@david_chisnall Interesting! Another reason to feel comfortable in the decision to not use Mastodon anymore ;) I wonder why they would have that design decision…
Per Axbom
As part of my overarching work on digital ethics I've more often lately been looking at all the data that lives on after our passing, how it affects family, friends and acquaintances, and the many parties that control our online legacy. If we don't talk about it more it's very hard to know what to expect, understand risks and express wishes and concerns.
”Refusal is not merely a negation, but a pathway to different futures.”
Louder, please. 🙏
Fediverse appreciation post.
I don’t know about the apps you’re using but at no point in the Fediverse have I lost track of a post I was reading because the feed auto-refreshed against my will.
Which happens to me regularly on LinkedIn. (And I remember happened on the other platforms I’ve left.) Which should make you acknowledge – if you haven’t done so already – how that is a design feature on the other platforms, not a bug.
This. Alastair has been a huge inspiration to me over the years. Do listen to our interviews with him on UX Podcast , and attend one of this workshops if you can. Or invite him to your organisation!
Clip from the wonderful UX Copenhagen conference.
@elhult
> One objective could be to generate text that is satisfying to customers. 🤷♂️
Haha, fair enough. It struck me that they had that sentence because they expected AI to be more specific, my assumption being that this definition came into being before the LLMs did, but I'm not sure.
> I don't think the term AI is very good because it is too broad. The OECD definition has that problem too.
Definitely agree that the term AI is really just confusing, especially since there are so many different definitions going around. The EU AI Act being another example which has lots of words but which really boil down to "something automated by a computer".
> do you have any AI definition that you like?
Not really, for the same reasons. I think we should divide whatever AI covers into at least 2 or 3 different segments. The fact that "AI" is also used so extensively in science fiction really just means that it's destined to give people the wrong idea.
In Moral Codes, Alan Blackwell talks about 2 different kinds of AI:
"The first kind of AI used to be described as “cybernetics”, or “control systems” when I did my own undergraduate major in that subject. These are automated systems that use sensors to observe or measure the physical world, then control some kind of motor in response to what they observe. [...]"
"The second kind of AI is concerned, not with achieving practical automated tasks in the physical world, but with imitating human behaviour for its own sake. The first kind of AI, while often impressive, is clever engineering, where objective measurements of physical behaviour provides all necessary information about what the system can do, and applying mentalistic terms like “learning” and “deciding” is poetic but misleading. The second kind of AI is concerned with human subjective experience, rather than the objective world."
This is a good start for understanding why the public discourse can be confusing. Especially when people point to the benefits of type 1 when one is critiquing type 2.
Source: https://moralcodes.pubpub.org/pub/1mn2q39n/release/6
As a kid I spent so many hours on Flight Simulator II (on my Atari 800XL), I finally believed I could cope well if I had to take over the controls of a commercial plane. 😅. I mean I had TWO joysticks!
@dcm
I was thinking from a user perspective that a tool like ChatGPT does not have a clear objective. But with the interpretation that the user has some sort of objective then everything does I suppose.
I was struck by how different a general-purpose chatbot is from an AI built for detecting anomalies in x-rays for example, where the objective is very clear. My assumption was then that the definition was once written with specific applications of machine learning in mind.
@elhult
Maybe it's just me but I don't get the impression that ChatGPT has an explicit or implicit objective for example.
Looking at the definition of AI by the Organization for Economic and Co-operation Development (OECD):
"An AI system is a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment."
Based on this, if a language model has no explicit or implicit objectives it is not an AI.
Nu har det gått lite mer än ett år sedan jag skrev min långa text om AI-hajpen, och spelade in den som podd. Jag vågar påstå att den fortfarande håller.
