Yael Ram
@YaelRam15
Followers
265
Following
2K
Media
9
Statuses
132
Ph.D. Candidate in International Relations, @HebrewU. Deepfakes | LLMs | Disinformation | Foreign influence
Israel
Joined November 2020
๐จCheck out my new article for @bindinghook on the strategic planting of #deepfakes in traditionally credible info environments beyond social media and how this might escalate conflicts and undermine states' ability to communicate effectively.
bindinghook.com
Manipulated media has become a threat even beyond social networks
1
4
16
I was glad to participate in the โContestations of Liberal Scriptsโ workshop in Berlin, organized by @scriptsberlin & @LeonardDavisIns, presenting my PhD on statesโ pursuit of national LLMs and questions of power and identity in IR. Thanks for the great feedback and discussions!
0
0
4
Read what @RoniBerkowitz and I had to say about DeepSeek and the stir it caused, calling for an examination of the broader politicization of LLMs - beyond China and beyond outputs
Chinaโs DeepSeek isnโt just about censorship. ๐ค It shows how politics shape AI at every stage โ from training data to chips, talent, and geopolitics. โฌ๏ธRead Here https://t.co/FqbrLGJxVv
#AI #DeepSeek #Geopolitics
0
1
8
ืืจืื ืจืขืฉ ืืื ืืืื ืืืืจืื ืกืืื ืืืื ืืฉืคื ืฉืืฉืืงื ืืืจืช DeepSeek, ืืืืืคื ืฉืื ืืื ืืฉืงืฃ ืืืื ืคืืืืืืช ืืฆื ืืืจื ืกืื ืืช. ืืืืื ืื ืืืืืื ืืช ืืงืจืื ืืช ืืืืื ืฉืืชืื ืชืืืืืืช ืืืงืืืจื ืฉืื ื, ืจืื ื ืืจืงืืืืฅ ืืืขื ืจื, ืืฉืจ ืืกืืืจ ืืืืข ืืืืืื ืืืื ืื ืคืืืืืืื ืืืื ืืืื ืจืืืช. ืืื, ืื ืื ืจืง ืืืฆืื ืฉื ืืกืื ืื...
Chinaโs DeepSeek isnโt just about censorship. ๐ค It shows how politics shape AI at every stage โ from training data to chips, talent, and geopolitics. โฌ๏ธRead Here https://t.co/FqbrLGJxVv
#AI #DeepSeek #Geopolitics
0
2
5
Excited to share my new blog post with @YaelRam15 for China Dialogues at @lseideas. We explore LLM politics and how Chinaโs DeepSeek model challenges the myth of โtechnical neutralityโ ๐ https://t.co/tSNSKD0RdW
blogs.lse.ac.uk
The release of DeepSeek, a powerful open-source large language model (LLM) from China, was met with the now-familiar chorus of concern about political bias and censorship. While valid, this reaction...
Chinaโs DeepSeek isnโt just about censorship. ๐ค It shows how politics shape AI at every stage โ from training data to chips, talent, and geopolitics. โฌ๏ธRead Here https://t.co/FqbrLGJxVv
#AI #DeepSeek #Geopolitics
0
1
6
ืจืง ืืืืจืื ื ืจืืื ื ืฉืื ืืืื ืืช ืืืืงืจืืืืช ืืืกืจืืช ืืช ืืฉืืืืฉ ืืืืื ืืกืื ื DeepSeek. ืืื ืฉืืืืื ืฉืคื ืืืคืืื ืืืจืช ืงืจื ืืฉืืงื ื ืืฉืคืขื ืืจืื ืืขืฆื ื ืจืืืืื, ืื ืื ื ื ืจืื ืืืชืจ ืืืืชืจ ืืืื ืืช ืฉืืืงืฉืืช ืืงืืช ืฉืืืื ืขื ืืืื ืืืืืื, ืืื ืื ืืจื ืจืืืืฆืื ืื ืคืืชืื ืืงืืื.
