Jef Allbright<p>I'm excited to see recent developments with LLMs may provide effective modeling of our aggregate values (multi-level, fine-grained), increasingly coherent over increasing context of meaning.</p><p>LLMs such as ChatGPT and Gemini already impress us with their ability to produce coherent (but not necessarily correct) abstractions from massive collections of data relevant to our interests.</p><p>Now, what if we run multiple instances of LLMs with diverse priors, collaboratively (at lower levels adversarial, at higher levels cooperative) selecting for increasing coherence over increasing context in the domain of present but evolving values?</p><p>Of the two orthogonal dimensions producing the expanding space of moral (right in principle) agency —our values, and our methods for their promotion — we already have much attention on our evolving methods (science, technology) and it appears we are on the cusp of rapid development of an effective model of our evolving values!</p><p>Cause for (Promethean) hope!</p><p>Research remains to be performed on theoretical and practical questions involving expected signal-to-noise improvements of "synthetic data" extracted from latent information in our datasets, regression due to synthetic data added to the data set, biases, gaming of the system, and on and on… Interesting, exciting, and dangerous.</p><p><a href="https://mathstodon.xyz/tags/LLM" class="mention hashtag" rel="tag">#<span>LLM</span></a> <a href="https://mathstodon.xyz/tags/LargeLanguageModels" class="mention hashtag" rel="tag">#<span>LargeLanguageModels</span></a> <a href="https://mathstodon.xyz/tags/AI" class="mention hashtag" rel="tag">#<span>AI</span></a> <a href="https://mathstodon.xyz/tags/GenerativeModels" class="mention hashtag" rel="tag">#<span>GenerativeModels</span></a> <a href="https://mathstodon.xyz/tags/SyntheticData" class="mention hashtag" rel="tag">#<span>SyntheticData</span></a> <a href="https://mathstodon.xyz/tags/DecisionMaking" class="mention hashtag" rel="tag">#<span>DecisionMaking</span></a> <a href="https://mathstodon.xyz/tags/CollectiveDecisionMaking" class="mention hashtag" rel="tag">#<span>CollectiveDecisionMaking</span></a> <a href="https://mathstodon.xyz/tags/Ethics" class="mention hashtag" rel="tag">#<span>Ethics</span></a> <a href="https://mathstodon.xyz/tags/MetaEthics" class="mention hashtag" rel="tag">#<span>MetaEthics</span></a> <a href="https://mathstodon.xyz/tags/Morality" class="mention hashtag" rel="tag">#<span>Morality</span></a> <a href="https://mathstodon.xyz/tags/ArrowOfMorality" class="mention hashtag" rel="tag">#<span>ArrowOfMorality</span></a></p>
Recent searches
No recent searches
Search options
Only available when logged in.
mathstodon.xyz is one of the many independent Mastodon servers you can use to participate in the fediverse.
A Mastodon instance for maths people. We have LaTeX rendering in the web interface!
Administered by:
Server stats:
3Kactive users
mathstodon.xyz: About · Status · Profiles directory · Privacy policy
Mastodon: About · Get the app · Keyboard shortcuts · View source code · v4.3.6
#CollectiveDecisionmaking
0 posts · 0 participants · 0 posts today
Timothy Shortell<p>One thing that I find really frustrating about work, even in an engaged sociology department, is how hard it is for us to get from talking about an issue (curriculum or pedagogy) to making a collective decision. So we end up with a model of social change in our department life that is largely voluntary, the aggregate of individual decisions to make changes or not.<br><a href="https://sciences.social/tags/Sociology" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Sociology</span></a> <a href="https://sciences.social/tags/AcademicLife" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AcademicLife</span></a> <a href="https://sciences.social/tags/Curriculum" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Curriculum</span></a> <a href="https://sciences.social/tags/Pedagogy" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Pedagogy</span></a> <a href="https://sciences.social/tags/CollectiveDecisionMaking" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>CollectiveDecisionMaking</span></a></p>
Jef Allbright<p>We need a workable system for reputation management that is </p><p>- decentralized and transparent in its operation</p><p>- robust, and immune to being gamed [see the first item...]</p><p>- multidimensional, since people's strengths and weaknesses, and impacts on others, vary with type</p><p>- discounts past behavior in favor of more the more recent</p><p>I think we have evidence that reputation management works in small groups, villages, families—so the challenge is how to apply our new tools for information management to our new, larger, social groups.</p><p>We need to apply new technologies to our new circumstances, to scale up our collective awareness and decision-making to match the exponentially increasing flood of data and connections.</p><p>And not by delegating to an independent agency such as an AI, or a 'trusted' leader or organization, or to an electoral system that fails on many of the criteria listed above.