"Plagiarize or Perish" at Harvard and MIT?

Trustless information will be a boon for journalists and academics, though it could put your favorite big media brand or university to rest.

Welcome, welcome, welcome!

Let me apologize for the inconvenience, but this week’s edition is too long for email readers to display.

Read the full post here: Immutabletype.substack.com

What’s covered in the full edition on Immutabletype.substack.com

• A raunchy joke from Sir Paul McCartney
• Alleged plagiarism by Claudine Gay
• Bill Ackman’s Twitter/X threads
• Alleged plagiarism by Oxman
• Trust vs Trustless systems
• Hard Reputations of an onchain world will save journalists

The following is a snippet from the full post.

Hardhats & Blocks

With the groundwork set to outline the headwinds for individuals within the legacy media system, we are now able to build a system to solve the needs of this lagging industry.

I will use some examples from the above to discuss the new onchain methods below. I recommend familiarizing yourself with Part One above if only to recognize the names of individuals below.

My opinion is we need to build in parallel with the existing media complex and migrate behavior to the new information network experience. There will be many opportunities to work within the legacy model, but the majority of those engagements will simply embed or fork the work we build. This is good news and is welcomed.

The media landscape is now a construction zone.

Trustless Systems > Trust

What’s better than trusting relationships? Trustless relationships!

The word “trustless” has taken me some time to fully internalize personally, but a good way to think about it is that a trustless system is built so it cannot lie, or at least, it can’t behave outside the verifiable public code. It is free from trust, because a trustless systemis unable to be trusted or mistrusted.

Can’t Lie is the easiest working concept to remember when thinking about a system being trustless.

“Trust Me” Systems

A system with an intermediary requires some level of trust to govern transactions between multiple parties, whether it’s peer-to-peer, one-to-many, or many-to-many.

No two words, however, raise more red flags than a person who says, “Trust me.” Why would that be? Because saying, “Trust me” announces that the potential to lie exists, so the party seeking trust must explicitly request it. A request to “Trust me” means the party requesting it does not have enough evidence available to make a convincing case that the agreement will be completed. Something could always go wrong after the handshake.

Despite that setup, trust me systems are what we are accustomed to in daily life. Banks, governments, media companies, all our subscription services, payment cards, car warranties, insurance companies, everything relies on trust to deliver upon an agreement. We all know this trust could be violated in some ways, but it’s the system we have, so we accept the likelihood of some loss of value by agreeing to accept the terms of service from the other party.

Media brands and academic institutions are information intermediaries that operate trust me systems to establish a consensus of what is true or accurate about information. Their business is to generate revenue by providing information to an audience that meets their internal standards for truth and accuracy. The consensus for what is true and accurate is reached internally and then published to an audience as information that has met the information intermediary’s standards. Our trust is put in the hands of the intermediary because we are not present to observe the execution of their policies. There is no mechanism to confirm they have done as promised, and there is no public dashboard to display their historical performance and track record.

Axel Springer SE, for example, conducted an internal review of the processes used by Business Insider to develop and publish the Oxman story. The standards of the review were not made available to the public, nor were the methods of evaluation. The audience was left only with a choice to trust or distrust a published report from Sam Jones of the Financial Times that Axel Springer SE was satisfied with its findings. The article by Jones was very thin for such a big story about media practices, and Jones himself has a very weak online reputation to be entrusted with such a story, yet it’s all we have. We have a headline, effectively, which waves a hand and assures that everything is fine. In short, we’re asked to “trust me” as an audience of the Financial Times.

Trust is the key for these businesses when operating “trust me” systems for consensus about what information is true and accurate. The job of an information intermediary is to gather, organize, format, and distribute uniquely actionable information. Trust makes that work actionable. If no trust, then no action. Action is the only value output for which the audience is willing to pay a fee. If no trust, no action, and no fees paid…

This seems so obvious, but the lack of public trust in media brands is precisely the problem we have today. The result of this lack of trust is an industry in crisis with widespread media company layoffs and closures. Information that is not both unique AND actionable is not valuable to any audience, so we churn from buying their services as the audience. The media brands and agents that fail to create actionable information not only destroy investor value, but ultimately damage the entire industry by operating at all.