« Det är aldrig så lätt att bli lurad som under en pågående hajp. Det är mars 2023 och vi är mitt i den. Det är sällan jag sett så många anamma en helt ny experimentell lösning med så lite ifrågasättande. Just i detta nu är det oerhört viktigt att skaka av sig hypnosen och granska innehållet i den här nya flaskan med AI som flertalet börjat smutta på, eller redan börjat tanka sina företagsdatorer med. Ibland utan ledningens vetskap. »
Lyssna eller läs: https://axbom.se/ai-ansvar/
Jag ser många företag som har hållbarhet på agendan, men få rapporter om hur de använder AI med tanke på den ökade el- och vattenförbrukning som det innebär. Mäter man det som en kostnad och betalar tillbaka i en annan ände? Utan den kalkylen och transparensen tänker jag att det är svårt att få förtroende för budskap om hållbarhet och cirkulär ekonomi.
Kanske är det enklast att bara ignorera och köra på som om den egna användningen av AI-tjänster inte spelar roll. Strunt samma om företagets bidrag till vatten- och elförbrukning ökar exponentiellt om man ändå precis bytt till ekologiskt kaffe och lågenergilampor i konferensrummet.
Om företaget vars tjänster man använder inte vill avslöja sin energiförbrukning är det en leverantör man gärna anlitar? Och hur förhåller man sig i då till ett hållbarhetstänk? Kalkylen blir svår tänker jag. Och i så fall har nog företag inte så bra koll på den hållbarhet och cirkulära ekonomi som de gärna marknadsför.
Jag tar gärna emot tips på företag och organisationer som gör en ansats att redovisa hur mycket mer energi deras verksamhet leder till i andra och tredje led, efter att alla anställda anammar AI-tjänster.
Alternativt som uttryckligen bara säger att de kör på och hoppas någon annan löser den där obetydliga frågan om “energiförbrukning”.
För mig personligen är alla rapporter kring energi- och vattenförbrukning ett stort skäl att undvika användningen av generativ AI, och då kanske särskilt verktyg för att generera bild och video. Dessa kostar inte bara oerhört i träning, tillverkning och användning utan även i till exempel lagring och bandbredd.
Så klart har företagsanvändare ansvar i detta även om man av ekonomiska skäl gärna spelar ovetande. Kanske tänds insiktens lampa när fakturorna för företagets molnlagring ökar markant. Tyvärr är insiktens lampa alltför ofta också av lågenergi-varianten just nu.
Men var är hållbarhetsrådgivarna i allt detta? Om man lär ut AI i skolan, pratar man om resurserna som krävs? Jag tänker att många ungdomar skulle ha åsikter om de framtida vägvalen när de förstår teknikens klimatpåverkan.
Jag läser om hur människor upprörs av någons flygresa, och hickar till när samma människor i nästa stund publicerar exempel på många av de bilder och filmer de sitter uppe på nätterna och genererar, och uppmuntrar andra att göra också. Att agera hållbart handlar idag mest om att inte göra det som riskerar ens eget rykte. De personliga relationernas accepterande attityd får styra, snarare än intresse för faktisk påverkan.
Opinionen har ännu inte, med klimatpåverkan i åtanke, bidragit till en bredare känsla av AI-skam. Kanske kommer det när man känner att den där halvlitern vatten kan komma till bättre användning än att ställa 50 frågor till ChatGPT.
“Du ersätts inte av AI, du ersätts av en annan människa som använder AI”, brukar det heta. Det är ett snyggt sätt att skrämma till användning utan eftertanke. Men låter det hållbart?
Hela texten, med vidare läsning, på bloggen: 👇
@rrwo 😁😂💸
The fact that you are on floss.social made this response epic.
@timmymac Haha, exactly. That’s identifying the real harms!
Imagine a ”smart” toothbrush. Through a bluetooth-enabled app on your phone, it is programmed to provide personalised feedback on your brushing habits based on time spent, pressure applied and area covered.
Now imagine worrying that adding more and more computing power to the toothbrush will create an all-powerful superbeing that erases mankind.
I get that vivid imaginations can end up there but that they are being taken so seriously, without any more coherent reasoning or explanation than ”it feels like it”, is truly baffling.
For a walkthrough of the many logical fallacies of believing computers will by themselves evolve into a "superbeing", I encourage you to read the first chapter of Moral Codes by professor Alan Blackwell.
The chapter also contains a great primer on the two main different kinds of AI. The first kind is what used to be called "cybernetics", or "control systems", with a sensor (like a thermostat) controlling an output (like a heating pump) . And the second kind of AI being the one that intends to imitate human behaviour for its own sake.
