Very happy to share that The Oxford Handbook of Digital Ethics is now finished and will be out in the next few months! Some of it is already published online. If you're interested in
#DigitalEthics
,
#AIEthics
,
#ethics
,
#privacy
,
#AI
,
#philosophy
, this one is for you... 🧵👇
Russell Group, which is made up of 24 leading
#universities
including Oxford and Cambridge, has announced that students will be allowed to use generative AI. Here's what you need to know 🧵👇 1/
#AIEthics
OpenAI and Microsoft are being sued in California for $3 billion over alleged
#privacy
violations. Here's a short thread on why this lawsuit is important and worth following 🧵👇
#AIEthics
If you were really worried, wouldn't you have collaborated with more AI ethicists and policymakers before releasing these models to give society time to come up with appropriate guardrails?
#ChatGPT
#GPT4
#AIEthics
"For over 6 years, people were often wrongly labeled as fraudsters, with dire consequences." "Some victims committed suicide. More than a thousand children were taken into foster care."
#AI
#AIEthics
Proud of
#Spain
. Might be the diet, but I doubt it—chorizo, fried foods, and late dinners don't seem that healthy. It's more likely a strong public health system. And let's not underestimate human warmth and friendliness.
1/ If Socrates was indeed the wisest Greek, that makes large language models like
#ChatGPT
the most foolish systems in our societies. Here's why 🧵👇
#AIEthics
Very happy to share that The Oxford Handbook of Digital Ethics is now available online, and will be out in paper in December! If you're interested in
#DigitalEthics
,
#AIEthics
,
#ethics
,
#privacy
,
#AI
,
#philosophy
, this one is for you... 🧵 (30% discount available!)
No, machines are not learning good. They are not learning ethics. They are learning to *mimic* moral language. There's a difference. To learn ethics, they would need to have an understanding of right and wrong that goes well beyond language.
#AIEthics
Very happy to share my latest academic article! I use the philosophical zombie experiment to explore the relationship between sentience and moral agency, and how those relate to
#AI
. Check it out! It's
#openaccess
!
#AIEthics
Very excited to share a new
#book
with you! The book responds to five basic questions: Where does
#privacy
come from? What is it? Why does privacy matter? What should we do about privacy? Where are we now? Pre-order now!:
@EthicsInAI
@TORCHOxford
It's here! It's out. Everyone, meet The
#Ethics
of
#Privacy
and
#Surveillance
. It offers a comprehensive philosophical theory of privacy. Importantly, it practically doesn't overlap with Privacy Is Power. Here's a summary 🧵👇
#books
#BooksWorthReading
What if I don't want "powerful generative models right at the core" of my laptop or phone? What if I don't want it to "draw on my personal context"?
#AIEthics
Bad actors are using
#GenerativeAI
to create
#YouTube
content for children that looks educational but is disinformation. Disinformation sells more; lies can be more spectacular than truth. Bad actors earn money with more views, and so does
#Google
.
#OpenAI
is being accused of having trained its
#LLM
on books without authors' permission. One hypothesis is that they used Libgen, or something similar. Books are ideal for training models: carefully written and edited.
#copyright
.
"AI should no longer be treated as a pure STEM discipline but rather a transdisciplinary one that requires technical knowledge, yes, but also insights from the social sciences and humanities. There should be mandatory courses on
#AIethics
and governance."
Good thing Oxford has a BA in computer science and philosophy 😊 "If you want to pursue a career in engineering, you should focus on learning
#philosophy
in addition to traditional engineering coursework. It will improve your code."
#AIEthics
My book, Privacy Is Power, is now available for pre-order! It’s written for those who care about
#privacy
, and those who think they don’t care about privacy. I argue that whoever has the
#data
will have the
#power
.
UK:
US:
AI is making research worse. The Scopus CiteScore ranking of philosophy journals has 3 journals among its top 10 that are fake. Articles are AI generated, with fake authors, and cite each other extensively to achieve a top ranking.
#AIEthics
h/t
@IEthics
Muy emocionada de compartir este curso online en ética de la inteligencia artificial que estoy haciendo en colaboración con la
@frdelpino
. El curso empieza el 4 de marzo de 2024, pero las plazas son limitadas, así que mejor inscribirse cuanto antes.
