Curation is the last best hope of intelligent discourse. — Joan Westenberg
Through tools like ChatGPT, anyone can conjure up rewritten Wikipedia
articles, essays, code, poetry, and more with just a few prompts. This
"democratisation" of content creation is pitched as The Great Promise to
empower voices previously unheard. But democratisation is, arguably, a
misnomer. It suggests an egalitarian shift in the power dynamics of content
creation, purportedly enabling a more diverse range of voices to be heard.
This perspective - at best, blindly idealistic and, at worst, cynically
manipulative - fails to acknowledge the underlying complexities and
potential pitfalls of an AI-dominated internet.
It assumes equal access to and understanding of these technologies across
different demographics, which is far from reality. The digital divide
remains a significant barrier, preventing equitable access to these tools.
The control and development of these AI technologies are concentrated in a
few major corporations and institutions, leading (inevitably) to
gatekeeping, bias, and the commercialisation of information.
The indiscriminate proliferation of AI-generated content will not empower
the underrepresented or democratise knowledge creation. Instead, it will
further dilute and fragment the authenticity and reliability of information
available to the public.
The AI glut significantly exacerbates the issue of misinformation and
low-quality content. The current state of AI technology lacks the nuanced
understanding and ethical judgment necessary to ensure the accuracy and
integrity of the content it produces. This gap in capability opens the
floodgates for misinformation to spread unchecked, fracturing discourse,
heightening divides and hampering decision-making.
This all boils down to one thing. Human curation is now more critical than
ever. As algorithms churn out vast quantities of information with varying
degrees of accuracy and quality, the discerning judgment of human curators
is the only defence against the tide of misinformation and mediocrity.
Human curators bring nuanced understanding, contextual awareness, and
ethical judgment to the table—qualities that AI, in its current state, is
fundamentally unable to replicate.
Human curators can distinguish between nuanced arguments, recognise
cultural subtleties, and evaluate the credibility of sources in ways that
algorithms cannot. This human touch is essential for maintaining the
integrity of our information ecosystem. It serves not only as a filter for
quality but also as a signal for meaningful and trustworthy content amidst
the overwhelming noise generated by AI systems.
The role of human curators is not just to select and present content but to
imbue the digital landscape with a sense of reliability and authenticity
that only human insight can provide. In an age where technology can easily
mislead or overwhelm, trusting in human curation becomes not a preference
but a necessity for preserving the quality of the information that shapes
our understanding of the world.
Across the Fediverse and beyond, respected voices are leveraging platforms
like Mastodon and their websites to share personally vetted links,
analysis, and creations following the POSSE model - Publish on your Own
Site, Syndicate Elsewhere. By passing high-quality, human-centric content
through their own lens of discernment before syndicating it to social
networks, these curators create islands of sanity amidst oceans of
machine-generated content of questionable provenance. Their followers, in
turn, further syndicate these nuggets of insight across the social web,
providing an alternative to centralised, algorithmically boosted feeds.
This distributed, decentralised model follows the architecture of the web
itself - networks within networks, sites linking out to others based on
trust and perceived authority. It’s a rethinking of information democracy
around engaged participation and critical thinking from readers, not just
content generation alone from so-called "influencers" boosted by
profit-driven behemoths. We are all responsible for carefully stewarding
our attention and the content we amplify via shares and recommendations.
With more voices comes more noise - but also more opportunity to find
signals of truth if we empower discernment.
This POSSE model interfaces beautifully with RSS, enabling subscribers to
follow websites, blogs and podcasts they trust via open standard feeds
completely uncensored by any central platform. RSS has been largely
forgotten in the age of the social media firehose, but it provides a
critical off-ramp from algorithmic feeds within walled gardens. It returns
agency to readers in determining their own information diet while giving
creators direct relationships with their audience rather than being
mediated through third-party networks.
RSS enables what technology essayist Venkatesh Rao calls "commonsism" -
through simple chronological updates from websites with full article text
included that let readers again own their own attention journey. With RSS,
the content stream becomes democratic again, with each person curating
their feed from diverse sites rather than submitting it to centralised
gatekeepers. This is a rethinking of information democracy as grounded in
context - trusting readers to determine their information intake aligned
with their interests and values.
The return of RSS and POSSE points to a revival of the personal website
ecosystem that thrived in the early blog era. Writers, researchers,
technologists and more are relaunching their independent homepages,
complete with feeds, as both a public notebook and a channel for sharing
insights. The personal website is the ultimate sovereign territory online,
enabling creators to share content on their own terms. These sites export
their ideas to the digital public square while filtering outbound
information to cultivate wisdom and perspective. They are living,
extensible documents evolving over time based on the site owner's journey.
In the Large Language Grift era, we need more visibility into source
provenance and credibility signals. Creators’ personal sites provide this
through clear origin stories and context. We know exactly who authored the
narratives and can evaluate them accordingly. These individuals seed
networks with reliable nodes by linking out from personal sites and
syndicating via social channels. Their sites become operations bases and
fortresses from generative excess - emitting only what they deem high
signal based on careful consideration.
Ecosystems in nature require careful balance across flora and fauna to
thrive. So, too, our information environment depends on thoughtful human
stewards curating nodes of insight across the digital public square -
editors, authors, reporters, analysts and engaged citizens. With care and
discernment, we can elevate human-centric content, rethinking information
democracy as a shared, collaborative process anchored in trusted human
judgment, not computational fiat.
Some will continue championing full automation and consolidation around a
few platforms, believing that is the only way to scale. But that creates a
monopoly of perspective, risking a monoculture vulnerable to cascading
failure and abuse. We need interoperable protocols enabling sovereign
voices to contribute to the broader knowledge ecosystem. Generative
algorithms can then be directed to reinforce existing islands of curated
insight instead of fully replacing them. We have seen how healthy
connectivity between online and in-nature networks contributes to
antifragility and collective intelligence surpassing any node.
There are early signs that ecosystem diversity may have an ally in these
large language models themselves. Tools like Anthropic's Constitutional AI
are designed to respect existing corpora rather than overwriting them with
automatic outputs. Inherently cautious and assistive systems could work
harmoniously with human-curated nodes to provide amplification without
displacing editorial integrity. Constructed carefully with hybrid
human-machine curation in mind, these technologies can be a positive force
- no matter how much I gripe and grumble about AI bros.
In many ways, we have outsourced the curation of our information diets to
distant platforms unaligned with our interests. Reclaiming even partial
agency in navigating knowledge networks is an act of self-care. Setting up
our own RSS feeds, linking out to personal sites, elevating voices that
inspire - these actions are the basis of a healthy information democracy.
One that is not defined by thoughtful pruning, editing and contextualising
via trusted guides, not by endless generation,
In every garden, seasons of growth require seasons of care - cultivating,
pruning, and even fallow periods to replenish the soil. May we embrace this
fertility cycle of information - rather than seek to dominate and control
it.
January 31, 2024 at 6:19:44 AM EST
*
- permalink
-
-
https://joanwestenberg.com/blog/curation-is-the-last-best-hope-of-intelligent-discourse
