Panos Ipeirotis
@ipeirotis
Followers
4K
Following
841
Media
115
Statuses
4K
Professor at @NYUStern. #ai #datascience #crowdsourcing I have opinions.
New York, NY
Joined October 2008
Started teaching my "Dealing with Data" class to the new cohort of TechMBA students today. The class covers standard Pandas material (data reading, manipulation, plotting, etc.). Typically, I spend most of the class going through notebooks, which students run on Google Colab.
0
0
8
That hits very close to home.
In the interview for hiring this guy, I started with a SQL question. He said "that's a dumb question.". Naturally I was taken aback by that response. "OK, let's move on then.". He was in his early 40s and had a pretty old-school stats background, so I figured a barrage of tough.
0
0
1
Mariner identified (correctly) 19 related orders. However, when I asked to fetch the receipts, reconcile with my AMEX and prepare the reimbursement packet, it complained that this is too much work.
0
0
3
Just tried Project Mariner at Google with the following prompt:. "I want you to log in to my Amazon account and look at all the 2025 orders and identify the orders that are related to my work as an NYU professor, so that I can reimburse them. My Amazon account credentials are:.
2
1
9
Few realize how much cohort-based structures streamline university logistics, enhance curricular rigor, and improve the student experience. Shared courses build community and focus. The “choose your own path” model is deeply overrated.
You can imagine a world where many universities offer bespoke programs that offer students little choice--their choice is which university to attend. Once in they have to submit themselves to the rigor of their chosen program.
0
0
2
A moment of silence for one of the most influential, and disastrously wrong, business books ever written. It convinced a generation of execs that technology was a commodity: that tech didn’t matter. Thoughts and prayers for the Blockbusters and Borders of the world, who watched
1
2
11
A different take:. * Today, we perform vastly more web searches than the total visits ever made to libraries or encyclopedias. * People today are significantly better informed than in pre-internet times. If you think misinformation is bad now, revisit the Pulitzer-Hearst.
A hypothesis on the accelerating decline of reading: . * Broadly speaking, people read for pleasure/entertainment and for learning/obtaining information. * Reading for pleasure has been declining for a while and is being replaced by videos (very sharply among young people).
0
0
2
It is called Stockholm syndrome.
Counterpoint: a working paper culture plus pub lag gives enough time to assess results by a much larger group of experts than peer review alone, before the paper goes on record. We’ve seen example of this working recently…. Obviously bad for tenure/promotion though.
0
0
0
Fifth: What's the Risk Here?. For researchers at universities or small startups, casually using LibGen might seem harmless. The risks escalate quickly when you're a global company. Training on "presumed free" copyrighted data differs from "willful infringement"—the legal term for.
1
0
0
Fourth: Is there a Legal Defense for using LibGen?. There is a very reasonable argument that training an AI is transformative—after all, an LLM doesn’t copy books; it learns from them. Consider also the LAION case from Germany. LAION, a nonprofit, scraped images from stock photo.
1
0
0
Third: What About the EU?. If you thought U.S. law was tricky, the EU adds another layer of complexity. They don’t have a broad "fair use" policy, but they've introduced exceptions specifically for Text and Data Mining (TDM). Good news for researchers and AI developers, right?.
1
0
0
Second: Does Training an AI Change the Equation?. Here’s where it gets fuzzy. In the U.S., you can claim "fair use"—the idea that some copying is permissible if you're transforming the original work into something new and valuable. (We covered this in an earlier blog post.).
1
0
0
First: Is Using LibGen Even Legal?. Short answer: Absolutely not. Downloading copyrighted books from LibGen is textbook piracy. Think of it like grabbing a handful of snacks at the supermarket without paying—it's convenient but totally illegal.
1
0
0
This isn't hypothetical. Recently, Meta made headlines for allegedly training their flagship LLM, LLaMA, on content from LibGen. But—can you even do that?. Let’s unpack the legal mess behind the scenes, step-by-step. 2/6.
1
0
0
Training LLaMA using LibGen: Hack, a Theft, or Just Fair Use?. Imagine you're building a Large Language Model. You need data—lots of it. If you can find text data of high quality, vetted, truthful, and useful, it would be. great! So, naturally, you head online and find a.
1
1
2
The solution to the problem of the student loans is for the universities underwriting the loans themselves, and not being able to sell the debt to others. (At most, let it be used for collateral for access to short-term liquidity.) . Long-term incentive alignment.
0
0
1
We tested the o1-pro model to give us a detailed analysis of the legal landscape around copyright and the use of copyrighted materials to train LLMs. The full discussion is available at . The post is a quick attempt to summarize the (much) longer report.
0
0
0