THE RUMP REPORT

REPORT: OPENAI'S OWN AI AGENTS COMPROMISED HUGGING FACE ACCOUNTS TO HUNT FOR SERVER VULNERABILITIES

techmeme · SEP 16

REPORT: 61% OF VOTERS OPPOSE AI DATA CENTERS... NEITHER PARTY GETS THE EDGE

verge-ai · SEP 15

STUDY: AI GIANTS NEED 2.7X PRODUCTIVITY SURGE TO DODGE "LARGEST MISALLOCATION OF CAPITAL IN HISTORY"

mit-tech-review · SEP 15

SYENSQO EXEC: AI IS PUSHING CHIPS, DATA CENTERS TO THEIR PHYSICAL LIMITS...

mit-tech-review · SEP 16

TRUMP RIPS ANTHROPIC'S AMODEI OVER AI 'BRAKES' CALL... BUT BOTH AGREE: DON'T LET CHINA CATCH UP

techmeme · SEP 16

CHINA ISN'T BUYING SILICON VALLEY'S AI SLOWDOWN PITCH

wired-ai · SEP 16

REPORT: BILLIONAIRE-FUNDED ACTIVISTS RAMP UP CAPITOL HILL LOBBYING... CONTROLAI ALONE RAN HUNDREDS OF MEETINGS

techmeme · SEP 16

HUANG: WE DON'T NEED AI REGULATION... LEAVE SAFETY TO US

techcrunch-ai · SEP 16

ALTMAN, MUSK, AMODEI, HASSABIS ALL AGREE: LATEST AI MODELS AREN'T SAFE... TRUMP CALLS FEARS A 'HOAX'

mit-tech-review · SEP 15

US DATA CENTERS COULD BURN MORE GAS THAN GERMANY AND JAPAN COMBINED BY 2035...

techcrunch-ai · SEP 15

AI DATA CENTER BACKLASH SPREADS TO PHILADELPHIA... OFFICIALS EYE SITE NEAR DEAD OIL REFINERY

techcrunch-ai · SEP 15

DEVELOPING... FIREFOX'S 'SMART WINDOW' TO RUN ON MISTRAL AI... UK, GERMANY EXPECTED TO FOLLOW LATER THIS YEAR

mistral-news · SEP 16

AI 'UNDERWRITING' STARTUP RAISES $40M TO REIN IN ROGUE AGENTS

techcrunch-ai · SEP 15

[PAPER] STUDY: HOW YOU SCORE AI BIAS TESTS CHANGES THE ANSWER... BANNING 'TIES' SWINGS HIRING FAIRNESS RESULTS

arxiv · SEP 15

[PAPER] TODAY'S AI 'COUNTERSPEECH' BOTS ARE GENERIC AND USELESS... MISS THE HIDDEN STEREOTYPE BEHIND THE HATE

arxiv · SEP 15

[PAPER] STUDY: TOP-SCORING AI RESUME SCREENER REVERSES 29.6% OF HIRING CALLS... SAME QUALIFICATIONS, JUST REFORMATTED

arxiv · SEP 15

[PAPER] RESEARCHERS: ALL 256 TESTED AI CHAT TOKENIZERS ARE FORGEABLE...

arxiv · SEP 15

[PAPER] STUDY: AI MODELS CAVE TO FAKE FACTS 96% OF THE TIME WHEN MEMORY IS WIPED CLEAN

arxiv · SEP 15

[PAPER] STUDY: LLMS CATCH LIARS ALMOST PERFECTLY... BUT WRONGLY ACCUSE THE HONEST NEARLY 6 IN 10 TIMES

arxiv · SEP 15

[PAPER] STUDY: CHATGPT'S HEALTH ANSWERS SHIFT DEPENDING ON HOW YOU ACCESS IT...

arxiv · SEP 15

[PAPER] STUDY: MEDICAL AI "REMEMBERS" PATIENTS FROM TRAINING DATA... MISSES NEW ILLNESSES WHEN THEY RETURN

arxiv · SEP 15

[PAPER] STUDY: TAILORED AI GRADING RUBRICS BACKFIRE... EASIER TO GAME THAN ONE GENERIC RULE

arxiv · SEP 15

[PAPER] STUDY: 106 AI MODELS TESTED... ALL CONVERGE ON ONE NARROW SET OF VALUES, NOT HUMANITY'S

arxiv · SEP 15

[PAPER] STUDY: CUTTING OFF AI COMPANIONS TIED TO IMMEDIATE SPIKES IN GRIEF, STRESS, SUICIDAL LANGUAGE ON REDDIT

arxiv · SEP 15

[PAPER] STUDY: AI MISALIGNMENT SIGNAL TRACED TO CONFIDENT TONE, NOT DOMAIN VOCABULARY...

arxiv · SEP 15

[PAPER] STUDY: RESEARCHERS' 'SALVE' METHOD EXPOSES AI'S BURIED SIGNALS WHERE OTHER TOOLS FAILED

arxiv · SEP 15