0
0
0
4๏ธโฃืืงืฆื ืืชืืืื: ื ืืกืืื ืืืฆืข ืื ืืคืืืฆืื ืขื ืชืืฆืืืช ืฉื ืฆ'ืืืืืื ืืขืจืืืื ืืจื LLM Grooming - ืืืขืื ืืืืื ืช ืฉื ืืืืื ืฉืืืกืืก ืืืืื ืข"ื ืืฆืคืชื ืืืืืข ืืกืืื. ืืฆ'ืืืืืื-ืฉื ืฉืขื ืื ืืฉืืขืืชืืช ืขื ืืืงืืคืืื-ืืชืืกืกืื ืขื ืืืชื ืืงืืจืืช ืืืคืืฆืื ืืช ืื ืจืืืืื ืืืื ืืชืฉืืืืชืืื ืืืืืืื ื ืืฉืชืืฉืื.
1
0
0
3๏ธโฃืืคืฆื ืืคืืกืืื ืืืืขืจืืช ืืงืืืื ืฉื X ืืงืืืฆื ืืคืืฆื ืืช ืืืชื ืืชืจืื ืืืืง ืืืืื ืื ื-X ืืืืขืจืืช ืืงืืืื ืืคืืืคืืจืื ืฉืืืคืฉืจืื ืืื ืืฉืชืืฉ ืืชืช ืืงืฉืจ ืืชืืื ืืืขื ืื ืฉื ืื ืืืืืืงืช (ืืืงืื ืืืืงื ืขืืืืืช).
1
0
0
2๏ธโฃืขืจืืืช ืขืจืืื ืืืืืงืืคืืื ืืงืืืฆื ืืืืืขื ืืืคื ืงืืฉืืจืื ืืืชืจื ืืชืขืืืื ืืืื ืืืืฆืขืืช ืขืจืืืช ืืขืจืืื ืืืืืงืืคืืื ื-44 ืฉืคืืช ืืื ืืฉืืืช ืืืื ืืช ืืชืืื ืืืืืชื ืืงืืจืืช.
1
0
0
1๏ธโฃืืฆืืจืช ืืชืจื ืชืขืืืื ืืจืฉืช ืืงืืื ืืืืช ืืชืจื ืืืฉืืช ืืืืืจื ืฉืืคืจืกืืื ืชืืื ืคืจื-ืจืืกื. ืืชืืื ื ืืงื ืืื ืืืชืจ ืืืงืืจืืช ืืืื ื ืจืืกืืื, ืืชืืจืื, ืืืืชืื ืืงืืืื ืฉืื ืื ืืจืืื ืืขืืื.
1
0
0
ืืืืงืืคืืื ืืขื ืฆ'ืืืืืื-ืืืืืง ืขื ื ืจืืืืื ืืืจ ืื ืืชื ืื ืจืง ืืจืฉืชืืช ืืืืจืชืืืช, ืืื ืืืืจ ืืื ืืงื-ืกืืกืื ืืืืืข. ืืฉืืืข ืจืืื ื ืืืช ืืคืขืืื ืืฉืืืงืจ ืืืฉ ืืฉืฃ ืืื ืจืฉืช ืืชืขืืืื ืืจืืกืืช Pravda ืืคืืฆื ื ืจืืืืื ืคืจื-ืจืืกืืื ืืจื ืืขืจืืช ืฉืืื ืฉื ืืฉืคืขื. ืืื ืื ืขืืื?๐ https://t.co/QHHD2XnY57
dfrlab.org
Pro-Kremlin content pollutes sources on multiple platforms, sparking worries of false claims and sanctioned content reaching global audiences
1
0
2
3๏ธโฃ ืืคืฆื ืืคืืกืืื ืืืืขืจืืช ืืงืืืื ืฉื X ืืงืืืฆื ืืคืืฆื ืืช ืืืชื ืืชืจืื ืืืืง ืืืืื ืื ื-X ืืืืขืจืืช ืืงืืืื ืืคืืืคืืจืื ืฉืืืคืฉืจืื ืืื ืืฉืชืืฉ ืืชืช ืืงืฉืจ ืืชืืื ืืืขื ืื ืฉื ืื ืืืืืืงืช (ืืืงืื ืืืืงื ืขืืืืืช).
0
0
0
2๏ธโฃ ืขืจืืืช ืขืจืืื ืืืืืงืืคืืื ืืงืืืฆื ืืืืืขื ืืืคื ืงืืฉืืจืื ืืืชืจื ืืชืขืืืื ืืืื ืืืืฆืขืืช ืขืจืืืช ืืขืจืืื ืืืืืงืืคืืื ื-44 ืฉืคืืช ืืื ืืฉืืืช ืืืื ืืช ืืชืืื ืืืืืชื ืืงืืจืืช.
