Tareq Amin
speaker
433 appearances
2 recordings
2 series
first heard Nov 2025
last heard 4 Nov
Tareq Amin’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 2 in all, peaking in Nov 2025 with 2.
Appearances
One of the acquisitions that has happened in Humane, and this is where I was really, really surprised,
to see the depth of science team I have today in Humane.
I mean, this is the first comment I get in my social post to say, yes, you have capital, but there is no people, no talent.
So we ended up building a foundation model.
And I'll tell you the reason we did this.
We built it from the scratch.
It was not distilled of any open source tool.
Well, for two reasons.
I just wanted to see, I mean, A, let's venture into this to understand the depth and the capability of the organization.
Second thing, when it comes to culture, language, and biases, I felt that it's important for us to really participate in this arena.
So we launched in Saudi Arabia what we call Humane Chat.
I just wanted to see the reaction from the people that uses this model and the whole objective of it.
Arabic first preference is the way we train this model, not English first preference.
Sorry, just on that.
I think for us, on any model that is used today, quality of data is always much better than, I think, just feeding it quantity of data.
Our model today has a proprietary set of data in Arabic language that you cannot find on the public internet.
So we had a proprietary data set that was important and obviously within the Arabic language the government today uses all its correspondence, transactions, translations is in Arabic.
So the preference of training first in Arabic is very different than you starting in English and then you add to it the Arabic components.
um so we built that we launched it it actually now became the number one uh app in the app store in in the country um it's really really an important thing for us and i'm going to be really specific that we didn't do this to say we're better than openai or we're better than x or rather than that
It was important for us to train the team on understanding how to build the entire stack.
Showing 61–80 of 433 · page 4 of 22
← Previous
Next →