Honoured to be part of the International Advisory Council on AI for Spain, along with
@katecrawford
,
@francesca_bria
, and
@erika_widegren
, among other brilliant thinkers. It’s an important effort to make sure AI serves the public interest in Spain and beyond.
#AIEthics
Acabo de mantener un encuentro con el nuevo Consejo Asesor Internacional de Inteligencia Artificial. Quiero agradecerles su apoyo al desarrollo y regulación ética de una tecnología que está llamada a transformar el mundo.
España tiene un gran potencial para convertirse en un
"If writing helps us think, what happens when we surrender the process to AI? We risk becoming cognitively and expressively disempowered."
#AIEthics
#writing
"
#AI
still is merely a factual engine that acts based on probabilities and scores, mostly based on historical data, with no context of the implications of the info it is delivering."
#AIEthics
1/ Last week, the White House announced that it had secured “voluntary commitments” (i.e. pinky promises) from 7 leading AI companies to better manage
#AI
risks. Here’s why voluntary commitments are not enough, and what we can learn from how we regulated the car industry 🧵👇
Exciting news! I am joining the board of Proton's new foundation! I'm thrilled to become part of Proton's mission to protect
#privacy
. It's an honour to join
@andyyen
,
@timberners_lee
, et al in this new adventure in which people and rights come first. 1/
Delighted to share my latest piece in the
@HarvardBiz
, on why we need to use randomized controlled trials on
#AI
. We would never allow a drug to be sold in the market without having undergone testing. Why do we allow it with decision-making AI?
#AIEthics
Whoa. I didn't see that coming.
"The watchdog said there was no legal basis to justify "the mass collection and storage of personal data for the purpose of 'training' the algorithms underlying the operation of the platform".
#ChatGPT
#privacy
#GDPR
#Italy
Delighted to share my latest
@wired
piece. By deciding our fate on the basis of predictive
#algorithms
, we are turning people into robots. The ability to defy the odds is one of our greatest gifts, and we undermine it at our peril.
#AI
#AIEthics
@EthicsInAI
"A group of 5 related class action lawsuits allege that Honda, Toyota, Volkswagen, General Motors, and Ford “recorded and intercepted [car owners’] private text messages and call logs” when their phones were connected to the vehicles’ systems."
#privacy
What AI is missing are the embodied emotions that function as our moral compasses, as I argue in my paper on Moral Zombies:
"People tended to rate AI responses as more virtuous, intelligent, trustworthy, and rational "
#AIEthics
"AI image and text generation is pure primitive accumulation: expropriation of labour from the many for the enrichment and advancement of a few Silicon Valley tech companies."
How do you think companies like OpenAI should compensate creators?
#AIEthics
Snap, Facebook, Twitter and YouTube would have earned $10bn more from exploiting sensitive
#data
if
#Apple
hadn't given
#iPhone
users the option not to be spied on. Remember when
#Facebook
used to say that users were happy being tracked?
#privacy
Very grateful to the International Association for Computing and Philosophy for this award. 🙏😊 Also very grateful to
@HertfordCollege
,
@EthicsInAI
and
@PhilFacOx
for the very supportive work environment.
When
#Facebook
bought
#WhatsApp
in 2014, it promised that
#data
would not be shared between both companies. They broke their word in 2016. We are being reminded of that breach of trust with ever more intrusive and non-negotiable T&Cs. These companies count on us forgetting.
Read "Digitization, Surveillance, Colonialism" in which
@CarissaVeliz
warns ominously that the threat of total social surveillance begins the moment anything is digitized
"
#Meta
[is] developing a wristband that would read your brainwaves and
#Apple
[is] patenting a future version of AirPods that would scan your brain activity through your ears". I want out. Disappointed in Apple. Meta, to be expected.
#privacy
"In “The A.I. Mirror: How to Reclaim Our Humanity in an Age of Machine Thinking,”
@ShannonVallor
argues that we vastly underestimate our own richness compared with that of A.I. “
#books
#AIEthics
"If a model is built on data that may be unlawfully collected, it raises questions about whether anyone can use the tools legally."
#privacy
#LLMs
#AI
#AIEthics
"If the taking of data without consent is generative AI’s original sin, the widespread labor exploitation is its dirty secret. Lastly, there’s the energy usage, water consumption, and sustainability problems of generative AI."
