Important news for AI Alignment
This news cited below means likely US Federal funding for a neglected class of AL/ML applications that could become key to the evolution of LLMs into AGIs. The problem with AI based solely on an LLM is that they don’t explicitly reason; rather, they pick up a semblance of reasoning, but often erroneously, from the training data. This problem has been experimentally highlighted by Dr. Lars Warren Ericson in his LinkedIn newsletter, GPTsplaining. https://www.linkedin.com/newsletters/gptsplaining-7057513503103873024/
Dateline March 6, 2024 On March 5, the Intelligence Advanced Research Projects Activity (IARPA) posted the Generative AI/ML Models for Math, Algorithms, and Signal Processing (MASP) request for information (RFI). Responses are due by 5:00 p.m. Eastern on March 29.
IARPA seeks information regarding innovative approaches to generative artificial intelligence (AI) or machine learning (ML) models to achieve a revolutionary leap in applications of science and engineering by generating smaller evolutionary products of MASP.
While significant progress has been made for generators of text, image, and audio (TIA) modalities, AI/ML generators for more complex sciences, including the MASP modalities, have not received the same attention. It is important to note this RFI is not for AI/ML solutions that perform such calculations of math, algorithms, or signal processing; rather, this RFI is looking for AI/ML solutions that create math, algorithms, or signal processing products themselves at the output of the generator.
Source: https://intelligencecommunitynews.com/iarpa-releases-masp-rfi/
See also https://sam.gov/opp/ff5ebc6c57954155a478a5815993f812/view to download the Request for Information.
Anyone is welcome to submit a white paper on this topic. At this stage in the procurement process, you also can email the program director, and even set up telecons or in-person meetings. This can supercharge your quest for a job in this field.
Dr. Adam L. Anderson, Program Manager
email: dni-iarpa-rfi-24-03@iarpa.gov
This has “AI alignment” in the headline, but it is not clear how this specifically has to do with alignment, and it sounds like capabilities research instead. Maybe the message you wanted to convey is something like “consider seeking funding from IARPA for you alignment project under this program which is otherwise not focused on alignment”? In any case, explanation what this has to do with alignment is missing from the post.