Notices by DW Innovation (dw_innovation@mastodon.social), page 2

(2/2) The following two articles may help you get a better understanding of Open AI's new models and their output:

Sora’s Surreal Physics (by Gary Marcus):
https://garymarcus.substack.com/p/soras-surreal-physics

Seeing Isn't Always Believing: Video Edition (by Conspirador Norteno)
https://conspirator0.substack.com/p/seeing-isnt-always-believing-video

#GenerativeAI #Verification #Disinformation

(1/2) By now you've probably heard about #Sora and the latest wave of #AI-driven text-to-#video tools. It's all quite spectacular, but far from perfect. If you take a closer look, it's still possible to detect synthetic content. Essential pro tips: Play everything in slow motion. Consider real world physics – and logic.

#Stract is a new, #opensource, #nonprofit #search engine. It has a lot of interesting functions. For example, users can exclude known copycat sites from search queries, focus on indie blogs, or do fediverse search only.

You can test Stract over here:
https://stract.com

And check out the repo on #GitHub:
https://github.com/StractOrg/stract

There's also an interesting article on Stract by @404mediaco:
https://www.404media.co/this-guy-is-building-an-open-source-search-engine-in-real-time

Still relevant: DWA's publication on the #safety and #security of #journalists.

▶️ What are the most promising strategies to support courageous journalists?
▶️ Where are the gaps in existing support mechanisms?
▶️ Who needs special attention, who has not been sufficiently taken into account so far?

Details and downloads (free PDF):
https://akademie.dw.com/en/safety-and-security-of-journalists-results-of-a-global-analysis/a-66850538

(1/2)

"We want the platforms to remove illegal content from the internet. That's important. But if a state abuses this, for example, by declaring any criticism of the government illegal, that is a law violation. Then you can use the same tools to make legitimate criticism disappear on the internet."

Journalists, activists, and media experts have been calling on search giant Google to do something about its technical framework, which frequently...

Did you know that IT/tech-related #risks rank *really high* in the new #WEF Global Risks report?

AI-generated #disinfo is considered to be the 2nd biggest risk (right after extreme weather events). And it's of course connected to the 3rd biggest risk: social/political polarization.

#Cyberattacks are ranked 5th – and can in turn be connected to the disruption of supply chains and critical infrastructure (a topic that also features prominently).

Food for thought.

https://www.weforum.org/publications/global-risks-report-2024/in-full/global-risks-2024-at-a-turning-point/

"Maybe we’ll get the interoperable shared, open social web that many of us have wanted for the last twenty plus years? Wilder things have happened."

Here's a *very* long post (by Tom Coates) on how #Threads / #Meta wants to be a part of the #fediverse:

http://plasticbag.org/archives/2024/01/how-threads-will-integrate-with-the-fediverse/

(You don't have to be a fan of the platform or the company, but one thing's for sure: They're way too big to ignore.)

Re: #AI #policy
(via @techreview)

1) "The US isn’t planning on putting the screws to Big Tech."

2) "It’s not going to be easy to grapple with the harms and risks posed by AI."

3) "AI is the next frontier for techno-nationalism and global competition."

4) "Watch closely what happens in the US election and those around the world."

More: https://www.technologyreview.com/2024/01/08/1086294/four-lessons-from-2023-that-tell-us-where-ai-regulation-is-going/

(1/3) Dear Mastodon community,

we need your expertise and support:

As you may have heard, we're working on several tools to counter #disinfo and facilitate #verification.

One of them is called *Spot*. It's basically an AI-driven natural language interface for geospatial searches in #OpenStreetMap (#OSM).

I.e. you ask the tool about...

The @Bellingcat team recently hosted another hackathon and asked participants to:

- make public open data more accessible and useful

- visualize climate change and its consequences

- understand the visual culture of disinfo

The outcomes of the event are summarized in this post:

https://www.bellingcat.com/resources/2023/11/29/here-are-the-results-of-bellingcats-first-in-person-hackathon/

#osint #journalism #data #dataviz #climatechange #disinfo

In this excellent scrollytelling piece, the FT explains basically everything you need to know about the genesis and concept of generative AI. RNNs, transformers, LLMs, beam search, hallucinations, grounding, RLHF, and the general recognition and prediction of motifs or patterns – it's all there.

https://ig.ft.com/generative-ai/

Dear Mastodon community!

Thanks to the kind mention by @FediFollows we've gained a lot of new followers in the last 24 hours. So maybe it's time for another quick intro.

Our mission statement: https://innovation.dw.com/about

Our team: https://innovation.dw.com/team

Our focus topics: https://innovation.dw.com/topics

Our collection of tools: https://innovation.dw.com/outcomes

And please remember: We're not a newsroom and not DW's management – just an R&D unit checking out #technology for #journalism.

#Doomscrolling. It's a thing.

"People are overwhelmed. They’re tired. Sometimes you want to just curl up in a ball and pull your comforter over your head. (...)

That’s the worst thing you can do for your mental health. It’s important to connect with people, (...) and sometimes you can connect with people and be part of the solution to a problem at the same time."

https://www.wired.com/story/doomscrolling-bad-news-mental-health/

"We gather. We source. We map."

That's the motto of the EU Disinfo Lab.

Have you checked out their web #newsletter yet? The *Disinfo Update* is a "trusted source for curated updates on news, events, and announcements in the #disinformation field from around the globe."

https://www.disinfo.eu/disinfo-update-14-11-2023/

Remember #NFTs? They were heavily criticized by some, and their production consumed a lot of energy. Well, most of the tokens seem to be completely worthless now:

https://dappgambl.com/nfts/dead-nfts/

2100 followers and counting. Wow. Thanks again.

More posts are coming up. And we'll also address your comments and questions asap.

Have a good start into the week!

From our blog archive: A candid conversation on #XR, the #metaverse, and #journalism.

With lots of great insights by Wesa Aapro of Finnish public broadcasting service #Yle.

https://innovation.dw.com/articles/xr-metaverse-journalism-conversation

Mark you calendars:

#AI4MEDIA will host an online #workshop on October 9th, 10h-12h (CEST).

⚡Learn about the consortium's #insights & #whitepapers.

🤝 Join the conversation between #AI R&D people & all kinds of #media practitioners.

More details & free tickets:
https://www.eventbrite.co.uk/e/artificial-intelligence-fit-for-media-sector-a-practitioner-perspective-tickets-711008163557

"Our journalists guarantee our quality. Our commitment to independence, diligence, transparency, respect, diversity of opinion and diversity applies no matter what. At the same time, we are keen to investigate how AI tools can support our journalism."

In a nutshell, this is DW's take on the use of #AI in #news. Details in this commentary by our editor-in-chief Manuela Kasper-Claridge: https://www.dw.com/en/what-is-deutsche-welles-approach-to-generative-ai/a-66868035

More on AI-driven projects & experiments at DW over here: https://innovation.dw.com/topics/ai-automization

👩💻 🤖

Good morning. 600+ followers in less than 24 hours – you are amazing! Please give us some time to answer comments and follow back. Cheers!