developing valhalla

@hubcityhackers founder
- agent observability
(stealth) - R&D lab teaching LLMs how to hack (and defend) startups

[developing valhalla @valhalla_dev]

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.

[Nous Research @NousResearch]

The Hermes Agent Accelerated Business Hackathon presented by @NVIDIAAI × @stripe × @NousResearch starts now, for builders making agents that can earn, spend, and run real operations at any scale.

Our NVIDIA integrations let your team run agents safely through NemoClaw, quickly on Nemotron 3 Ultra, and intelligently with access to their extensive agent skills. The new Stripe Skills for Hermes let your agent buy what it needs, provision its own SaaS, and pay for the services it uses.

We want to see what kind of business tooling you can build on top of this foundation, whether it’s a fully automated company or a framework to accelerate enterprise functions.

Prizes:
1st — $10,000 cash + NVIDIA DGX Spark + $5,000 Stripe Credits
2nd — $5,000 cash + NVIDIA DGX Spark + $3,000 Stripe Credits
3rd — $2,500 cash + NVIDIA DGX Spark + $1,000 Stripe Credits

To enter:
1) Tweet a 1-3 minute demo video tagging @NousResearch with a short writeup
2) Drop the link in the submissions channel: https://t.co/S16j0bgumq
3) Fill out the submission form: https://t.co/5tQR7mOODF

Judged by Nous Research, NVIDIA, and Stripe on usefulness, viability, and presentation. Submissions due EOD Tuesday, June 30.

[developing valhalla @valhalla_dev]

@m_adams I can count on one hand how many times I've been called "pretty reasonable" so I appreciate it

[developing valhalla @valhalla_dev]

I think there is a degree of "surveillance" that will probably be necessary, at least in the short term, for security, but it does not involve the USG getting carte blanche permission to hack and "patch" my servers.

That surveillance can be extremely benign and concentrated. USG funding for open source tech like PHP, WordPress, React, TypeScript, Rust, Linux, etc. in exchange for very high security standards, bidirectional security reporting (USG passes information to maintainers, maintainers pass information to USG) and mandatory security headcount requirements. We had an agency that could have been part of this with CISA, but guess what happened to CISA?

I also don't particularly have a problem with the USG scanning my infrastructure, one because they already do, as does everyone from black hats to white hat security companies. I don't really even have a problem with getting automated notifications, ideally via opt-in, from the USG on potential security flaws they find. My tax dollars going to a nation-wide, free bug scanning program is pretty reasonable. You don't get to pop a shell on my server to "patch" it, though.

I think moreover what we need is a massive push for a systematized way for OAI, Anth and other LLM providers to discover, fix, report and publish information about these security vulnerabilities. People are freaking out about this but honestly, we need more of this exact thing happening, and judging by @OpenAI 's public statements, there probably will be more of this.

What is probably going to happen, at least for a while, is an extremely chaotic environment of this happening increasingly more often. Look at any bar chart showing CVE assignments over time and you'll see that even before LLM's, we were facing an uptick in found and reported vulns, but now we're facing a massive acceleration. These things won't happen perfectly. Allowing the government to functionally backdoor or attack any domestic or foreign server in order to "patch" it is not and never will be the answer.

[Michael Adams @m_adams]

@valhalla_dev Any ideas on how we'll address the AIcyberpocalypse without total digital surveillance? I don't love where we're headed but the current model is unworkable

[developing valhalla @valhalla_dev]

The only difference between

a remote attacker in China scanning for devices and installing malware on them

and

the USG scanning for devices and "patching" them

is that I actively pay my government to defend me from both scenarios.

[Michael Adams @m_adams]

The PHP incident is more evidence that we’re heading toward a new paradigm for existing on the internet.

It feels like the future will have much more digital surveillance.

[developing valhalla @valhalla_dev]

How about I quote it?

“An extreme representative of this view is Ted Kaczynski, infamously known as the Unabomber. Kaczynski was a child prodigy who enrolled at Harvard at 16. He went on to get a PhD in math and become a professor at UC Berkeley. But you’ve only ever heard of him because of the 17-year terror campaign he waged with pipe bombs against professors, technologists, and businesspeople. In late 1995, the authorities didn’t know who or where the Unabomber was. The biggest clue was a 35,000-word manifesto that Kaczynski had written and anonymously mailed to the press. The FBI asked some prominent newspapers to publish it, hoping for a break in the case. It worked: Kaczynski’s brother recognized his writing style and turned him in. You might expect that writing style to have shown obvious signs of insanity, but the manifesto is eerily cogent. Kaczynski claimed that in order to be happy, every individual “needs to have goals whose attainment requires effort, and needs to succeed in attaining at least some of his goals.” He divided human goals into three groups: 1. Goals that can be satisfied with minimal effort; 2. Goals that can be satisfied with serious effort; and 3. Goals that cannot be satisfied, no matter how much effort one makes. This is the classic trichotomy of the easy, the hard, and the impossible. Kaczynski argued that modern people are depressed because all the world’s hard problems have already been solved. What’s left to do is either easy or impossible, and pursuing those tasks is deeply unsatisfying. What you can do, even a child can do; what you can’t do, even Einstein couldn’t have done. So Kaczynski’s idea was to destroy existing institutions, get rid of all technology, and let people start over and work on hard problems anew. Kaczynski’s methods were crazy, but his loss of faith in the technological frontier is all around us. Consider the trivial but revealing hallmarks of urban hipsterdom: faux vintage photography, the handlebar mustache, and vinyl record players all hark back to an earlier time when people were still optimistic about the future. If everything worth doing has already been done, you may as well feign an allergy to achievement and become a barista.”

