Sara Hooker on compute thresholds for AI; CrowdStrike breaks the internet
Safe Mode Podcast25 Juli 2024

Sara Hooker on compute thresholds for AI; CrowdStrike breaks the internet

AI governance regimes the world over have seized on compute thresholds as a mechanism for implementing various controls on artificial intelligence systems. The basic idea is that if an AI model relies on a sufficiently large amount of computing power, then various controls kick in. As AI models get larger, the thinking goes, they also get riskier, and this means that AI governance regimes should focus on the largest models, as measured by computing power. But does this idea make any sense as a governing tool for the models of today and tomorrow? Sara Hooker leads Cohere’s research operation, and she’s looked hard at whether compute thresholds can be applied to AI systems to mitigate risks. On this episode of Safe Mode, she sits down with host Elias Groll to discuss her research on compute thresholds. CyberScoop’s newly minted editor in chief, Greg Otto, also joins the show to discuss how an errant CrowdStrike software update took down a huge number of critical services across the internet. Links: On the Limits of Compute Thresholds as a Governance Strategy | arXiv CrowdStrike Falcon flaw sends Windows computers into chaos worldwide | CyberScoop

Det här avsnittet är hämtat från ett öppet RSS-flöde och publiceras inte av Podme. Det kan innehålla reklam.

Avsnitt(139)

Inside Operation Disruption Week: Taking Down Southeast Asia's Scam Machine

Inside Operation Disruption Week: Taking Down Southeast Asia's Scam Machine

What does it actually take to dismantle an industrial-scale scam operation running bulletproof hosting, distributed ASNs, and crypto laundering across multiple countries? Mike Sweeney of Silent Push w...

25 Juni 35min

Zero days, zero order: The chaos reshaping vulnerability disclosure

Zero days, zero order: The chaos reshaping vulnerability disclosure

The rules of responsible disclosure were written for a different era — one where humans found bugs, humans reported them, and 90 days felt like plenty of time to patch. That era is over. In this epis...

18 Juni 40min

Why the autonomous SOC Is the wrong goal

Why the autonomous SOC Is the wrong goal

On this week's episode, we're joined by Mike Nichols, General Manager of Security at Elastic, fresh off the Gartner Security and Risk Summit in the D.C. area, where AI dominated every conversation on ...

11 Juni 33min

The last layer standing

The last layer standing

What happens when an "assume breach" scenario turns into a total corporate wipeout? In this episode of Safe Mode, host Greg welcomes Brandon Willitts, Director of Cyber Resilience at Everpure, to pull...

4 Juni 35min

From Two Weeks to Three Days: The KEV Deadline Debate

From Two Weeks to Three Days: The KEV Deadline Debate

Drawing on his experience from his time in government working directly on CISA’s Known Exploited Vulnerabilities (KEV) catalog, Todd Beardsley, VP of Security Research at runZero, explains what it act...

29 Maj 37min

Can specialized security survive Daybreak and Mythos?

Can specialized security survive Daybreak and Mythos?

In this episode, we sit down with Lior Div, CEO of 7AI, at a moment when the ground is shifting under the entire security industry. With AI lowering the barrier to entry for attackers, supply chain co...

21 Maj 38min

Why access brokers have stubbornly remained successful

Why access brokers have stubbornly remained successful

Anna Pham of Huntress joins Safe Mode to discuss the current landscape of initial access brokers and how their tactics continue to support ransomware operations. She explains that attackers are still ...

14 Maj 31min

Can you prove which agent did what?

Can you prove which agent did what?

In this week's episode, Greg Otto talks with Howard Ting, CEO of Opal Security, about the growing security challenges created by AI agents inside the enterprise, especially around identity governance,...

7 Maj 28min

Populärt inom Politik & nyheter

svenska-fall
aftonbladet-krim
p3-krim
spar
motiv
de-fyras-gang
tv4-nyheterna-story
rss-expressen-dok
flashback-forever
aftonbladet-daily
rss-sanning-konsekvens
rss-vad-fan-hande
rss-krimreportrarna
svd-ledarredaktionen
rss-flodet
rss-frandfors-horna
kungligt
svd-dokumentara-berattelser-2
grans
dagens-eko