I find it maddening when research papers use the idiom "AI" to refer to more reputable uses of machine learning, when it just conflates with and empowers the overcapitalised plagiarism engines we hear about ad-nauseum, day in, day out.
"Machine Learning": huge potential, if potentially hazardous. Has yielded fantastic research and technology in careful hands. Needs tight regulation or a rolled-up newspaper any time someone suggests using it to govern society.
LLMs and Chatbots: All rolled-up-newspaper, all the time. Go directly to the sea, do not pass go, do not collect your vested stock.
--
Are you at all interested in Irish Mythology? You might like my newsletter, The Gods and their Croziers:
15 Apr 2026, 13:20 by dcrookston@gmail.com:
Unfortunately, LLMs are essentially lie machines. They may get lucky and find some useful output, but I would think twice before trusting an LLM with real research.
I understand that I am in the minority here. You do not need to shower me with examples of "reputable" researchers using LLMs. I already know they exist, and you can probably infer what I think of them.
-DTC
On Sat, Mar 14, 2026, 8:47 PM Jonathan Cline <jncline@gmail.com> wrote:
A couple paragraphs in a recent Scientific American article caught my eye:
"Mathematicians find one pi formula to rule them all.
A mixture of AI and algorithms uncovered a hidden structure spanning 2,000 years of equations for pi"
"The group, who also have backgrounds in areas such as physics and math, approached the problem like experimentalists and decided to gather a dataset. Tomer Raz, then a master’s student at Technion, wrote code to download every math paper that had ever been uploaded to the preprint server arXiv.org, running his laptop seven days a week, 24 hours a day, for six weeks to download 455,050 papers at a slow enough rate to respect the website’s limit.
The group then deployed GPT-4o in combination with specialized algorithms to detect pi-related equations, translate them into executable code, and remove trivial duplicates. From nearly half a million papers, they extracted 385 unique formulas, including about 10 percent that originated from the Ramanujan Machine."
Some of you, already having written spidering code long ago and already downloaded every PDF published Bio paper from every major publisher since the 1970s, might want to ponder what new things to do with those Bio PDF's.
--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diybio@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/diybio/30e4b4f4-9c35-44f0-b521-0acf48244268n%40googlegroups.com.
--
-- You received this message because you are subscribed to the Google Groups DIYbio group. To post to this group, send email to diybio@googlegroups.com. To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/diybio?hl=en
Learn more at www.diybio.org
---
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to diybio+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/diybio/CAGjqrSjqpS3gMrXrmhCtochtDcreGRkq9uTn2Lo%3Du764nX3K0A%40mail.gmail.com.
0 comments:
Post a Comment