you say in English "slap on the wrist" it means no punishment or light reprimand.
I think pretty much any interviews with Zelensky and or Putin are helpful. I was skeptical that the Tucker Putin interview would be too pro Russian but there was Putin going on about how Hitler was ok and WW2 was the Poles fault for resisting Hitler's desire to take their land and even Tucker looking awkward.
It's a bad look for Lex Fridman but at least Zelenskyy comes comes out looking pretty good.
It's funny that Lex hasn't sensed the shifting tides, with Putin turning down Trump's plan and Trump becoming ever more pro-Ukrainian.
Apart from that it's interesting to see the difference between Lex and Zelenskyy. They're similar ages, both Jewish, both have Ukrainian roots, both speak Russian. But Lex's family left at a time when Ukrainians and Russians had positive relations and basically missed the separate development of the two nations through the last 30 years (and apparently has been completely asleep through the last 10 years).
He’s just a guy trying to use what leverage he thinks he has to negotiate, or pretending he can negotiate, but in the end he will do whatever his boss tells him.
A three-hour interview with four language tracks – Ukrainian, English, Russian and mixed.
I believe we live in a great time where language barriers are being erased before our eyes.
The work of Eleven Labs is certainly impressive and warrants respect.
Of course, there are certain technical imperfections.
Besides language, AI still cannot reproduce the speaker's breathing or diction flaws.
I really hope these aspects will be overcome as well.
I haven’t seen any practical implementations of duplex TTS/ASR, although some preprint research exists on this topic.
It would also be beneficial to see a large language model capable of voice-to-SSML conversion, in addition to the existing text-to-voice functionality.
Finally, I am hoping for more competitive solutions to emerge, particularly from companies like Play.ht, Fish Audio, and Deepgram, among others. This could drive innovation and improve overall quality.
26 comments
[ 4.4 ms ] story [ 75.8 ms ] thread- Used AI to overdub "translation" using Zelensky's own synthesized voice. Actually putting words into Zelensky's mouth.
- Manipulated translation to replace Zelensky's "we hit people" into "slap everyone on the wrist" while no such expression exist in russian.
- Moved that question to the very beginning (0:40 second) so every American listening will hear this first.
- Posted on Social media about "innocent mistake" https://t.me/s/lexfridman/329
"Hi everyone, I would like to fix a translation in the audio / subtitle that better captures what President Zelenskyy was saying. I will delete this post once we find a good translation. The President said:"
"UPDATE #3 (FINAL): I went with 5: "We cracked down hard on everyone" to make intended meaning absolutely clear."
except 'slap on the wrist' is still there and will be "fixed" after everyone who wanted to hear the interview already listened.
Good job agent Alexei Fedotov!
I’m reminded of an interview William Buckley gave Malcolm X in which the first question was “but what is your real name?” Its a request he knew 100% would be denied, doesn’t pertain to any important issues, and frames his interviewee as uncooperative from the first question.
Now I’m not saying Lex has any responsibility to give Zelensky softball questions. On the other hand he is just some schmuck* in a suit interviewing the commander in chief of a country at war and asking if they could use the language of their aggressors. I don’t like it
* ok maybe I’m biased against lex a little already…
It is an odd request though considering Zelenskyy speaks very good English and Lex normally does all his podcasts in English.
He's a native Russian speaker who grew up in Moscow. Zelensky is also a native Russian speaker.
There is no "slap on the wrist" expression in russian language, there is however "схватить вора за руку" meaning catch the thief red-handed while word for word it goes "grab thief by the hand". When you say in russian 'beat their hands' it means punishment, when you say in English "slap on the wrist" it means no punishment or light reprimand.
> When you say in russian 'beat their hands' it means punishment
Do you speak russian? They have "получишь по рукам" "получить по рукам" which has the same meaning as "slap on the wrist"
It's funny that Lex hasn't sensed the shifting tides, with Putin turning down Trump's plan and Trump becoming ever more pro-Ukrainian.
Apart from that it's interesting to see the difference between Lex and Zelenskyy. They're similar ages, both Jewish, both have Ukrainian roots, both speak Russian. But Lex's family left at a time when Ukrainians and Russians had positive relations and basically missed the separate development of the two nations through the last 30 years (and apparently has been completely asleep through the last 10 years).
The work of Eleven Labs is certainly impressive and warrants respect.
Of course, there are certain technical imperfections. Besides language, AI still cannot reproduce the speaker's breathing or diction flaws. I really hope these aspects will be overcome as well.
I haven’t seen any practical implementations of duplex TTS/ASR, although some preprint research exists on this topic.
It would also be beneficial to see a large language model capable of voice-to-SSML conversion, in addition to the existing text-to-voice functionality.
Finally, I am hoping for more competitive solutions to emerge, particularly from companies like Play.ht, Fish Audio, and Deepgram, among others. This could drive innovation and improve overall quality.