Internet Culture and Technology

Life, Liberty, and Superintelligence

Are American institutions ready for the AI age?

3QuarksDaily pointed me to an essay in Arena on Life, Liberty, and Superintelligence. The essay starts with the question that Dario Amodei tackled in Machines of Loving Grace, namely, what might be the benefits of artificial intelligence (AI). It then questions whether we could actually achieve the potential benefits without the political will and changes needed to nimbly pivot.

Benefits: Amodei outlined a set of domains where intelligence could make a real difference, including:

Biology and health,
Neuroscience and mind,
Economic development and poverty, and
Peace and governance.

Amodei concluded with some thoughts on Work and meaning, though the loss of work and meaning may not be a benefit.

It is important that we talk about the benefits as massive investments are made in infrastructure for AI. We should discuss what we think we are going to get other than some very rich people and yet more powerful companies. Discussion of benefits can also balance the extensive documentation of risks.

Institutions: The essay then focuses on whether we could actually see the benefits Amodei outlines even if we get powerful AI. Ball points out that everyone (like JD Vance) believes the USA should lead in AI, but questions if we have the political will and appropriate institutions,

Viewed in this light, the better purpose of “AI policy” is not to create guardrails for AI — though most people agree some guardrails will be needed. Instead, our task is to create the institutions we will need for a world transformed by AI—the mechanisms required to make the most of a novus ordo seclorum. America leads the world in AI development; she must also lead the world in the governance of AI, just as our constitution has lit the Earth for two-and-a-half centuries. To describe this undertaking in shrill and quarrelsome terms like “AI policy” or, worse yet, “AI regulation,” falls far short of the job that is before us.

There could be other countries (read China) who may lag when it comes to innovation, but are better able to deploy and implement the innovations. What sort of institutions and politics does one need to be able to flexibly and ethically redesign civil institutions?

Metaculus on AGI Outcomes

Listening to Jacob Steinhardt on The Hinton Lectures™ I learned about Metaculus, which is a forecasting service which is a public benefit company. It has a focus area on AI Progress with lots of AI related forecasts, (which seems to be a huge area of interest.) This service coordinates human forecasts and builds infrastructure to facilitate others in forecasting.

Neat!

UNESCO – Artificial Intelligence for Information Accessibility (AI4IA) Conference

Yesterday I organized a satellite panel for the UNESCO – Artificial Intelligence for Information Accessibility (AI4IA) Conference. This full conference takes place on GatherTown, a conferencing system that feels like an 8-bit 80s game. You wander around our AI4IA conference space and talk with others who are close and watch short prerecorded video talks of which there are about 60. I’m proud that Amii and the University of Alberta provided the technical support and funding to make the conference possible. The videos will also be up on YouTube for those who don’t make the conference.

The event we organized at the University of Alberta on Friday was an online panel on What is Responsible in Responsible Artificial Intelligence with Bettina Berendt, Florence Chee, Tugba Yoldas, and Katrina Ingram.

Bettina Berendt looked at what the Canadian approach to responsible AI could be and how it might be short sighted. She talked about a project that, like a translator, lets a person “translate” their writing in whistleblowing situations into prose that won’t identify them. It helps you remove the personal identifiable signal from the text. She then pointed out how this might be responsible, but might also lead to problems.

Florence Chee talked about how responsibility and ethics should be a starting point rather than an afterthought.

Tugba Yoldas talked about how meaningful human control is important to responsible AI and what it takes for there to be control.

Katrina Ingram of Ethically Aligned AI nicely wrapped up the short talks by discussing how she advises organizations that want to weave ethics into their work. She talked about the 4 Cs: Context, Culture, Content, and Commitment.

In Ukraine War, A.I. Begins Ushering In an Age of Killer Robots

Driven by the war with Russia, many Ukrainian companies are working on a major leap forward in the weaponization of consumer technology.

The New York Times has an important story on how, In Ukraine War, A.I. Begins Ushering In an Age of Killer Robots. In short, the existential threat of the overwhelming Russian attack is creating a situation where Ukraine is developing a home-grown autonomous weapons industry that repurposes consumer technologies. Not only are all sorts of countries testing AI powered weapons in Ukraine, the Ukrainians are weaponizing cheap technologies and, in the process, removing a lot of the guardrails.

The pressure to outthink the enemy, along with huge flows of investment, donations and government contracts, has turned Ukraine into a Silicon Valley for autonomous drones and other weaponry.

