Camille lives at Praxis Now · @camille
36 followers · 46 posts · Server social.praxis.nyc
Japheth Campbell 🇺🇦🌻 · @japheth
5 followers · 37 posts · Server social.vivaldi.net
ニトロン · @2YzwfXIsC2bLGVT
104 followers · 12742 posts · Server mstdn.jp
Sam Panini · @spanini
32 followers · 237 posts · Server me.dmI published a hot take on Medium: https://sampathpanini.medium.com/hard-decisions-for-dikw-b9c40b0292ed
#AI #culture #leadership #GIGO #datagovernance #strategy #management #datamanagement #stewardship #ethics #decisionmaking
#ai #culture #leadership #gigo #datagovernance #strategy #management #datamanagement #stewardship #ethics #decisionmaking
David Raygoza Gómez · @hashraydamon
54 followers · 485 posts · Server me.dm
Destiny 2 fans use Glorbo to trick AI-written websites - Polygon
https://www.polygon.com/23808925/destiny-2-glorbo-ai-sites-secret-boss-running-joke
Ruth Mottram · @Ruth_Mottram
6745 followers · 9877 posts · Server fediscience.org
Habsburg AI: #MachineLearning models fed on input that has been produced by other ML models...
A beautiful yermt I learned from @pluralistic showing why #GIGO (garbage in, garbage out) also explains why dictators can't rely on #AI to detect stirrings of revolt.
As always, worth a read.
Kenny Smith · @kennysmith
123 followers · 2732 posts · Server mstdn.social“#GarbageInGarbageOut … is a very inconvenient truth …”
What an incredibly salient point in this instance. #GIGO
From: @pluralistic
https://mamot.fr/@pluralistic/110781205598033675
Cory Doctorow's linkblog · @pluralistic
44869 followers · 43000 posts · Server mamot.frBut adding more unreliable data to an unreliable dataset doesn't improve its reliability. #GIGO is the iron law of computing, and you can't repeal it by shoveling more garbage into the top of the training funnel:
When it comes to "AI" that's used for decision support - that is, when an algorithm tells humans what to do and they do it - then you get something worse than Garbage In, Garbage Out - you get Garbage In, Garbage Out, Garbage Back In Again.
7/
Dennis Alexis Valin Dittrich · @davdittrich
570 followers · 738 posts · Server fediscience.orgPitfalls of #DigitalScholarship Machiavelli and Matching
https://marketdesigner.blogspot.com/2023/07/pitfalls-of-digital-scholarship.html?m=1
…are the pitfalls of badly labelled data. Without at least a minimum of quality assessment and control of the source data, we just end up with nonsense. #gigo
AI6YR · @ai6yr
3853 followers · 22644 posts · Server m.ai6yr.org
三頭 :verified: · @santou_hong
323 followers · 907 posts · Server g0v.social
Alberto Cottica · @alberto_cottica
376 followers · 814 posts · Server mastodon.green
"Training a model on its own output is not recommended." #GIGO #AI
https://www.theregister.com/2023/06/16/crowd_workers_bots_ai_training/
Eka A. · @Eka_FOOF_A
83 followers · 2889 posts · Server spacey.space@delta_vee @davidgerard If the training data contains error, ie garbage, then garbage is what you get out. #GIGO #AI currently in this case means Artificial Idiot #ArtificialIdiot
Kay :heart_bi: · @Kay
637 followers · 6571 posts · Server mastodon.nz@XauriEL #GIGO = #GarbageInGarbageOut gets worse with greater volumes of garbage.
Unless AIs/ ML systems or LLMs are trained using systematically cleansed data or information the problem will escalate like a #fatberg in the drains.
#gigo #garbageingarbageout #Fatberg
Hobson Lane · @hobs
689 followers · 1936 posts · Server mstdn.social
Damien Riehl · @damienriehl
115 followers · 138 posts · Server law.buildersReminder: In #PromptEngineering, accurate/helpful results require #SMEs.
If you say "ChatGPT sucks" remember #GIGO (Garbage In, Garbage Out).
Google: bad queries get bad results; good queries get good results. Same with #GPT + #ChatGPT + #LLMs.
#Lawyers = SMEs.
#Lawyers as #SMEs can use #legaltech to be good #PromptEngineers.
Good article on this point:
https://www.law.com/legaltechnews/2023/04/27/a-lot-of-prompt-engineering-how-law-firm-traverse-legal-built-a-gpt-powered-client-tool/
#PromptEngineering #smes #gigo #gpt #chatgpt #llms #lawyers #legaltech #promptengineers
Flock of Nazguls · @flockofnazguls
1076 followers · 3066 posts · Server mastodon.flockofnazguls.com
AndiMann · @AndiMann
345 followers · 933 posts · Server masto.aiThe big turn on the hype cycle for #AI is reinforcing the need for #data quality.
#GIGO ('Garbage In Garbage Out' - one of the oldest principles of computing) is magnified exponentially in AI use cases.
If your #ML/AI training dataset is even slightly tainted, your AI will be deeply compromised.
ghost_shit 🚮 · @ghost_shit
434 followers · 6900 posts · Server aus.social
Eka A. · @Eka_FOOF_A
73 followers · 2266 posts · Server spacey.space@UlrikeHahn @ct_bergstrom I see it as a classic case of #GIGO (Garbage In, Garbage Out).