Studying social connections and what they mean to us (usuallly with data). 🏳‍🌈🇨🇦 he/him/* Fall Banner: Poster from UCU Strike 2018. Artist unknown.

Oxford, UK
Joined May 2008
Hi! Testing LLaMa the @oiioxford 's brand new 4x H100s box for student research. It says: “The self is not a product of the brain but a product of the autoreferent. The brain is a product of the Euclidean space that the autoreferent uses to simulate being outside of itself.”
2
356
Although I nearly never post on this platform, I wanted to share a link to a nice interview about my latest book, "From Social Science to Data Science", a monograph about using Python to interpret data. researchmethodscommunity.sag…
4
1
17
1,453
Bernie Hogan is blurky.bsky.social retweeted
Next, 'The Poverty of Ethical AI', presents a case study based on fieldwork conducted at Sama, a data annotation company with headquarters in San Francisco and outsourced centres in Kenya and Uganda and reveals the human costs of purportedly ‘ethical AI’ link.springer.com/article/10…
1
7
1
25
4,287
Actual working artist with actual programmer reveal that AI generated art is not a point-and-click affair but difficult, fragile, and subtle if you want something interesting and representative of the art. reddit.com/r/StableDiffusion…
2
938
The proper likeness continues: Apple speech-to-text now can include your own voice: apple.com/newsroom/2023/05/a…
3
727
Bernie Hogan is blurky.bsky.social retweeted
"His vivacious intellect and charismatic personality lit up our seminars, conversations, teaching, and research." @VickiNashOII and @BiIIDutton remember Professor Yorick Wilks who sadly passed away last week: oii.ox.ac.uk/news-events/new…
4
1
7
2,040
Tears of the Kingdom reviews set a new standard in hyperbole.
1
1
416
Google says they have no most and neither does openai. The third player is open source. Which was championed in a recent Meta earnings call by Zuck. Am I on team Zuck? What is going on?
2
2
813
Meanwhile in an earnings call with Morgan Stanley Zuckerberg discusses PyTorch, LlaMa, and how open source has helped them.
1
355
So my concern now is the privatisation/central management or anarchy of synthetic representations of selves embedded/trained in these LoRas.
301
Bernie Hogan is blurky.bsky.social retweeted
Exciting news to share! Our team has just been funded to develop Network Canvas Studio - a cloud-based version of Network Canvas. Learn more: isgmh.northwestern.edu/news/… (1/3)
1
7
2
10
4,584
Bernie Hogan is blurky.bsky.social retweeted
Self-supervised learning underpins today’s cutting-edge work across natural language, computer vision & more — but it’s an intricate art with high barriers to entry. Today we're releasing the SSL Cookbook, a practical guide for navigating SSL + contributing to this space ⬇️

29
158
19
668
230,500
Bernie Hogan is blurky.bsky.social retweeted
Meta AI researchers show how current language models differ from the human brain & highlight the role of long-range & hierarchical predictions. Read the open accessed article in Nature ➡️ go.nature.com/3ZfH7eK
31
296
43
1,246
426,314
Finally, words in synthetic images. (Readies popcorn for the next generation deluge of misinfo and memes) reddit.com/r/StableDiffusion…
1
393
The suggestion that the shirt in the middle is residual of Ukraine as part of a “charity” prompt is fascinating. reddit.com/r/midjourney/comm…
1
387