There isn’t necessarily any “human in the loop” in the cheap systems they are developing. One wonders how the development of this industry will affect other conflicts. Could we see a proliferation of terrorist drone attacks put together following plans circulating on the internet?

The Deepfake Porn of Kids and Celebrities That Gets Millions of Views

It astonishes me that society apparently believes that women and girls should accept becoming the subject of demeaning imagery.

The New York Times has an opinion piece by Nicholas Kristof on deepfake porn, The Deepfake Porn of Kids and Celebrities That Gets Millions of Views. The opinion says what is becoming obvious, that deepfake tools are being used overwhelmingly to create porn of women, whether celebrities, or girls people know. This artificial intelligence technology is not neutral, it is hurtful of a specific group – girls and women.

The article points to some research like a study 2023 State of Deepfakes by Home Security Heroes. Some of the key findings:

The number of deepfake videos is exploding (550% from 2019 to 2023)
98% of the deepfake videos are porn
99% of that porn women subjects
South Korean women singers and actresses are 53% of those targeted

It only takes about half an hour and almost no money to create a 60 second porn video from a single picture of someone. The ease of use and low cost is making these tools and services mainstream so that any yahoo can do it to his neighbour or schoolmate. It shouldn’t be surprising that we are seeing stories about young women being harassed by schoolmates that create and post deepfake porn. See stories here and here.

One might think this would be easy to stop – that the authorities could easily find and prosecute the creators of tools like ClothOff that lets you undress a girl whose photo you have taken. Alas, no. The companies hide behind false fronts. The Guardian has a podcast about trying to track down who owned or ran ClothOff.

What we don’t talk about is the responsibility of some research projects like LAION who have created open datasets for training text-to-image models that include pornographic images. They know their datasets include porn but speculate that this will help researchers.

You can learn more about deepfakes from AI Heelp!!!

OpenAI’s GPT store is already being flooded with AI girlfriend bots

OpenAI’s store rules are already being broken, illustrating that regulating GPTs could be hard to control

From Slashdot I learned about a stroy on how OpenAI’s GPT store is already being flooded with AI girlfriend bots. It isn’t particularly surprising that you can get different girlfriend bots. Nor is it surprising that these would be something you can build in ChatGPT-4. ChatGPT is, afterall, a chatbot. What will be interesting to see is whether these chatbot girlfriends are successful. I would have imagined that men would want pornographic girlfriends and that the market for friends would be more for boyfriends along the lines of what Replika offers.

On Making in the Digital Humanities

On Making in the Digital Humanities fills a gap in our understanding of digital humanities projects and craft by exploring the processes of making as much as the products that arise from it. The volume draws focus to the interwoven layers of human and technological textures that constitute digital humanities scholarship.

On Making in the Digital Humanities is finally out from UCL Press. The book honours the work of John Bradley and those in the digital humanities who share their scholarship through projects. Stéfan Sinclair and I first started work on it years ago and were soon joined by Juliane Nyhan and later Alexandra Ortolja-Baird. It is a pleasure to see it finished.

I co-wrote the Introduction with Nyhan and wrote a final chapter on “If Voyant then Spyral: Remembering Stéfan Sinclair: A discourse on practice in the digital humanities.” Stéfan passed during the editing of this.

All the (open) world’s a stage: how the video game Fallout became a backdrop for live Shakespeare shows

Free to roam through the post-apocalyptic game, one intrepid group has taken to performing the Bard. They have found an intent new audience, as well as the odd mutant scorpion

The Guardian has a nice story today about a Shakespeare troupe who are staging plays in Fallout 76, All the (open) world’s a stage: how the video game Fallout became a backdrop for live Shakespeare shows.

It seems to me that this is related to the various activities that were staged in Second Life and other environments. It also has connections to the Machinima phenomenon where people use 3D environments like games to stage acts that are filmed.

Of course the problem with Fallout 76 is the performers can get attacked during a performance.

How AI image generators work, like DALL-E, Lensa and stable diffusion

Use our simulator to learn how AI generates images from “noise.”

The Washington Post has a nice explainer on how text to image generators work: How AI image generators work, like DALL-E, Lensa and stable diffusion. They let you play with the generator, though you have to stick with the predefined phrases. What I hadn’t realized was the role of static noise in the diffusion model. Not sure how it works, but it seems to train the AI to recognize and then generate in noisy images.