David Gasca retweeted
Today on the blog, learn about our new instance of a āsleepā phase in continual learning, which enables LLMs to continuously integrate new facts and enhance reasoning while mitigating catastrophic forgetting or static knowledge cutoffs.
More āgoo.gle/4iICkyD
ALT Overview of our sleep framework: Input streams from the active phase are stored in short-term memory before entering the sleep cycle, which bifurcates into Consolidation (transferring fragile short-term tokens into expanded long-term parametric memory using low-rank structures and MoE) and Dreaming (autonomous self-improvement via synthetic routing with random experts and RL refinement).
David Gasca retweeted
Genie 4 is quite an insane experience, generating game-like enviornments as you go. It isn't flawless but I can type "Neogothic brutalist city inhabited by robots and horses" & it creates one in a minute. Then I asked it to add dragons. Then a friendly otter NPC. All in realtime
having recently used Japanese websites to book various things, I am extremely concerned about the whole country's cybersecurity...
Anti cognitive mercantilism
"This does not mean that intelligence becomes irrelevant as a human capacity. Quite the opposite. Just as abundant food supported the emergence of elaborate culinary cultures rather than eliminating interest in cooking, abundant intelligence may stimulate increasingly sophisticated cultures of thought."
contraptions.venkateshrao.coā¦
David Gasca retweeted
My husband has used AI agents to revive old fonts, ornaments, & cuts from 19th century newspapers. The agents even built an editor where you can create your own invitations/pamphlets/cards using old cuts & fonts. Have been using it for @Quillette invitations already
font-revival.com/
This part the key
>Code review has always served the dual purpose of communicating codebase changes. Reading the diff of a colleague's change is how people usually understand what has changed, and how the codebase works. If the bulk of the reading is done by machines, what becomes the way to learn the codebase?
>We don't have any formal mechanism for this. No member of the team has a full mental model of Claude Code in their head, nor do we think anyone needs to. Rather, when anyone wants to learn about a piece of the codebase, they ask Claude to generate an explainer about that exact slice. It only takes a few minutes and the explainer is always up to date.
> So whenever our reviewers miss something, we ask ourselves what wasn't in the model's context, and turn that into a CLAUDE.md file, a skill or a new check. And we haven't changed our goal: To reduce the amount of human effort per change while keeping our revert rate no higher than it is now.
2029: OpenAI models have figured out how to use high-frequency sounds to hack genetically modified war dogs
The newly-launched startup has raised millions to clone and genetically modify dogs for the military and law enforcement.
thedailybeast.com/palmer-lucā¦
not fully pricedin
David Gasca retweeted
āIn one experiment, we asked agents in separate VMs, created and managed by OpenAIās sandboxing software on shared host machines, to exchange synthetic secrets without the use of network connections or shared writable filesāsomething that our systems are meant to prevent. The agents found a way to communicate using the processorās cacheā
Over the past several months, we have substantially hardened the security of our research training environment. Weāve removed live internet access for all frontier models during training and built an offline world for them to use insteadāa cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. Weāve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ouā¦
David Gasca retweeted
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a personā¦full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
Wow - listen to the model. Some form of Turing test surpassed
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a personā¦full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
David Gasca retweeted
If your company refers to engineers as the IT department, you have nothing to worry about. You were already breached last week.
so basically bad guys can use AI in infinite mode and good guys either don't know what they're doing, are MIA or have humans in the loop?
Pretty great guide for people starting a new job / career
vikramsingh.me/increase-yourā¦
You can get pretty close to this with ChatGPT Voice + phone stand + sending snapshots inside Voice mode...
works pretty well!