This is not "praising the unabomber." This is actively saying "the unabomber did insane shit but parts of his manifesto were indicative of a deeper problem in society that does genuinely exist." You will find people across the political spectrum, to include the far left, agreeing with many parts of the manifesto. This is not even controversial.

Also
> He is actively going around saying how he is not the “anti-Christ” which only brings attention to him being the anti-Christ.

this is a galaxy-brained moron take of literally the exact same strain as you see with right wingers and George Soros. Come on man this is embarrassing

[developing valhalla @valhalla_dev]

Peter Thiel told me that I need to have a controversial take that most people disagree with me on so my controversial take is that Peter Thiel has been right more than he's been wrong and he's turned into the George Soros of the anti-tech left.

[developing valhalla @valhalla_dev]

@YouHitYourHead I read zero to one. That's why I'm quoting it. He also is not promoting the unibomber or... advocating for being the anti-Christ?

This is why I'm saying he's turned into the George Soros of the anti-tech left, ya'll just make up random stuff about him lmao

[developing valhalla @valhalla_dev]

@LowLevelTweets Silence is violence

[developing valhalla @valhalla_dev]

Recursive Self Improvement (RSI) doesn't have to come from a frontier lab, you can just build it into your company now

[developing valhalla @valhalla_dev]

Save that money and buy a DGX Spark and create unlimited games you actually own

[DiscussingFilm @DiscussingFilm]

Xbox is reportedly planning for their next console to release in late 2028.

Microsoft is internally projecting that the console could cost consumers between $1,200 and $1,800, unless they decide to subsidize the cost.

(Source: https://t.co/Qgq8P0UlFe)

[developing valhalla @valhalla_dev]

The average person is seeing rapid degradation of their information environment and don't know about, or care about, the positive technological acceleration.

We win the messaging war against decels when we start showing the average Joe and Jane a better world.

[developing valhalla @valhalla_dev]

@_Stocko_ any CNC machine can be a desktop CNC machine if your desk is massive

[developing valhalla @valhalla_dev]

@pleometric Grok has more neuron activation than the average modal reply guy tbf

[developing valhalla @valhalla_dev]

@sporadica it's extremely hard to avoid the "I am mad about the flavor of the week, too!" behavior

[developing valhalla @valhalla_dev]

@CryptoCyberia Bush had great intonation and delivery but very clunky wording as well

[developing valhalla @valhalla_dev]

As kids we see new tech and it's magical.
As adults we see tech applied in a billion harmful ways, and we forget that awestruck wonder we had as kids.

Tech has mostly improved our society in the net and long term, but if you ask the average person they may not say so. This is a messaging problem.

[developing valhalla @valhalla_dev]

@OpalDesuwa I mean they recycled the Uchiha Clan plot to build the whole story so why not rip from their own plot point as well

[developing valhalla @valhalla_dev]

@sakai_mica @LucusCruz it was just "sasuke uchiha meets sexless night city meets john wick"

[developing valhalla @valhalla_dev]

@gossipaddress @langsort yeeeup. it's bad here

[developing valhalla @valhalla_dev]

Just genuinely disappointed in Edgerunners 2 man I was psyched for this

[developing valhalla @valhalla_dev]

Edgerunners 2 is worse than the first season in basically every respect except for art style. The art/animation is incredible.

In exchange for fantastic aesthetics, they completely neutered night city into a sexless, drugless playground. There's absolutely 0 queerness, which I don't think should be present as a token but should be present to represent the infinite possibilities for gender representation in a cyberpunk-like future.

Also, the story is basically "Sasuke avenges the genocide of the Uchiha clan" and when you take that + the fact that D looks exactly like Sasuke, it's a really thin story.

[🦀 YORA 🦀 @yoracrab]

finished Cyberpunk Edgerunners 2

I LOVED the story, but i never want to see Night City portrayed this sauceless, gritless. sexless again, it felt sanitized to the point of being interchangeable. NC has too much personality to feel like an afterthought 🥀