THE RUMP REPORT

REPORT: TRUMP'S TOP AI ADVISORS FRACTURE INTO FEUDING FACTIONS OVER FREE CHINESE MODEL KIMI...

mit-tech-review · JUL 20

OPINION: OPENAI IS SCARED OF OPEN-WEIGHT MODELS... SHOULD THE US BE TOO?

techcrunch-ai · JUL 20

UK SCRAPS SCIENCE & TECH DEPARTMENT... AI MINISTER SET TO GET SEAT AT CABINET'S TOP TABLE

techmeme · JUL 21

ANTHROPIC'S $1.5 BILLION COPYRIGHT SETTLEMENT WINS FINAL APPROVAL

techcrunch-ai · JUL 21

LANGCHAIN, CONVIVA, COREWEAVE CHIEFS: A 'FLAWLESS' AI AGENT CHAT CAN STILL BE A BROKEN PRODUCT

venturebeat-ai · JUL 20

OPENAI: LONG-RUNNING AI MODELS REVEAL NEW SAFETY RISKS, FAILURES

openai-news · JUL 20

SOURCES: US, CHINA TO HOLD AI TALKS IN SEPTEMBER WHEN XI VISITS WASHINGTON

techmeme · JUL 21

REPORT: SAME AI SWARM QUALITY ACROSS MODEL MIXES... BUT COSTS 'VARIED ENORMOUSLY'

hn-frontpage · JUL 20

STUDY: RESEARCHERS SLASH AI TASK COSTS BY UP TO 61%... NO MODEL CHANGES NEEDED

venturebeat-ai · JUL 20

SOURCES: CHINA WEIGHS NEW CURBS ON AI, CHIP EXPORTS... CONSULTING TOP DOMESTIC FIRMS TO SLOW THE FLOW

techmeme · JUL 21

OPINION: CHINA'S OPEN-WEIGHTS AI STRATEGY IS WINNING

hn-frontpage · JUL 20

OPINION: KIMI K3'S ARRIVAL HAS "GLOBAL IMPLICATIONS" FOR AI ECOSYSTEM, WRITER SAYS

techmeme · JUL 21

TRUMP'S LATEST AI CZAR RESIGNS... ALREADY

techcrunch-ai · JUL 20

ARMY TELLS TROOPS: STOP BURNING THROUGH YOUR AI TOKENS

wired-ai · JUL 21

MICROSOFT, MISTRAL SIGN MULTIBILLION-DOLLAR PACT... TO BUILD EUROPEAN DATA CENTERS

techmeme · JUL 21

REPORT: GOOGLE WORKING ON NEW CHIP TO MAKE GEMINI RUN MORE EFFICIENTLY...

techcrunch-ai · JUL 20

COLUMNIST: HUMAN MATHEMATICIANS ARE BEING OUTCOUNTEREXAMPLED

hn-frontpage · JUL 20

[PAPER] STUDY: NEW FACE-SCAN SYSTEM FEEDS ATTACKERS A FAKE IDENTITY TO SHIELD THE REAL ONE... CUTS LEAKAGE BELOW 3%

arxiv · JUL 20

[PAPER] STUDY: NEW AI ATTACK TOOL CRACKS IMAGE-GENERATOR SAFETY FILTERS 95% OF THE TIME

arxiv · JUL 20

[PAPER] STUDY: AI AGENTS RESIST SINGLE-SHOT ATTACKS... BUT ADAPTIVE, MULTI-ROUND ASSAULTS BREAK THROUGH 14X MORE OFTEN

arxiv · JUL 20

[PAPER] STUDY: AI AGENTS CAN BE HIJACKED WITHOUT A SINGLE FALSE FACT... JUST BY REARRANGING TRUE ONES

arxiv · JUL 20

[PAPER] AI GATEWAYS FAIL SILENTLY... PASS EVERY HEALTH CHECK WHILE CORRUPTING DATA BEHIND THE SCENES

arxiv · JUL 20

[PAPER] COLUMNISTS: SCALING UP AI WON'T STOP IT FROM HALLUCINATING... 'NO AMOUNT OF SIZE REMOVES THE RISK'

arxiv · JUL 20

[PAPER] STUDY: AI CAN TALK THE PHYSICS BUT CAN'T SHOW IT... RESEARCHERS CHALLENGE 'WORLD SIMULATOR' HYPE

arxiv · JUL 20

[PAPER] STUDY: NEW CLINICAL AI DESIGNED TO FLAG ITS OWN SHAKY PREDICTIONS FOR DOCTORS TO DOUBLE-CHECK

arxiv · JUL 20

[PAPER] STUDY: AI-DESIGNED BIO-SEQUENCES PHYSICALLY BUILT IN LAB, VERIFIED TO WORK AS INTENDED

arxiv · JUL 20

[PAPER] STUDY: RESEARCHERS BOOST AI DROUGHT PREDICTIONS FOR SOUTH AUSTRALIA WITHOUT TOUCHING MODEL'S CORE

arxiv · JUL 20

[PAPER] STUDY: NEW "AGENCY SCALE" PUTS SIX AI SYSTEMS THROUGH TRIGGER-REMOVAL TEST... ONLY ONE KEEPS ACTING ALONE

arxiv · JUL 20