As Blackwell writes: "Many public discussions of AI do not acknowledge this distinction between the objectively useful and (sometimes) straightforward engineering of practical automated machinery, and the subjective philosophical enterprise of imitating the way that humans interact with each other."
Blackwell goes on to argue that many AI researchers have blurred the lines between computing and philosophy, without formal training in the latter. There is a great deal of magical thinking going on, with no basis in scientific reasoning.
«The magical powers of genies and deities are not precisely defined in stories, because clear definitions would limit the power of the narrative device. If anyone did commit to a technical definition of AGI, the contradictions would become clear, because the definition itself has to be coded in some way. Of course the answer can’t be a circular one - you can’t say that the necessary code definitions will be written by the AGI itself! Intelligence that could manifest itself in any way, without the inconvenience of a clear definition, is a plot device like the shape-changing T-1000 of Terminator II which might be described as an “artificial general body” in the way it is unlimited by any definition of shape, and thus able, in the imaginary universe of the movie, to become anything at all.»
Read the chapter here.
https://moralcodes.pubpub.org/pub/1mn2q39n/release/6
And by all means, keep on reading after that 😉
Jag rockar sockorna varje dag men idag är dagen när många fler gör det också. Älskar att kika under konferensbordet idag ❤️🎉🧦
#WorldDownSyndromeDay firas den 21/3 för att uppmärksamma den unika uppsättningen med 3 exemplar av kromosom nummer 21.
Några länkar jag gillar att dela:
En film som beskriver hur Trisomi-21 funkar:
https://www.youtube.com/watch?v=o0VV3C_ydak
George Webster avfärdar 5 myter om Downs syndrom:
https://www.bbc.co.uk/programmes/p09kyydg
Frank Stephens tal i FN 2018:
https://www.youtube.com/watch?v=1d8ocuPrlT8
Den senaste kampanjfilmen ”Assume that I can so maybe I will”, inklusive ett porträtt av Madison Tevlin, stjärnan i filmen, och om inspirationen till filmen:
https://www.ds-int.org/Blog/assume-that-i-can
Från svenska Downföreningen:
”I Sverige och i flera andra länder uppmärksammar man Världsdagen för Downs syndrom med att Rocka sockorna. Genom att ta på dig olika strumpor 21 mars deltar du i firandet av dagen och visar att du står upp för alla människors lika värde och rättigheter.”
This.
The bias in generative AI is not a reflection of bias in society. Why would it be? It's a reflection of bias in the training data, which of course is very very far from a complete representation of the world. The training data is whatever the makers decided to throw in there and often even they have no idea how the soup is made or how well it represents any society or culture. And generative AI will then tend to amplify and exacerbate bias, because... maths.
Bonus insight: if you try to mitigate bias with inclusive prompting, it's not always that simple.
@ambivalena Ja, jag hade själv missat den i december så det kändes plötsligt jätteviktigt att hjälpa den att få spridning. 🙏
A reminder that my chart The Elements of AI Ethics is available as a free PDF download. Use it as a reference and guide in teaching, evaluations, risk assessments and mitigation strategies.
I’m currently also producing a ’worksheet for AI trials in the workplace’ as I’m hearing of many organisations experimenting with generative AI without really documenting expected and real learnings and outcomes. I expect to publish it by next week.
The harmful effects of technology can be managed and mitigated. But not if we do not talk about them. Nor if we allow ourselves to be misled to focus on misleading narratives. In the chart The Elements of AI Ethics I map out harms that we are already seeing reports of, that have been ongoing and many of which were predicted before they happened. As a tool it can provide guidance and talking points for understanding how to prioritise your work with "smart" tools, and acknowledge that all teams that deploy or make use of AI need a mitigation strategy for many different types of harms.
This chart builds on The Elements of Digital Ethics (2021). The harms in the original diagram are all still relevant, and what this chart does is provide a focused overview of the types of harm we are seeing proliferate with the ongoing advancement of AI within different industries, and especially general-purpose and generative tools. You can refer to the original chart to get an idea of how a tool like this can be used in education and project work.