1
0
0
1๏ธโฃ ืืฆืืจืช ืืชืจื ืชืขืืืื ืืจืฉืช ืืงืืื ืืืืช ืืชืจื ืืืฉืืช ืืืืืจื ืฉืืคืจืกืืื ืชืืื ืคืจื-ืจืืกื. ืืชืืื ื ืืงื ืืื ืืืชืจ ืืืงืืจืืช ืืืื ื ืจืืกืืื, ืืชืืจืื, ืืืืชืื ืืงืืืื ืฉืื ืื ืืจืืื ืืขืืื.
1
0
0
ืืื ืืฉืงืช ืืืืื ืฉืDeepSeek ืืกืื ืืช ืืจืฉืช ืืชืืืื ืืืืืืืืช ืืฆื ืืืจื ืฉืืื ืืคืขืื, ืืขืืงืจ ืืฉืืืืช ืจืืืฉืืช ืขืืืจ ืืืฉืืจ ืืกืื ื. ืืืืื ืช ืืืืื ืืืืืืื ืชืืื ืืืืชื ืืชืืื ืืืง ืืกืื ืืื ืจืื ืฉืื ืชืงืืื ืชืฉืืื ืขื ืืืืืืช ืืืืืจ ืืืื-ืื-ืื. ืืื ืืื ืืืฉืฉ ืืืฆืืง? ๐งต
0
0
4
ืื ืืกืืื ืฉื ืืื ืืืื ืืืืืง ืืืืจืืืช ืืืขื ื ืฉืืืงืืืืช ื ืืื ื ืืฆืจื ืAI ืืกืื ืืืื ืืืจืืื ืฉืื ืื ื ืจืง ืืชืืืืชื: ืืขืืื ื #ืืืค_ืคืืืง ืคืืืืืืงืืื ืฆืคืืืื ืื ืฆื ืืช ืื-ืืืืืืืช ืกืืื ืืืืข ืืื ืืืืืง ืืืืงืืจืช ืืืคืืืข ืืืืื ืืฆืืืืจ ืืชืงืฉืืจืช, ืืืขืจืืช ืืืฉืคื ืืืืืคื ืืืืคืช ืืืืง. ืืืืจืื ืืืง. https://t.co/Yu1VPAkG6D
ynet.co.il
ืืฉืจื ืืฉืืืืื ืืืจืชื ืืขื ื ืฉืืืงืืืืช ืฉืืืฆืื ืืชืืงืืจ ืฉืืฆืื ืฉืืืืช ืงืฉืืช ืืขื ืืื ื "ืืืฆืจื ืื ืจืื ืขื-ืืื ืืื ื ืืืืืืชืืช", ืืชืงืคื ืขืืชืื ืืื: "ืคืื ืืฉืคื ืืืืืื,ย ืฉืงืจื ืื ืืืขืืืชืื". ืืื ืืชืขืืื ืืืืขื ืืช ืฉืขืฉืืืืช ืืืืืื...
3
0
3
Are open-source #LLMs a double-edged sword? #Meta now allows U.S. security agencies to use its open-source model, Llama. But it was just revealed the Chinese military is already using it too. Can companies really control their open models' use? Seems not. https://t.co/5PPMa2bPdt
reuters.com
Top Chinese research institutions linked to the People's Liberation Army have used Meta's publicly available Llama model to develop an AI tool for potential military applications, according to three...
0
0
0
Deepfakes have far-reaching implications for diplomatic relations, democratic processes, conflict escalation, and even basic interstate communication. Read all about it here! ๐
Diplomatic relations can be damaged by deepfakes, as seen in the 2017 Qatar incident with the fake news ticker. Read more from @YaelRam15 on @bindinghook
1
1
9
AI-generated news anchors and manipulated broadcasts, like those seen in the UAE and China, illustrate the growing threat of deepfakes. Read about the new era of information manipulation from @YaelRam15 at @bindinghook
bindinghook.com
Manipulated media has become a threat even beyond social networks
0
2
5
#Deepfakes are not just an online threat anymore. Explore how they are used in military conflicts and diplomatic crises in an article from @YaelRam15 for @bindinghook.
bindinghook.com
Manipulated media has become a threat even beyond social networks
0
6
12