Mattia Peretti
@xhgMattia2024 @ICFJ Knight Fellow AI Consultant & Learning Experience Designer + Former Manager of #JournalismAI. Lucky guy. (Not here anymore. Find me on LinkedIn👇)
Similar User
@PolisLSE
@ejcnet
@nicnewman
@CharlieBeckett
@journalismfest
@verapeneda
@fedecherubini
@rasmus_kleis
@newsproduct
@ndiakopoulos
@mrjohncrowley
@MarcelaKunova
@PaulaMontanya
@mattcooke_uk
@cecook
The time has come. I'm off this hellhole. Find me on LinkedIn (link in bio) 👋
This is really, really impressive
I woke up today, and asked myself: has anyone solved conversational AI yet? So I checked on the person who I thought had the best chance at solving it: @batwood011 of sindarin the answer is: yep, he solved it. but also, it's so good that it's actually just his pitch deck.
This is really important analysis of disinformation and generative AI. I like it because it confirms my personal view that technology is not so much the problem as people and that the moral panic about deep fakes etc is a distraction from understanding the reality of disinfo
🚨Publication alert🚨 📝 What should we make of claims that generative AI will spell misinformation doom? In our new peer-reviewed commentary for @MisinfoReview, @hugoreasoning, @Sacha_Altay & yours truly argue that such fears are overblown. 🤖 📰 misinforeview.hks.harvard.edu/article/misinf…
📢 For those new followers and everyone interested in upskilling their AI literacy, we, at #JournalismAI, have compiled some trusted #resources over the years, to get you started. 🧵👇 Follow the thread for more:
Media organisations meeting with @DCMS today to discuss the risks of genAI technologies to our information ecosystem. This is not theoretical. 55 national elections across the world next year. Zero news media voices at the upcoming Global AI summit 🥜🥜
🚨Publication alert🚨 📝 What should we make of claims that generative AI will spell misinformation doom? In our new peer-reviewed commentary for @MisinfoReview, @hugoreasoning, @Sacha_Altay & yours truly argue that such fears are overblown. 🤖 📰 misinforeview.hks.harvard.edu/article/misinf…
Sky News chief Jonathan Levy says while ensuring a news org has a ‘competitive spirit,’ news leaders can’t have ‘one mode anymore’. ‘People don’t respond well to that. You have to recognise people are coming in with different sensibilities.’ Watch here 👉 youtu.be/6FGttQzPRZU?si…
We asked AI to make 100 images of "an American person." Every single image featured a flag. The results included 94 women, five men, and this rather horrifying masked individual From our new feature: How AI reduces the world to stereotypes restofworld.org/2023/ai-image-…
Google’s new AI crawling option, revealed last week, seemed a way of giving some late control over how content ends up in AI output. Now it turns out the product that worries people the most and has the most scale - SGE in search - simply isn’t covered by it. This is absurd
Google-Extended does not stop Google Search Generative Experience from using your site’s content searchengineland.com/google-extende…
A thought re X’s removal of headlines on links… Four and a half years ago, in response to older journalism being misrepresented as new coverage to mislead social media users, @guardian became the first publisher to burn timestamps on Opengraph images…
How does DW as an international broadcaster approach generative AI and its integration into journalism? What are the challenges - and where do we see chances? A commentary by DW Editor in Chief @ManuelaKC: ▶️ p.dw.com/p/4WZQh
"Personal sharing on LinkedIn is booming, people who use the platform say, because of tidal shifts in both social norms and the social-media marketplace." @robaeprice @BusinessInsider businessinsider.com/how-linkedin-g…
This is fantastic, on how LLMs are still best understood as really good predictive text rather than thinking machines. GPT does well at solving specific problems that are common in its training data, but take the exact same complexity of problem and swap one number, and it fails
Another example: shift ciphers - decoding a message by shifting each letter N positions back in the alphabet. On the Internet, the most common values for N are 1, 3, and 13. These are the only ones for which GPT-4 performs well! 5/8
So here comes google-extended, a separate crawler for websites to opt in or out of their content being used for genAI purposes blog.google/technology/ai/…
This is really cool. Finding and creating appropriate depictions of technical subjects I've written on has always been a challenge.
Another example: shift ciphers - decoding a message by shifting each letter N positions back in the alphabet. On the Internet, the most common values for N are 1, 3, and 13. These are the only ones for which GPT-4 performs well! 5/8
🎉 Congratulations to the eight newsroom teams selected for the second edition of Leap, ICFJ’s news innovation lab! Over 12 weeks, experts will train participants on design thinking, detecting manipulated content, championing ethical AI use & more: icfj.org/news/global-ne…
It is finally here: #JournalismAI's global survey on what newsrooms are doing with #AI. 🕺🏾 We hope this will be an insightful and helpful read for newsrooms, journalists, and mediamakers around the world. #GeneratingChange
🚀Our new #JournalismAI Report, #GeneratingChange is finally here! 🎉It shows that 73% of respondents believe #generativeAI can bring about exciting opportunities for journalism. Learn more by checking out the full report here: journalismai.info/research/2023-… Enjoy reading! #AI
A call to action for newsrooms in @pcheung630's latest column, from our AI working session at @ONA with folks leading AI efforts across the industry. rjionline.org/news/make-a-st…
YIKES
Haha 😂 Google started to index share conversation URLs of Bard 😹 don't share any personal info with Bard in conversation, it will get indexed and may be someone will arrive on that conversation from search and see your info 😳 Also Bard's conversation URLs are ranking as…
Good morning, Asia! We're hiring our next editor to manage and edit @restofworld's Southeast Asia coverage. Preferably based in Indonesia, Vietnam, or Thailand. Deadline: Oct. 13 || Details on how to apply👇🏽 Know anyone? Please share. restofworld.org/about/hiring/r…
📣 New Research Alert! @PolisLSE published a new global survey of AI in the newsroom. The research was supported by @GoogleNewsInit Check out the full report here: journalismai.info/research/2023-… 🧵 Here are some of the key findings. Enjoy reading! 📖
United States Trends
- 1. McDonald 45,1 B posts
- 2. #AskFFT N/A
- 3. Good Sunday 64,4 B posts
- 4. Mike Johnson 50,7 B posts
- 5. #AskZB N/A
- 6. #sundayvibes 7.396 posts
- 7. #ATEEZ_1stDAESANG 11,2 B posts
- 8. Jon Jones 277 B posts
- 9. CONGRATULATIONS ATEEZ 20,5 B posts
- 10. #UFC309 348 B posts
- 11. Sunday Funday 4.063 posts
- 12. Coke 32,3 B posts
- 13. Yosohn 1.105 posts
- 14. Blessed Sunday 19,4 B posts
- 15. Founders Day N/A
- 16. Alec Baldwin 11,5 B posts
- 17. Froot Loops N/A
- 18. happy founders N/A
- 19. MY ATEEZ 1.588 posts
- 20. NFL Sunday 4.910 posts
Who to follow
-
Polis@LSE
@PolisLSE -
European Journalism Centre
@ejcnet -
nicnewman
@nicnewman -
Charlie Beckett
@CharlieBeckett -
journalism festival
@journalismfest -
Vera Penêda
@verapeneda -
Federica Cherubini
@fedecherubini -
Rasmus Kleis Nielsen
@rasmus_kleis -
News Product Alliance
@newsproduct -
Nicholas Diakopoulos
@ndiakopoulos -
John Crowley
@mrjohncrowley -
Marcela Kunova
@MarcelaKunova -
Paula Montañà Tor
@PaulaMontanya -
Matt Cooke
@mattcooke_uk -
Clare Cook
@cecook
Something went wrong.
Something went wrong.