</p><p>Next, a system for modeling our values, multilayered and fine-grained, with increasing coherence over increasing context of meaning, and increasing scope of consequences…</p><p><a href="https://mathstodon.xyz/tags/ReputationManagement" class="mention hashtag" rel="tag">#<span>ReputationManagement</span></a>, <a href="https://mathstodon.xyz/tags/CollectiveIntelligence" class="mention hashtag" rel="tag">#<span>CollectiveIntelligence</span></a>, <a href="https://mathstodon.xyz/tags/CollectiveDecisionmaking" class="mention hashtag" rel="tag">#<span>CollectiveDecisionmaking</span></a>, <a href="https://mathstodon.xyz/tags/TechnologyAndSociety" class="mention hashtag" rel="tag">#<span>TechnologyAndSociety</span></a> <a href="https://mathstodon.xyz/tags/Democracy" class="mention hashtag" rel="tag">#<span>Democracy</span></a> <a href="https://mathstodon.xyz/tags/Values" class="mention hashtag" rel="tag">#<span>Values</span></a></p><p>-</p>
Jef Allbright<p>Dusting off an old idea that may be nearing practicality with the rise of large language models:</p><p>A model of human universal values—multilevel, fine-grained, probabilistic, intertwingled, and evolving—could facilitate useful discourse and collective decision-making.</p><p>It seems to me that our worldwide, multicultural corpus of fiction must contain examples, within context, of nearly all human values and situations in which they may apply.</p><p>Such fictional examples are of course highly biased toward the dramatic and the sensational, since that's what people tend to enjoy reading, but might we now be close to having the tools to help us collect, categorize, and make sense of such a data-store and apply it to effective discussion and decision-making?</p><p><a href="https://mathstodon.xyz/tags/decisionmaking" class="mention hashtag" rel="tag">#<span>decisionmaking</span></a> <a href="https://mathstodon.xyz/tags/values" class="mention hashtag" rel="tag">#<span>values</span></a> <a href="https://mathstodon.xyz/tags/HumanValues" class="mention hashtag" rel="tag">#<span>HumanValues</span></a> <a href="https://mathstodon.xyz/tags/HumanUniversals" class="mention hashtag" rel="tag">#<span>HumanUniversals</span></a> <a href="https://mathstodon.xyz/tags/CollectiveDecisionmaking" class="mention hashtag" rel="tag">#<span>CollectiveDecisionmaking</span></a> <a href="https://mathstodon.xyz/tags/TextProcessing" class="mention hashtag" rel="tag">#<span>TextProcessing</span></a> <a href="https://mathstodon.xyz/tags/CorpusAnalytics" class="mention hashtag" rel="tag">#<span>CorpusAnalytics</span></a> <a href="https://mathstodon.xyz/tags/analytics" class="mention hashtag" rel="tag">#<span>analytics</span></a> <a href="https://mathstodon.xyz/tags/LLM" class="mention hashtag" rel="tag">#<span>LLM</span></a> <a href="https://mathstodon.xyz/tags/Cooperation" class="mention hashtag" rel="tag">#<span>Cooperation</span></a> <a href="https://mathstodon.xyz/tags/emotion" class="mention hashtag" rel="tag">#<span>emotion</span></a></p>
Jef Allbright<p>Dusting off an old idea that may be nearing practicality with the rise of large language models:</p><p>A model of human universal values—multilevel, fine-grained, probabilistic, intertwingled, and evolving—could facilitate useful discourse and collective decision-making.</p><p>It seems to me that our worldwide, multicultural corpus of fiction must contain examples, within context, of nearly all human values and situations in which they may apply.</p><p>Such fictional examples are of course highly biased toward the dramatic and the sensational, since that's what people tend to enjoy reading, but might we now be close to having the tools to help us collect, categorize, and make sense of such a data-store and apply it to effective discussion and decision-making?</p><p><a href="https://mathstodon.xyz/tags/decisionmaking" class="mention hashtag" rel="tag">#<span>decisionmaking</span></a> <a href="https://mathstodon.xyz/tags/values" class="mention hashtag" rel="tag">#<span>values</span></a> <a href="https://mathstodon.xyz/tags/HumanValues" class="mention hashtag" rel="tag">#<span>HumanValues</span></a> <a href="https://mathstodon.xyz/tags/HumanUniversals" class="mention hashtag" rel="tag">#<span>HumanUniversals</span></a> <a href="https://mathstodon.xyz/tags/CollectiveDecisionmaking" class="mention hashtag" rel="tag">#<span>CollectiveDecisionmaking</span></a> <a href="https://mathstodon.xyz/tags/TextProcessing" class="mention hashtag" rel="tag">#<span>TextProcessing</span></a> <a href="https://mathstodon.xyz/tags/CorpusAnalytics" class="mention hashtag" rel="tag">#<span>CorpusAnalytics</span></a> <a href="https://mathstodon.xyz/tags/analytics" class="mention hashtag" rel="tag">#<span>analytics</span></a> <a href="https://mathstodon.xyz/tags/LLM" class="mention hashtag" rel="tag">#<span>LLM</span></a> <a href="https://mathstodon.xyz/tags/Coopertion" class="mention hashtag" rel="tag">#<span>Coopertion</span></a> <a href="https://mathstodon.xyz/tags/emotion" class="mention hashtag" rel="tag">#<span>emotion</span></a></p>
SearchLive feeds
Mastodon is the best way to keep up with what's happening.
Follow anyone across the fediverse and see it all in chronological order. No algorithms, ads, or clickbait in sight.
Create accountLoginDrag & drop to upload