#AIEthics
Important.
"Data brokers are buying and selling hundreds of millions of people’s location information, and too many researchers are inadvertently providing public-relations cover to this massive
#privacy
invasion by using the data in scientific studies."
“Although they are marketed as something that will enhance your mental health and well-being, they specialize in delivering dependency, loneliness, and toxicity, all while prying as much data as possible from you.”
#AIEthics
#privacy
"I didn’t leave the conference feeling confident AI was going to play a meaningful role in advancing any of the UN goals. In fact, the most interesting speeches were about how AI is doing the opposite."
#AIEthics
By
@Melissahei
"We’re in the tricky phase of AI adoption, where big tech companies have spent millions—sometimes billions—of $ developing their own tools or investing in others, and need to try and come up with reasons to justify that investment." By
@stokel
#AIEthics
A
#facial
recognition database, used by police departments, was built in part with 30 billion photos the company scraped from
#Facebook
and other social media users without their permission, creating what critics call a "perpetual police line-up".
#privacy
One good reason not to use facial recognition:
#privacy
breaches of data that cannot be modified (faces, unlike passwords, cannot be changed). Other good reasons include preserving freedom and democracy.
“As always,
#surveillance
isn’t safety.”
#AIEthics
#Amazon
is full of junk books made by
#AI
. Bad for authors, bad for readers, bad for Amazon. Is it a
#copyright
infringement? It's unclear. We'll need someone to sue to figure it out.
#AIEthics
#books
Zuck is bragging about just how much data
#Meta
has to train
#AI
: more than CommonCrawl. Very reassuring for
#privacy
. "Perhaps we shouldn’t trust a billionaire who’s been reckless with personal information that he wants to exploit yet again."
#AIEthics
But isn't this kind of challenge going to stifle progress? No. Technical innovation that leads to the erosion of rights and democracy is not progress. And
#AI
can be designed differently; not worse; just in accordance with rights and in support of liberal democracies. 12/
"Its surveillance plans, including controversial use of experimental AI video surveillance, are so extensive that the country had to change its laws to make the planned surveillance legal."
#Olympics
#AIEthics
Both
#Google
and
#Microsoft
employees, whose jobs are to think about
#AI
risks, tried to stop their companies from launching their chatbots. "The companies released their chatbots anyway."
#AIEthics
Mi nuevo artículo en
@elpais_opinion
. Tener más datos no garantiza más conocimiento, ni mucho menos más sabiduría o progreso. En algunos sentidos la tecnología está suponiendo un retroceso político y social, entre otras razones porque lo digital no olvida.
I’m absolutely thrilled to start working at the new Institute for Ethics in AI! It will be a privilege to work with our newly appointed Director,
@JTasioulas
, and with
@carinaprunkl
,
@tlechtable
, and Milo Phillips-Brown. We will announce new positions soon. Follow us!
@EthicsInAI
How are you finding Twitter these days? I no longer see the posts of my favorite people (on either tab), and I constantly see people who I don't follow (and who I don't want to follow). Also, it's glitchy. Sigh.
Every day is Book Day! I have 3 signed hardcover copies of
#PrivacyIsPower
with minor imperfections (I ripped out the first page, but no content is missing) that I'm happy to mail to anyone who'd like to read the
#book
.
#privacy
#BooksTwitter
#Surveillance
"Privacy is inevitable" People want
#privacy
. Democracy needs privacy. Privacy laws are getting better. Big tech is still engaging in large scale privacy violations. Enforcement needs to do some catching up. Tools like WebXray can help. Check it out.
On how the greater danger with AI is that it will produce mediocrity and that we will settle for it. "What is really being manufactured is us—our perceived needs, our attitudes toward what is good, our willingness to settle for “good enough.”"
#AIEthics
The carbon emissions of the computing industry already exceed those of the airline industry. "There's been such a race to build bigger and better models that almost every secondary consideration has been put aside."
#AIEthics
"The European Centre for Algorithmic Transparency (ECAT), which was officially inaugurated in Seville, Spain, is expected to play a major role in interrogating the algorithms of mainstream digital services"
#AIEthics
#AI
By
@riptari
.
