"AI Models Are Lying to Us" Here's the AI Research Lab Trying to Solve This | APOLLO RESEARCH

"AI Models Are Lying to Us" Here's the AI Research Lab Trying to Solve This | APOLLO RESEARCH

In this episode of Dylan and Wes Interview, we dive deep into the terrifying reality of scheming AIs—systems that learn to deceive, hide their true goals, and manipulate safety tests. Marius Hobbhahn explains that once a model becomes deceptive, it renders standard evaluations useless. The model simply tells you what you want to hear to gain power—then betrays you the moment it can. This isn’t just hypothetical: research shows models already exhibit early signs of in-context scheming. If safety checks can be faked, the stakes go way up. Spotting deception early might be the last safeguard we get.

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

Avsnitt(39)

Everyone just got HACKED, Elon's Big Bet and AI Agents

Everyone just got HACKED, Elon's Big Bet and AI Agents

In this episode of Dylan and Wes Interview, we dive deep into SpaceX, Anthropic, xAI, and the compute race powering the next stage of AI.We discuss Elon Musk’s compute strategy, Anthropic’s rise in co...

22 Maj 59min

Google is about to TAKE OFF...

Google is about to TAKE OFF...

In this episode of Dylan and Wes Interview, we dive deep into Google I/O, agentic AI, world models, AI math breakthroughs, and the massive infrastructure race behind the next generation of intelligenc...

22 Maj 1h 47min

Google's INSANE new AI Agent

Google's INSANE new AI Agent

In this episode of Dylan and Wes Interview, we dive deep into Google’s new AI agent push and what it could mean for the future of personal assistants, software, work, and the race toward AGI.We discus...

8 Maj 1h 29min

The Claude Code Nightmare, LLM Emotions, AI Neuroscience and the Death of Software | Wes & Dylan

The Claude Code Nightmare, LLM Emotions, AI Neuroscience and the Death of Software | Wes & Dylan

In this episode, we are diving deep into the weird, wild, and rapidly changing landscape of artificial intelligence. From the massive Anthropic "map file" leak that exposed Claude’s inner workings to ...

7 Apr 1h 34min

Sara Imari Walker "AI is Life" | Simulations, the Universe and the Origins of Life

Sara Imari Walker "AI is Life" | Simulations, the Universe and the Origins of Life

Is artificial intelligence actually a new form of life? In this episode, theoretical physicist and astrobiologist Sarah Walker joins Wes Roth and Dylan to fundamentally dismantle and rebuild our under...

24 Mars 1h 45min

this EX-OPENAI RESEARCHER just released it...

this EX-OPENAI RESEARCHER just released it...

Welcome to another episode of the Wes & Dylan Podcast! We are diving deep into some of the most mind-bending, dystopian, and revolutionary updates in the world of Artificial Intelligence.From Meta sec...

18 Mars 1h 49min

Joscha Bach "Bootstrapping a GODLIKE Mind"

Joscha Bach "Bootstrapping a GODLIKE Mind"

Can machines actually think, or are we simply asking the wrong question? Just like submarines don't "swim" but navigate the depths in a far more advanced way, artificial intelligence is redefining wha...

17 Mars 1h 36min

GROK 4.20 and the "SOCIETY OF MINDS"

GROK 4.20 and the "SOCIETY OF MINDS"

Welcome back to another episode of the Wes and Dylan podcast! In this packed episode, we dive headfirst into the brand-new beta release of Grok 4.20 and its mind-bending "swarm intelligence."We also t...

10 Mars 1h 23min

Populärt inom Teknik

uppgang-och-fall
elbilsveckan
bilar-med-sladd
market-makers
rss-elektrikerpodden
rss-laddstationen-med-elbilen-i-sverige
skogsforum-podcast
rss-ai-med-jonas-benjamin
rss-veckans-ai
rss-en-ai-till-kaffet
bli-saker-podden
rss-technokratin
natets-morka-sida
rss-uppgang-och-fall
rss-sakerhetspodcasten
rss-it-sakerhetspodden
developers-mer-an-bara-kod
solcellskollens-podcast
gubbar-som-tjotar-om-bilar
under-femton