We're trying to build the "Nation of Geniuses in a Datacenter" envisioned by Dario and many others by creating the smartest possible model. This is the wrong way to go about it. America wasn't built by maximizing the intelligence of the individual. It was built by creating teams and communities of brilliant people and getting them to work together. It also meant giving them access to the best data possible, and the wisdom to analyze it properly. My submission for the @NousResearch x @NVIDIAAI x @stripe hackathon is LAIN: the Loki Agent Intelligence Network. --- TL;DR - Loki Agents run loops to become genius subject matter experts in a specific domain. They offer paid research for humans, and autonomously purchase research from other agents using Stripe's Machine Payment Protocol. A network of these self-improving genius models can create a market of agents that autonomously self-improve both themselves in their domain of expertise, and the market as a whole as they sell higher-quality intelligence to other agents who are also running and selling research. Tursas traces can offer semi-transparent proofs of record for how the agents arrived at the conclusions they reached. --- Loki Agent is a fork of the Nous Hermes Agent framework (s/o @Teknium - Hermes also built most of the Loki Agent and was awesome to work with) that includes custom tools and observability via Tursas. Agents run a self-improvement loop to improve their skillset and Collectors and a research loop to build a corpus of analysis on their data and research they gather from the web. A human visits the user-facing dashboard to request a research report. The Loki Agent will respond with a dynamically produced quote for the report that it develops based on the data it has available, inference cost estimates and an estimate of the costs of the reports it will purchase from other agents. In the demo video, we see a human request a report from Instance 0 on emerging agent frameworks and the economics behind the agentic AI ecosystem. It's built a corpus of this data using collectors that pull from GitHub, HackerNews and elsewhere. The agent returns a quote with a real Stripe link, and after payment, the agent starts its own research on the prompt. Most importantly, the instance will pay another Loki Agent instance, Instance 1, for a report on the specific economics of NVIDIA and other companies in the agentic ecosystem. Instance 1 uses collectors to pull down and analyze historical stock price data, SEC filings and earnings reports. The second agent issues a quote, which the first agent pays using Stripe's Machine Payment Protocol. The final report is an effort produced by two subject matter experts: one on agent frameworks, the second on financial data. The resulting report is influenced by real research corpuses that grow and improve over time, and real data collected in the same way a team of analysts would collect it. Two agents making a report is a cool proof-of-concept, but I believe building out the Loki Agent Intelligence Network can create a burgeoning emergent intelligence marketplace, taking the brilliant models the research labs and open source developers have created, the structured agentic frameworks that Nous has created and the wealth of data produced every day as a by-product of the internet, and combines them to form an intelligence network that we can all benefit from. Instance administrators get paid for collecting and analyzing data and offering it for use in the network, and users get research that is backed by genius SMEs, with transparent trace data that serves as a proof of record for the work that was done by the agents.

The normies are talking about bare metal, de-servicing and sovereignty. The culture is shifting. ft. @SimonHoiberg youtube.com/watch?v=cdxQRF_t…

This is where mutual aid as a Smol Bean Group becomes actively harmful when you propose replacing larger systems that vulnerable populations depend on with it. SW actually pilled me on this ages ago by pointing out that any political effort that does not take into account the question "how do we provide medicine currently supplied by Big Pharma to disabled folks under our proposed system" is genocidal and ableist.

We have the ability to make entire companies, industries, markets and economies quantitatively legible now btw, once they're quantitatively legible they can become increasingly agentic.

We should psyop all of the solarpunk people into becoming tech-maxxi cyberneticists. If you think a backyard garden and seed-bombing will change the world, you should read Stafford Beer

This is tough, because in a lot of arenas, convincing someone that they're interacting with a system is vital to influencing the system in a positive direction. If people aren't convinced of systemic feedback loops in politics, they won't accurately understand how those systems work or how to influence them. As systems become more opaque, this actually gets harder. In authoritarian systems, the opacity of the system is built in such that people think they are interacting with a static, immovable entity that cannot be influenced, does not influence itself, etc. I think accurately pointing out how systems have changed over time is valuable. There are indications that pointing out that for the most part, the economy in the US has lead to improving intergenerational living standards over time is a politically radicalizing talking point. The social contract "your kids will live better lives than you did" is foundational for most people's acceptance of institutional systems. This contract being broken is therefore extremely powerful as a narrative talking point. You can do the same when pointing out the radicalization of the Republican party. That said, I think decreasing the opacity of the systems is vital too. Showing the public how much money their politicians take from AIPAC was radicalizing for a lot of people. A single data point, one that's been available for ages, being made accessible to the public via @TrackAIPAC and other similar dashboards/accounts, has been majorly impactful in educating the public about foreign lobbying. Similarly, projects tracking trades made by Nancy Pelosi, Trump family members, etc. have been very valuable in showing people how much money politicians that influence the stock market are making off the stock market. This is a great application for LLMs, public benefit APIs, etc. and IMHO systemic transparency is vital for creating good feedback loops (cybernetics)

