PGF (@philfeld) header
PGF (@philfeld) profile picture

NULL

Joined September 2012
Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)Photo by PGF (@philfeld)
1/There’s a reason people read threads but balk at long walls of text thrown up by Blue Check Essayists: oral vs. written storytelling styles. Twitter evolved largely as an oral medium, stylistically. It was well suited to small bits of info one might glean in casual exchange.
1
33
1
151
17,911
What if scientific articles had a short section titled: “What I really think.” No references, no justification, just transmission of your current internal model in <250 words. Would save a lot of reading between the lines.
38
43
5
486
31,211
We are conducting a study to reduce "hallucinations", where AIs like ChatGPT generates convincing yet false information. If you would like to help, and are a researcher with publications and a work history, we'll ask the model questions about your cv. philfeldman.com/Context_test
75
Hey @zachariahCurtis. This may help your search problems
A big failure mode with RAG is retrieval quality - the retrieved context might be irrelevant, or it matches keywords but has 0 information. @cohere’s new Embed v3 model ranks documents not just on topics, but also the content quality - it’s optimized for RAG Gives SOTA performance on MTEB and returns higher-quality docs than ada-002. Thanks to @ravithejads, it’s now integrated with @llama_index. Full notebook/Colab here 📗: github.com/run-llama/llama_i… Embed v3 blog post: txt.cohere.com/introducing-e…
1
68
Honored to receive the @sosym_journal First Paper Award at #models23 @modelsconf for our paper on "Gamifying Model-Driven Engineering: The PapyGame Experience". Huge thanks to our amazing co-authors and especially the young researchers #PhD who made the experiments a reality.
1
7
1
25
1,337
HAPPY 30-50 FERAL HOGS DAY TO ALL WHO CELEBRATE
3
33
3
164
7,393
Fantastic news: the pumping of the oil from FSO Safer has just started! An enormous environmental and humanitarian disaster has been prevented! Congratulations to @DavidGressly @fathi_fahem @beleefboskalis and @UNDPYemen for an incredible job. Today is a great day for #Yemen!
19
96
21
263
77,064
The original
In the interest of balance...
1
24
151
10,989
Sneak preview. I'll be auctioning 10 of these (really, REALLY heavy) brass plaques on eBay soon to celebrate the Twitter diaspora. All profits to the charity that annoys Elon Musk the most.
8
10
1
138
59,190
Indian War elephants in mediaeval times were on another level. The original battle tanks.
5
18
1,236
Hm… so to prevent a runaway feedback loop of GIGO, humans will still need to do art and knowledge work, and produce original, correct information? So we *will* have jobs in the future? Just to be clear, here. 😂 venturebeat.com/ai/the-ai-fe…
5
22
60
6,112
Did you know there is NIH funding for available to support a student transitioning from a toxic or unsafe lab space to a new lab? Faculty can apply for “Re-entry and Re-Integration Supplements”. I had no idea this resource existed! #AcademicTwitter #phdchat
18
433
35
1,049
245,051
Also, if anyone has a good idea about what the next thing is that will be called AI, please let me know. I want to work on the new new thing :)
5
1
9
989
Enshittification in action: futurism.com/top-google-resu…
3
15
2,727
PHILADELPHIA, USA: Mapping Afrofuturism: Samuel R. Delany in conversation with Science Fiction Scholar Alex Wermer-Colan Thu, June 15, 2023 6:00 p.m. Heim Center for Cultural and Civic Engagement at Parkway Central Library libwww.freelibrary.org/calen…
3
1
6
4,645
1) Please, please, generate alerts for when usage jumps dramatically. Every platform alerts you of abnormal usage. As AI models as a service become the next technological platform, this is a must.
3
3
1
135
27,766
To fix echo chambers, we need to create feedback loops that reward people for engaging with diverse info. Automated tools can identify & intervene when an individual is searching for echo chamber-enforcing content. Search engines already do this for self-harm related searches.
1
34
The use of computers to mediate information makes it easier to fall into filter bubbles, but algorithms can also detect dangerous behaviors caused by the narrowing of content. By exposing people to diverse information, we can combat cult behavior and promote critical thinking.
1
11