The case for RAMP - A visual summary

How the RoboNet Artificial Media Protocol can deliver global enforcement for AI

I wasn’t expecting to turn this into a series but really new solutions apparently need to be translated into many narratives nowadays. Maybe this is an effect of our generation information overload, or the fact that everybody wants the microphone for their own interests and everything else is background noise, we all are lost in many translations. What I do know is that for whatever reason, more AI experts have misunderstood or literally overlooked RAMP’s core propositions than I would’ve expected, so I will not waste anyone’s time with even more words: this post will just visually stack and compare AI regulatory governance initiatives against RAMP, a visual executive summary if you will.

The executive oneliner:

RAMP is a global, frictionless, apolitical compliance instrument for AI that uses an Internet protocol to deliver technical and normative resources for the enforcement of any AI law, policy or regulation, no matter the political jurisdiction, at the same time that it institutes an unique set of capabilities for AI as technology.

Where RAMP stands in the AI lifecycle:

RAMP fluidly plugs with today’s AI technological stack and everything else in between and does so by being its own thing, not yet another intermediary.

Figure 1: RAMP works just as the HTTP, nothing needs to be changed at the end user side as complexity is handled at service level. This approach makes RAMP out-of-the-box compatible with all known deployment packaging routes

RAMP comparison to other AI regulatory instruments:

To compare other initiatives to RAMP is not exactly something fair as RAMP is a capability multiplier element to pretty much all it can be compared to, but for simplicity’s sake a capability comparison is offered on Table 1 to expose RAMP’s unparalleled combo of features:

Table 1: Qualifying AI regulatory instruments by macro capabilities. Column items: RAMP - RoboNet Artificial Media Protocol - link, Stanford HAI Shibboleth Rule - link, European Union AI Act - link, Nist AI Risk Framework - link, International AI Organization (IAIO) - link, China’s Interim Measures for the Management of Generative Artificial Intelligence Services - link. While marked boxes do reflect an inherent capability, the range of compared items include instruments of distinct end use lineages where the intersection of capabilities may well be incremental in nature. TEVV > Test & Evaluation, Validation & Verification

How RAMP stacks with other regulatory instruments:

RAMP not only stacks well with current AI regulatory efforts, it introduces logical governance order to AI products and services:

Figure 2: RAMP provenance instrumentation at service delivery level sets a solid foundation for the introduction of business propositions that stand on top of its worldwide market capabilities in a standardized and enforceable fashion

What exactly are RAMP core features/advantages?

At a glance:

How RAMP and AI provenance initiatives stack?

RAMP is a vehicle for provenance initiatives of all kinds and flavors but RAMP also stacks its own provenance flags with an unique advantage: as it operates at Internet protocol level, RAMP provenance can identify AI products, data and services *before the content is actually delivered to end user devices, no other initiative offers this. In short, RAMP not only works with any sort of AI provenance initiative, it doubles their capabilities with its own provenance standard:

Figure 4: Above: Today HTTP enables provenance metadata to be embedded into media objects. Below: RAMP replaces HTTP and inform edge devices with the specific type of AI product being served before* any metadata or watermark ever reaches an user device, no matter if the request was filled or not

The advantages of such approach are vast, but in a nutshell RAMP allows for the enforcement of AI policies of all kinds at numerous moments of the AI product lifecycle:

Figure 5: Detail of RAMP governance loop enforcement reach with the many enforcement bottlenecks RAMP enables for policymakers to act on

What RAMP isn’t?

So here we are, hopefully this list means nothing to you at this point but for whatever case, it is here as a token. RAMP is not:

• A blockchain initiative or anything even remotely connected to cryptocurrencies;

• A web3 initiative;

• A product, a good, something that can be owned, patented or sold;

• A silver bullet: RAMP empowers AI regulations with a new Internet service design to forge new possibilities so that AI may evolve with a plan in its own environment, not as a technical imposition to all Internet users.

So RAMP is a perfect solution?

No. But it offers an unique set of capabilities that pretty much all other initiatives can use, both in technical and the regulatory domains, it is also free to all AI users no matter where, imposes little technical conditions to AI services and is the same to all countries, establishing a common and unique regulatory enforcement mechanism for AI as technology, unchained from political and corporate interests, allowing edge devices to identity AI content without relying on third parties to do so.

Any doubts? Let me know in the comments. Interested in learning more about the RAMP protocol? Grab a tea, it takes a good 30/50 minutes to read it.