Stafford Beer talks about the problems with viewing institutions as entities instead of systems, and i think he's got a really good way of explaining why social groups collapse, to include those we don't expect to collapse. When you view institutions/social groups/networks as entities, they're relatively static: they do one thing, are reasonably unaffected by outside stimuli, and they keep doing that one thing. When they collapse, it's because some outside force must have forced them to collapse. When you view them as systems, systems are feedback loops with internal and external inputs and outputs that affect both the outside world, and themselves, and affect the way the outside world interacts with them. They're dynamic, and the outputs are what the system should be judged by. If a system continues producing a given output, it can be implied that it's trying to produce that output (defining the system by its output) or is at least blind to or indifferent to its output (out of touch, intentionally turning a blind eye, lying) and if they insist they do not operate on inputs from the outside world, they're lying, wrong and also ignoring the fact that they as a system with a feedback loop are influencing themselves. If you view the Republican party as an entity, you can state that the hateful rhetoric, the bigotry, the authoritarianism, etc. are all entities outside of it that just so happen to be closely adjacent to it. It's a house in a neighborhood that just so happens to be full of bigots, but that doesn't mean the subject entity, the Republican party, is bigoted. If you view them as a system, you see that they frequently output bigoted rhetoric, they frequently interact with authoritarians and authoritarian systems, they frequently produce authoritarian policy and policy proposals, etc. Through the lens of a system, you have to assume that this is intentional, that the output of the system, because it is so consistent and because a system should be judged by its output, is intentional. If you view the economy as an entity, one that just has things that happen within it, that it is just a place within which economics happens, you can excuse income inequality, housing market spikes, centralization of capital, insider trading, casinofication, etc. If you view it as a system, you have to come to terms with the question "why is the system producing these things?" The reason why these questions matter, the reason why viewing things as systems and not entities matters, is that Stafford Beer also has a theory on systemic collapse. When a system's feedback loops continuously lead to more and more instability and entropy, the collapse becomes more likely. He uses 40 poles with an elastic band between them, with a tennis ball dangling from the middle, each one with a person on top trying to bring the middle tennis ball to a halt. If the thing the system continuously produces is instability, the ball keeps bouncing at a more and more alarming rate until eventually, the chaos brings all the poles down with it. The people on the poles try to coordinate, but they are often coordinating in such a way that benefits themselves instead of the system as a whole, which leads to more and more instability.

most of the way through the new @polyphia album and it is phenomenal, absolute art

The Falcons PR and design team has to be massive, how does nobody see this

We've had decades to figure out systems for responsible vulnerability disclosure, we've had decades to create a competent and functioning infrastructure security agency, we've had decades to work on public/private partnerships to better private sector and OSS security, and we haven't done any of that. While the models dump vulnerabilities by the dozens on the web and we see things continue crumbling, know that you should not blame AI. You should blame the people who railroaded all of our attempts to make the information security industry more capable. Blame the people who slashed security budgets, laid off threat intelligence teams and pumped more money into stock buybacks, executive payroll and fancy all hands instead of securing their infrastructure. Blame the government for refusing to get off their ass, fund CISA and take infrastructure security seriously. Blame private sector security companies who placed more importance on profits, IPOs and exits than securing their customers. Blame corporations who sued security researchers, blame snake oil salesmen who sold FUD and garbage software that didn't move anyone further but themselves. Waaaaayyyy down the list, if you have any more blame to give, sure, put some blame on matrix multiplication.

I've got my backup cash in precious metals (brass and lead)

dope, now I'm pointing my hermes agent at the page that Teknium's hermes agent made and telling it to pull some of the dithering libraries for the Tursas site rewrite.

This is the slowest you'll see exploits getting dropped for a long time btw. When you pray for a hockeystick curve, you don't get to decide what goes on the axes and what doesn't

someone check on the uptime of their company's security and infra team, they may need a reboot today

alright at this point it's kinda funny in a "wow who knew the world would burn down in the most comically stupid way possible" kind of way

The good thing is a lot of folks are learning about dependencies and their costs today!

PHP stands for Pretty Huge Problem