Tina Eliassi-Rad
speaker
154 appearances
1 recordings
1 series
first heard Jan 2025
last heard Jan 2025
Tina Eliassi-Rad’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsNo recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.
Appearances
The problem is that I can hold a human accountable. I can sue a human being. Who am I going to sue? You know what I mean? And especially in America, we're very litigious. And so then this gets into accountability. And in fact, there's a lot of work in the government.
For example, our government is putting a lot of our tax dollars into like trustworthy machine learning, trustworthy AI, et cetera, et cetera. And to me, it rings a little hollow because there's no accountability. Like, how can I trust you if there's no accountability? I feel like they go hand to hand. And so there's some of that going on, which is like, You know, who am I going to sue?
Am I going to sue OpenAI because it's sexist and misogynist? Like one of its products is sexist and misogynist. You know, that's not the case right now.
Perhaps, right? The thing is, at this point, what it gives out is what's the most probable and what it believes you will like, right? So it's a two-place function, what's probable and what you will like. But yes, you could definitely do that.
And there's this comedian, unfortunately, I forget his name now, but he was saying the secret to a long marriage is to never say what comes to your mind first or second. Always say the The third thing that comes to your mind, right? And this goes back to what you were just saying. Maybe you should just say this third thing, the third most probable thing.
And in fact, along those lines, usually the students who use these generative AI tools for like math problems, math homeworks, the first answer is usually wrong because a lot of the answers that have been uploaded into like Course Hero, et cetera, et cetera, they're wrong. Usually it's the second answer that's the correct answer.
These are just anecdotal, right? Like I haven't had anybody do like a systemic study of this, but that like usually the first answer is not quite there, right?
Yeah. So in the true computer science, AI, machine learning sense, we're very good at coming up with names for our system. So we called it Life2Vec. So we're just putting your life into a vector space, whether you like it or not.
But you're just a vector in this vector space. Now, basically, the idea is that if you look at these large language models, right, so they're analyzing sequences. And so as human beings, we also have a life story. That's a sequence. Right. And so I was lucky enough to work with a group of scientists in Denmark.
So if America has a surveillance capitalism, in Denmark they have surveillance socialism. So there is a department there, Department of Statistics, they call it, like Ministry of Statistics that collects information about people. And so we had information for about 6 million people who have lived in Denmark from 2008 to 2020.
And we were like, well, can we write stories for these people in a way and then feed it to what is the heart of these large language models, a transformer model, which is basically just the architecture of a neural network that learns association weights for within some context window.
um and that's what we did so but instead of so for example chat gpt goes online and gobbles up all this bad data that that or that people have put in all the misogynistic sexist data we didn't do that so we had very good data from this department of statistics and we created our own artificial symbolic language
And then we fit that artificial symbolic language for these six million people into a transformer model. And then we were able to predict life events. And so one of them that caught the media's eye was, will somebody between the age of 35 and 65 pass away in the next four years? And we picked that age range because that's a harder age range to predict for.
Like if you're over 65, then it's easier to predict whether you're going to pass away in the next four years. And if you're younger than 35, it's also easy. The other, right, you're unlikely to pass away. And so that's one of the things. The other prediction task was like, will you leave Denmark? You know, so then you can predict for that.
But it had this similar technology as these large language models, which is like you have this one, what they call like predefined, where you just learn based on the data that you have what's likely to happen next. And then you fine tune it for whatever prediction task that you have.
It's a logical encoding because the data that the Department of Statistics has in Denmark is all tables. So it is not like this kind of sequence. So then you could say, like, Tina was born in Copenhagen in December, blah, blah, blah, right? And we could generate a natural language, but that's difficult. Why would we do that?
So then we generated a vocabulary for this artificial symbolic language, and then we And that was actually a lot of the intellectual property of the work is like, okay, well, how do you take these tables and then create this artificial symbolic language that then you can give to a transformer model?
Well, the thing that we found, which was very interesting, I think, so like the accuracy in terms of the model was about like 78%, et cetera. And I think that's why people were showing a lot of interest in it. But to me, that wasn't really the takeaway.
The takeaway actually was that labor data is a very good indication of whether somebody in that age range is going to pass away in the next four years or not, because health data is very noisy and inconsistent. So even in Denmark, where they have universal health care, it's not like everybody goes to the doctor all the time and you have good data for them.
And then the other stuff was basically just which sector you were working in. Right. So if you're like an electrician. It's a bad thing. It's not a very good thing. Right. As opposed to like an office worker. So the labor data was actually very, very helpful than the health data.
Showing 81–100 of 154 · page 5 of 8
← Previous
Next →