Allow me to be the first on Canadian Mastodon to say:
This is not going to end well for anyone.
"Treasury Board President Anita Anand says she wants to increase the use of artificial intelligence in Canada's public service but insists there's no question of using AI tools like ChatGPT for confidential information like cabinet documents.”
Aside: In a way, the rise of "AI" truly has emphasized the word “Artificial" in Artificial Intelligence. lol
On answering programming questions: "We found that 52 percent of ChatGPT answers contain misinformation, 77 percent of the answers are more verbose than human answers, and 78 percent of the answers suffer from different degrees of inconsistency to human answers." #AIhttps://werd.io/view/6654cee735684fb99a004842
Es gibt viele Beispiele die aufzeigen, wie die vielbeschworene Digitalisierung und "Smartifizierung der Welt" vor allem eine Transformation in eine noch autoritärere und repressivere Zukunft bedeuten kann.
Die biometrische Überwachung/Verfolgung von Frauen in Iran und von Kriegsdienstverweigernden in Russland sowie die Zero-Covid-Politik des chinesischen Staates sind offensichtliche Beispiele dafür. Und auch im Überwachungskapitalismus des "demokratischen Westens" werden solche Entwicklungen schon seit einer ganzen Weile auf hohen Ebenen herbeigesehnt, wie z.B. das Dokument "Smart City Charta" von Bundesinstituten und dem Bundesumweltministerium aus dem Jahr 2017 zeigt.
Eine schön deutliche Einordnung dieses Dokuments (im Rahmen einer vergangenen Veranstaltung) findet sich hier:
Gesichtserkennung in Großbritannien: Unschuldige verdächtigt
"In London und Manchester wurden zwei Personen aufgrund einer Gesichtserkennungs-KI zu Unrecht wie Kriminelle behandelt.
(...)
Die Metropolitan #Police nutzt Kameras, um Tausende von Gesichtsbildern zu erfassen und diese mit Personen auf einer polizeilichen Beobachtungsliste abzugleichen..."
While AI is expected to be top of the bill at the unveiling of iOS 18 at WWDC on June 10, it seems Apple could be keeping things a little more basic than some of the other AI announcements made recently.
Mark Gurman’s Power On newsletter suggested Apple’s approach to AI will focus on “being practical,” reports @theverge, with features like voice memo transcriptions, auto-generated emojis, and AI-enhanced photo edits and Spotlight searches. The biggest announcement? A rumoured partnership with OpenAI, which will allow ChatGPT to “integrate more deeply” into the Apple operating system. Here’s more.
There was a time when professional design & presentation helped separate the fly-by-night from the well-established. Even those who couldn’t articulate why or how one ad (commercial, website) was better than another could tell one was. It was no guarantee, but it was a helpful clue.
Facebook stripped all that away. On Facebook, the guy selling stolen shit from his van looks as legit as Target or Home Depot.
Google’s AI-generated search results do this, too. #AI#Google
@rafadc
Ohhh nice! Might make a follow up blog post explaining how these things are done on the fly to acquire semiconductor information at the nanoscale. With a technique called Scanning Kelvin Probe Microscopy used with an #AFM with an electrically conductive tip and a lock-in amplifier.
I wonder if #LLM or other #AI use #FPGA lock-in amplifiers for their Fourier transforms. Coding must be really slow.
"The significance of this development is profound: If AI provides the answers to all the searches that you’re looking for, there is no need to click on the source articles that provide the answers. If you don’t click on the source articles, the publishers do not receive any ad revenue, and if the publishers do not receive ad revenue, they cannot pay their writers to provide the answers for Google’s AI to steal."
It’s amazing the things some people will say to you if you are not 100% aboard with #ai.
Do these people really believe in Roko’s Basilisk? Is there a quota of tiny twitter/mastodon accounts to bully in order to not get tortured for eternity?
"AI headphones let wearer listen to a single person in a crowd, by looking at them just once"
Basically, you look at a person in a crowd (maybe at a party, or a crowded restaurant) and the #AI focuses on that person, does noise cancellation on all other sounds.
To me, THIS is the great side of AI, where we should be focusing. Obvious benefits, no big disbenefits.
Interesting data from a new edition of the Foundation Model Transaprency Index - collected six months after the initial index was released.
Overall, there's big improvement, with average score jumping from 37 to 58 point (out of a 100). That's a lot!
The interesting fact is that researchers contacted developers and solicited data - interactions count.
More importantly, there is little improvement, and little overall transparency in a category that researchers describe as "upstream": on data, labour and compute that goes into training. And "data access" gets the lowest score of all the parameters.
Having massively over-promised on an immature technology with really quite a limited range of productive use-cases, #AI firms are now contriving to deliver.
However, because there is no tangible product, they are using their financial might and institutional power to bludgeon the Internet and society as a whole into an AI-shaped hole.
It's like a company that makes anti-diarrhoea meds pumping laxatives into the water-supply to engineer enough demand to justify billion of investment.
Can Generative #AI improve social science? https://www.pnas.org/doi/10.1073/pnas.2314021121?TOC_v121_i21 That's a reasonable discussion of the advantages and risks, including limited reproducibility of using LLMs.
An open AI infrastructure for LLMs seems reasonable. Maybe funding agencies can support first steps in this direction @academicchatter