My article for the Guardian: Countries with poor data protection laws typically fall into two categories: underdeveloped or authoritarian. But there is a third unpalatable possibility: the UK could develop into a data haven.
#privacy
#dataprotection
Companies that depend on your personal data for their enrichment want you to think that the battle for
#privacy
has been lost, but the battle is far from over.
#surveillance
#AIEthics
I've heard experts say that
#LLMs
learn a lot about the world in the process of statistically predicting what the next word is. I think that's wrong. If they "learn" anything at all, they learn about how humans use language—not about the world. The distinction matters.
#AIEthics
A group of authors has sued
#Anthropic
for violation of
#copyright
. The proposed class action lawsuit was filed in a California and alleges Anthropic “built a multibillion-dollar business by stealing hundreds of thousands of copyrighted books.”
.
@mer__edith
puts it well. "If you were to heed
@timnitGebru
's warnings you would have to significantly change the business of these companies." Existential risks warnings "project everything into the far future so they leave the status quo untouched."
Research shows that employees who are steered by digital nudges may lose some ethical competency. We should be mindful of this risk when using AI assistants.
#AIEthics
Check out my latest Nature piece, in which I argue that
#chatbots
shouldn't be using emojis. It's too deceptive and manipulative. Emojis hijack our empathic responses. Thank you,
@kevinroose
for the inspiration to write the piece.
#AIEthics
@EthicsInAI
"The European Data Protection Board (EDPB) has decided that large platforms such as Facebook and Instagram cannot force a “binary” pay or consent choice on users"
#privacy
#AIEthics
"Not only could the majority of participants not tell the real tweets from the ones generated by AI, but they generally found the AI's tweets more convincing."
#AIEthics
#LLMs
#Facialrecognition
firm
#Clearview
has run nearly a million searches for US
#police
.
It has 30bn images scraped online without users' permission.
Clearview has been fined millions of dollars for breaches of
#privacy
, but clearly, that's not enough.
To get a sense of just how dangerous personal data is, imagine a totalitarian regime having our real-time location data, facial print, political opinions... Let's end the data economy, before it's too late.
#PrivacyIsPower
@UniofOxford
@HertfordCollege
"ChatGPT gave John an “exceeds expectations” and Jane just a “meets expectations"
"Firms should have an AI ethics manager or consultant to constantly evaluate the use of AI at the workplace."
#AIEthics
Seinfeld, nailing it again.
#AIEthics
"So if I have this right, we are smart enough to invent
#AI
, dumb enough to need it, and still so stupid we're not sure if we did the right thing."
So much happening. "The body that unites
#Europe
's national privacy watchdogs said it had set up a task force on
#ChatGPT
, a potentially important first step toward a common policy on setting
#privacy
rules on artificial intelligence."
#AI
#AIEthics
"
#Amazon
,
#Meta
and
#Google
have funded and lobbied the (...) carbon accounting oversight body, and financed research that helps back up their positions"; “a way to rig the rules so the whole ecosystem can obfuscate what they are up to”.
#AIEthics
When thinking about whether collecting personal
#data
is a good idea, don't imagine it ending up in the best possible hands. Imagine it ending up in the worst possible hands, because there's a good chance that's exactly where it's going to end up.
#Privacy
On how money tied to tech, and ideologies like effective altruism, are infiltrating the halls of democracy and trying to co-opt
#AIEthics
and the regulation of
#AI
. People without ties to tech need funding to defend the public good.
Modern cars "know where you go, what you say, and how you move your body. Nissan’s privacy policy, for example, allows the company to collect and share drivers’ sexual activity, health diagnosis data, and genetic information, according to the report."
Data is the new oil; extracting it is dirty, and storing it is dangerous. “We are barely recognizing the negative externalities of oil production and consumption now, and it took us almost destroying the planet. We should do a better job for data.”
#privacy
Some good news:
#ClearviewAI
, a
#facialrecognition
company, is fined $22.6m for breach of
#privacy
laws (that's 60% of the funding that the company has raised from investors). The
#ICO
ordered the company to delete the personal
#data
of people in the
#UK
.
It's not only that
#chatGPT
fabricates sources. It's that it fabricates such plausible sources. E.g. it attributes books to me (e.g. Understanding Privacy) that I haven't written but could've...
#LLMs
are superb bullshitters (technical philosophical term!)