Luca's Avatar

Luca

@lucab.phd.bsky.social

ML Ethics, Tech Policy, Italian food, Typos

716 Followers  |  116 Following  |  79 Posts  |  Joined: 25.04.2023  |  1.5508

Latest posts by lucab.phd on Bluesky

My contribution for the @techpolicypress.bsky.social and @datasociety.bsky.social series β€œIdeologies of Control” where we explore the relationship between power and democracy

04.06.2025 16:01 β€” πŸ‘ 5    πŸ” 5    πŸ’¬ 0    πŸ“Œ 0
Post image

how the email found me:

13.05.2025 16:59 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

how the email found me:

26.04.2025 03:39 β€” πŸ‘ 3    πŸ” 2    πŸ’¬ 1    πŸ“Œ 0

LLMs are powerful tools that are being used to automate more and more tasks. When nuances are importantβ€”such as in content moderationβ€”we should ask a lot of questions about their impacts on human rights.

@ecnl.bsky.social wrote a great report, where I offered some comments. Read it below:

15.04.2025 11:07 β€” πŸ‘ 3    πŸ” 2    πŸ’¬ 0    πŸ“Œ 0
Post image

how the email found me:

14.04.2025 07:11 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Drawing on my own experience in auditing Twitter's systems (from the inside), I explore the crucial role business leaders can play in navigating these emerging challenges.

27.03.2025 00:06 β€” πŸ‘ 1    πŸ” 1    πŸ’¬ 0    πŸ“Œ 0

Audits are going to be a more critical business practice, especially given the increasing regulatory requirements. However they are not a straightforward and linear process, rather a web of different decisions. By their nature, they don't only bring technical challenges, but organizational ones.

27.03.2025 00:06 β€” πŸ‘ 1    πŸ” 1    πŸ’¬ 1    πŸ“Œ 0
Preview
What Leaders Need to Know About Auditing AI While audits are becoming a core feature of working with AI, they don’t have a predetermined process that follows a straight line; rather, they are a web of different decisions, both from the business...

I wrote about Algorithmic Audits for Harvard Business Review!

hbr.org/2025/03/what...

27.03.2025 00:06 β€” πŸ‘ 2    πŸ” 1    πŸ’¬ 1    πŸ“Œ 0
Post image

how the email found me:

22.03.2025 21:38 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Post image

per my previous email

16.03.2025 15:31 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

Hulu that's not how words work

20.02.2025 20:41 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

earthquake?

13.02.2025 22:00 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
COMUNICATO STAMPA - IA: il Garante privacy chiede informazioni a DeepSeek. Possibile rischio per i dati di milioni di persone in Italia

The Italian Data Protection Authority @gpdp.it t has requested information to DeepSeek to confirm what data is collected and used. A similar move happened with ChatGPT in 2023, just before incognito mode was introduced.

www.garanteprivacy.it/home/docweb/...

30.01.2025 16:33 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

i do wonder a lot if things would have gone differently if we had an union

29.01.2025 20:08 β€” πŸ‘ 8    πŸ” 1    πŸ’¬ 0    πŸ“Œ 0

wow they didn't even change the subject of the email

29.01.2025 20:04 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image Post image 23.01.2025 02:10 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
General-Purpose AI Code of Practice The first General-Purpose AI Code of Practice will detail the AI Act rules for providers of general-purpose AI models and general-purpose AI models with systemic risks.


The next plenary meeting is going to be Monday 27th January at 5pm CET, or 8 am if you are in the West Coast!

digital-strategy.ec.europa.eu/en/policies/...

16.01.2025 00:42 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

☞ Energy consumption (known on estimate) is also something that regulators can demand. This really helps put into perspective the performance of models. There has been a race towards bigger models, but not every task requires the latest and biggest, (e.g. locally run tasks)

16.01.2025 00:42 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Considering the various lawsuits that are happening in the US on this exact topic, it's going to be interesting to see what features and models are NOT going to be released in EU.

16.01.2025 00:42 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

☞ Furthermore, regulators can ask more information about how the data was filtered to remove illegal and undesirable content, and how providers have acquired the legal rights to the data.

16.01.2025 00:42 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

☞ Providers will have to disclose to regulators and downstream providers β€œinformation on data used for training, testing and validation”. This is particularly important as training data (and their provenance) is considered part of the β€œsecret sauce” of their models

16.01.2025 00:42 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

☞ Providers will have to disclose to the regulators a list of the name of the organization that label data for them. No more hiding behind 3rd party firms

16.01.2025 00:42 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

☞ They acknowledge that Generative AI models β€œpresent unique innovation opportunities, but also challenges to artists, authors, and other creators, and to the way their creative content is created, distributed, used, and consumed” (Recital 105 AI Act)

16.01.2025 00:42 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
General-Purpose AI Code of Practice The first General-Purpose AI Code of Practice will detail the AI Act rules for providers of general-purpose AI models and general-purpose AI models with systemic risks.

Comments on the second draft of the European Commission GPAI Code of Practice are due tomorrow am EU time.

The document is an important step in the operationalization of the AI Act, together with the harmonized standards by CEN and CENELEC

Some (random) interesting bits in no particular order:e

16.01.2025 00:42 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

How do I get in in this research group?

04.01.2025 18:00 β€” πŸ‘ 7    πŸ” 1    πŸ’¬ 0    πŸ“Œ 0

a cursed ai image for a cursed mood

01.01.2025 00:07 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Post image

going in 2025 like

31.12.2024 20:56 β€” πŸ‘ 4    πŸ” 0    πŸ’¬ 2    πŸ“Œ 0

While everyone is focused on the GPAI Code of Practice, this is an important milestone to achieve meaningful transparency for large online platform and search engines operating in the EU. In fact this is unprecedented access from external researchers!

12.12.2024 19:22 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
Data Sharing and the Delegated Act of Europe's DSA | TechPolicy.Press Researchers will have a chance to get meaningful data of the most precious kind: unaggregated, non-public platform data, writes Luca Belli.

In late October the @ec.europa.eu released a draft of the Delegated Act of Article 40 of the DSA, which mandates data sharing with external researchers.

My thoughts on why it matters for @techpolicypress.bsky.social . Special thanks to Justin Hendrix.

www.techpolicy.press/data-sharing...

12.12.2024 19:22 β€” πŸ‘ 7    πŸ” 4    πŸ’¬ 1    πŸ“Œ 0
Post image

how the email found me:

10.12.2024 18:38 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

@lucab.phd is following 20 prominent accounts