This void in trust has spurred the development of new tools and services from outside the academic and media systems. The goal of the tools is to help improve the audience’s understanding of the accuracy of information within trust me systems, such as plagiarism detection technology and media rating services. Each product type is effectively a monitoring solution, so the product really lacks sufficient features to increase information clarity for consumer audiences at scale in real time. The technology just doesn’t have the juice to keep up with such a soft system.

Plagiarism software, for example, has been used by educational institutions for years with the purpose being to prevent students and faculty from passing off plagiarized works as their own. Recently, however, activist researchers have turned the plagiarism software upon academic work published before the adoption of the software by educational institutions. The result is the activists have identified significant failures of the trust me system of peer review at Harvard, and one must believe this type of research will turn up similar damaging results across most academic institutions once AI-enabled tools are tasked with the job. Regardless of the scope of the peer review failures that come to light, the minimum finding will be that trusting an intermediary to deliver accurate information will not be entirely reliable. It’s time to externalize that job to a network and stop trusting intermediaries to behave properly.

Media monitoring services, for their part, do address bias by the media brands, but the results don’t reach the audience at the time when the audience needs that information, which is at the point of engagement. While these media monitoring services do appear to provide relevant and accurate results, the ratings only address the historical performance of the brands, rather than the individual stories published under the banner of the brand. Additionally, the agents of the intermediaries (journalists & editorial management) are not ranked or indexed, so the range of bias of any individual story is unknown at any time. Monitoring services do provide value, but their impact is too low to be powerful.

We can solve these blind spots with trustless systems.

“Trustless” Systems

A trustless system does not require an intermediary to validate transactions between peers. It can only execute the public code.

Blockchains are trustless systems, for example, where the parties that agree to a transaction need not know each other directly, nor rely on the reputation of the other peer’s intermediary to have enough confidence to complete a transaction. Unlike a trust me system, an effective trustless system cannot lie.

As noted, Business Insider, MIT, and Harvard are information intermediaries because they are the entities that have established the policies that govern their agents as the layer of promises between the raw information source and the audience. The act of governing the performance of its agents is conducted by officers and directors of the intermediary away from public view. Again, we must trust the intermediary to not lie when reporting its conclusions.

A trustless system eliminates the need for the services of information intermediaries because trust is no longer required by the audience. Trust is only a criterion when the reputation of an intermediary is necessary to assure the audience that information has been gathered, organized, formatted, and distributed in a manner that contributes uniquely actionable information to the audience experience.

The trustless system, however, does the job of assuring compliance within governance criteria, so agents can operate freely between the information source and the audience as a commentator of subject matter. All the jobs previously completed by the information intermediary are executed publicly according to coded criteria that are verifiable and immutable. The journalists and the academics become the information stewards, and the code validates the criteria that have been met on behalf of the networked audience.

In a trustless system, “code is law” therefore any behaviors outside the encoded criteria may NOT pass through the network, regardless of the reputation of the actor. This is a hard system to replace our existing soft systems. The code is incorruptible governance, so we may have confidence as the audience that the information we access has met the publicly verifiable criteria of the trustless system. The audience doesn’t need to hope the intermediary honors journalistic integrity as part of their obligation to a professional guild, or question if such a commitment is little more than a magic cloak to ward off inquiry when the business reviews the work of its agents. We will simply know the code executed as designed.

One of the key features of an effective trustless system is that it is immutable. The transactions are incapable of being corrupted, bribed, or attacked successfully, and this is quite appealing when we’re dealing with information integrity. History, for example, may not be rewritten to serve the needs of those who would like to change it. In a trustless system, what is stored onchain may only be appended to previously published information. Histories stored onchain may not be augmented or entirely wiped away at the command of any new power, as the code will not allow that action. We won’t lose Alexandria again; cultures won’t be obliterated; and, histories won’t be dictated tops down.

Continue to the full post here: Immutabletype.substack.com