Tweets
-
š¼š¶š¤ Merci patron - Les Charlots
https://on.soundcloud.com/ddiKsKe624eyRLRZA
Regardless of our position and even understanding of AI we must keep an eye on exploitation so often hidden behind a thin veil of technological democratisation.
(original)
-
Replying to @irinimalliaraki
Did you work in the past with policymakers thus know their workflow and tools? If so let me know I know people working on this who would surely be happy to talk.
(original)
-
RT @timnitGebru: You too should model your scientific research agenda based on what Mark Zuckerberg is doing šš½
(original)
-
Replying to @krishnanrohit
For something slightly more technical https://fabien.benetou.fr/Content/SelfHostingArtificialIntelligence but not necessarily wiser.
(original)
-
Replying to @krishnanrohit
Sure, itās called getting few beers: highly confident while spewing BS. We can all do it. Itās fun for a bit then fades off. Better also keep this to decisions or actions that wonāt have too serious consequences.
(original)
-
Weāll thatās one way to study AI riskā¦ just be the risk, for profit. /$ https://twitter.com/WillOremus/status/1643687459627180057
(original)
-
Replying to @pogue25 and @newsycombinator
Flexible honeypot?
(original)
-
Replying to @waitbutwhy
Learn more math
(original)
-
Replying to @max__drake and @imaginary_dev
thinks they are runtimes.
(original)
-
Interesting if true but the risks of āhallucinationsā (factual mistakes in AI/LLMs parlance) makes it less reliable than it looks. Still we do all want to find the appropriate explanatory level relative to our current understanding. I just worry about the backend. https://twitter.com/max__drake/status/1643594487971676160
(original)
-
Replying to @mitopenlearning, @medialab and @MitRaise
Cc @ya_lb
(original)
-
Replying to @Emil_BB, @thepine64, @remarkablepaper and @splitkb
Hi, back after weeks with the Corne-ish Zen and love it but unfortunately doesnāt fit in a pocket, at least not without putting it at risk. Did you get something better since?
(original)
-
Replying to @ruchowdh
Iād argue itās simpler than that, namely that capitalistic production endgame is no workers, only customers. Still ridiculous but maybe slightly less about a quasi religious original sin. We are flawed though, like everything else, including the machines we design.
(original)
-
Replying to @simonw
I like the analogy of shortcuts to possibly get faster from A to B but you need to know where A and B are in the first place to be able to efficiently verify, otherwise you might end up in a ditch or an intellectual cul-de-sac.
(original)
-
Ok ok itās not just me https://twitter.com/utopiah/status/1498041374226235395 and a couple of other former āfansā (check my history) who did notice a turn away from tech proper and a weirdly āinclusiveā (as in WTFā¦ not the positive sense) of the podcast.
Sadly dangerously out of his depth for such a reach. https://twitter.com/mjnblack/status/1642875488505933825
(original)
-
Replying to @rsms
Canāt help but wonder if thereās any ALife research https://en.m.wikipedia.org/wiki/Artificial_life on bitrot as the foundation for e-evodevo https://en.m.wikipedia.org/wiki/Evolutionary_developmental_biology š¤
(original)
-
RT @ouestware: š£ We are proud to announce Gephi Lite š
It is a free and open-source web application to visualize and explore networks. Itā¦
(original)
-
Replying to @dishitaturakhia and @peilingjiang
Cc @ChSimon__ for Clara Rigaudās work e.g https://dl.acm.org/doi/abs/10.1145/3490149.3503584 From the Makerspace to the Web: Creating Knowledge Resources from Fabrication Activities
(original)
-
RT @dishitaturakhia: How can we support #learning through #selfReflection and critical thinking?
Our poster at #UIST2022 on #TheReflectivā¦
(original)
-
.@Azoteacirculo Madrid šŖšø
(original)
-
What if desires paths, on places and objects, lead to better design in VR and beyond?
Stigmergy shaped worlds. https://twitter.com/utopiah/status/1643216850031968256
(original)
-
Replying to @utopiah
Also makes me wonder if after a museum visit anybody put their favorites on display back home.
Iād imagine between search and OCR it became trivial.
(original)
-
Some installations at @MuseoThyssen frame š¼ļø large flat screens.
Iāve seen setups like https://www.samsung.com/us/tvs/the-frame/highlights/ get the idea but want to support the process yet sadly often end up limiting it.
Any recommendation if you have such a setup showing solely content you curated? DIY?
(original)
-
Replying to @utopiah
Deforming with the raycaster
(original)
-
.@OnDeck for mathematics, tinkering with @threejs after watching @3blue1brown then hopefully integrating later on to designing in VR https://mobile.twitter.com/utopiah/status/1640078038535749633 to blend direct manipulation with code.
(original)
-
RT @franziroesner: Weāre excited to announce a special issue of IEEE Security and Privacy magazine exploring #security and #privacy in andā¦
(original)
-
Replying to @TfTHacker
First note on June 28, 2008 https://fabien.benetou.fr/Site/AllRecentChanges so nearly 15 years.
(original)
-
RT @FrontVR: New Research: Design-thinking skill enhancement in virtual reality: A literature study https://www.frontiersin.org/articles/10.3389/frvir.2023.1137293?utm_source=S-TWT&utm_medium=SNET&utm_campaign=ECO_FRVIR_XXXXXXXX_auto-dlvrit #VirtualReality
(original)
-
Replying to @utopiah
Form from function over time, design responding to changes
(original)
-
Replying to @utopiah
IMHO itās also in this kind of setups that plasticity is key, not just animated but rather reshaping from constraints this where code has a central place.
(original)
-
Maybe I be shouldnāt think of live coding as such, in VR or otherwise, but rather as improv where we collaboratively shape back space itself š¤ https://twitter.com/andreeavr/status/1642979654872059912
(original)
-
Replying to @GattaKat
(original)
-
RT @theentreflaneur: Young tech #entrepreneurs at the cusp of making business-, life- decisions should be given an opportunity to invest thā¦
(original)
-
Replying to @zachernuk, @Blackberu and @donalddhoffman
FWIW tried a while ago the great result (optimised a bit since via gltf-transform) https://mobile.twitter.com/utopiah/status/1616071822591815682 but didnāt have the chance to participate to the collaborative design process, hopefully soon!
(original)
-
Replying to @_akhaliq
cc @readyplayerme
(original)
-
Type your expected motion and… watch the avatar moveš® https://twitter.com/_akhaliq/status/1643077969215205376
(original)
-
Replying to @rachel_l_woods
I wish that “People want their data to be private.” but for now it seems very few actually take the small extra effort, either paying for a service or hosting themselves, to actually do so.
There might be a shift but for now mostly virtue signalling, sadly.
(original)
-
Imagine an open model, e.g @BigCodeProject, @berkeley_ai’s Koala, @lmsysorg’s Vicuna hosted locally then run as your new @GiteaIO Action parsing tagged issues, generating tests, a la CodeT, then proposing PRsš¤© https://twitter.com/_akhaliq/status/1643082187321794562
(original)
-
“Through their teaching, research, and projects likeĀ Work It Out Wombats!, Dr. Bers and her colleagues demonstrate that education can address not only intellectual and practical concerns, but also enduring moral, civic, aesthetic, and spiritual questions.” https://twitter.com/DevTechResearch/status/1641843636529922049
(original)
-
Replying to @zachernuk, @Blackberu and @donalddhoffman
too expensive, is even achievable in the first place.
Anyway… I haven’t listen to that episode but thanks a lot for sharing it, downloadedš§ and can’t wait, going to brighten my commute I have no doubt!
(original)
-
Replying to @zachernuk and @Blackberu
outside every morning. What is real? What is the “quotidien” and my experience of it? My habits?
It’s clearly easier to ignore that and “just” navigate in our comfortable illusion.
Actually it loops back to @donalddhoffman’s The Case Against Reality. Access to reality is…
(original)
-
Replying to @zachernuk and @Blackberu
That being said it can be tricky to disentangle precisely how poorly or how imprecisely we describe what seems, again naively, immediate.
A good example IMHO is Michel de Certeau and daily life. We know we know. We need that confidence otherwise we wouldn’t take a single step…
(original)
-
Replying to @zachernuk and @Blackberu
inside wanting to bluntly come back saying he had no idea what reality was, none. That it was an extremely naive perspective.
It’s honestly infuriating to see how quickly the topic can be dismissed for a variety of reasons but the first one, sadly, being confident ignorance.
(original)
-
Replying to @zachernuk
Indeed. Few days ago I was discussing the topc with architects, with @Blackberu https://twitter.com/utopiah/status/1641122569012101141 and I asked who actually has been in the “metaverse”.
The first replied “I haven’t.” I asked why, he replied, kind of proudly “I prefer reality”.
Honestly I was boiling…
(original)
-
RT @zachernuk: I just finished listening to this *brilliant* conversation between @andreeavr and @kentbye. One thing Andreea mentioned I wā¦
(original)
-
Replying to @AntonioCasilli
FWIW https://twitter.com/voxdotcom/status/1642995312049389569
(original)
-
Replying to @aoighost
Well that’s humanity for you. We get to try to decide what our future is. Based on our current response to numerous events, including climate change, I’d argue that ransomware on critical systems isn’t even that bad.š¤·āāļø
HerešŖ
(original)
-
Why, despite focusing on #XR and #WebXR, I still share so much about AI, both criticism and enthusiasm?
Because AI, generative or not, matters. It’s “just” a tool to build more but it’s very important to use the right tool from your moral framework : https://twitter.com/utopiah/status/1306566303839391744
(original)
-
“learning from high-quality datasets can mitigate some of the shortcomings of smaller models, maybe even matching the capabilities of large closed-source models in the future. […] enable safer, more factual, and more capable models than simply increasing the size” https://twitter.com/berkeley_ai/status/1643053599390593024
(original)
-
RT @lmsysorg: Introducing Vicuna, an open-source chatbot impressing GPT-4!
š Vicuna reaches 90%* quality of ChatGPT/Bard while significantā¦
(original)