Author Archives: pgfeldman

Phil 11.10.2022

See if I can get BSO tix for Friday -yes!

Creative Writing with an AI-Powered Writing Assistant: Perspectives from Professional Writers

  • Recent developments in natural language generation (NLG) using neural language models have brought us closer than ever to the goal of building AI-powered creative writing tools. However, most prior work on human-AI collaboration in the creative writing domain has evaluated new systems with amateur writers, typically in contrived user studies of limited scope. In this work, we commissioned 13 professional, published writers from a diverse set of creative writing backgrounds to craft stories using Wordcraft, a text editor with built-in AI-powered writing assistance tools. Using interviews and participant journals, we discuss the potential of NLG to have significant impact in the creative writing domain–especially with respect to brainstorming, generation of story details, world-building, and research assistance. Experienced writers, more so than amateurs, typically have well-developed systems and methodologies for writing, as well as distinctive voices and target audiences. Our work highlights the challenges in building for these writers; NLG technologies struggle to preserve style and authorial voice, and they lack deep understanding of story contents. In order for AI-powered writing assistants to realize their full potential, it is essential that they take into account the diverse goals and expertise of human writers.

SBIRs

  • More reading. Need to search each paper for “loop”, “centaur”, and “team” and check those paragraphs at least. As you might think, the reality is more complex. All the papers have some parts of the concepts, but they often don’t use the terms
  • Chat with Aaron. Really good. I think I was able to explain my concept. We’re going to write some sections and worry about the structure later
  • 9:15 standup
  • 2:00 Presentation. Went ok. Steve needs to join Toasmasters

Book

  • Roll in more edits – DONE!
  • Set up new template?

GPT Agents

  • Add to the spreadsheet a week of numbers with min/max/avg to see what the pull size should be – done

Phil 11.9.2022

Fooling around with Mastodon a bit. The lack of advertising and the associated visual clutter is… remarkable

Book

  • Rolling in edits
  • Replied to my new Editorial Project Manager. This is starting to feel very real and scheduled

SBIRs

  • More reading. Going to work through human factors

GPT-Agents

  • Having an interesting discussion with Jack Chen about synthetic story spaces
  • Continuing documentation while watching Twitter implode. I’ll probably rework the tools to wok with the Reddit API
  • 4:00 meeting

Phil 11.8.2022

Election day! Absolutely no idea how any of this is likely to play out

Also, create a Mastodon account? I probably have enough info at this point. Applied at fediscience.org – done and set up! I have even tooted

SBIRs

  • 9:00 planning meeting
  • 10:00 MC meeting?
  • Paper
    • Finish populating annotated bibliography
    • Add category and “LMN ranking” to spreadsheet
    • Read top 3(?) papers for each
    • Search all papers for Hi/otL statements and add those quotes to the spreadsheets.
    • Tempted to do some embedding clustering, but that’s overkill
  • Add a task to Rukan to check out MinGPT as possible NN for out modules? Done

Book

  • Roll in changes

GPT Agents

  • More documentation – finished TweetEmbedExplorer
  • Start Twitter pull?

Phil 11.7.2022

Move hotel to January

SBIRs

  • Adversarial Policies Beat Professional-Level Go AIs
    • We attack the state-of-the-art Go-playing AI system, KataGo, by training an adversarial policy that plays against a frozen KataGo victim. Our attack achieves a >99% win-rate against KataGo without search, and a >50% win-rate when KataGo uses enough search to be near-superhuman. To the best of our knowledge, this is the first successful end-to-end attack against a Go AI playing at the level of a top human professional. Notably, the adversary does not win by learning to play Go better than KataGo — in fact, the adversary is easily beaten by human amateurs. Instead, the adversary wins by tricking KataGo into ending the game prematurely at a point that is favorable to the adversary. Our results demonstrate that even professional-level AI systems may harbor surprising failure modes. See this https URL for example games.
  • 9:00 Sprint Review
  • More reading
  • Used the LMN tools to figure out what to emphasize and find more papers

GPT Agents

  • More documenting
  • Figure out some keywords for various groups and start pulling tweets. I think 10k per group a week would be manageable.
    • Watching Twitter implde. Maybe I should just use the pushshift API?
  • Reply to First line with some examples

Book

  • Meeting with Brenda

Phil 11.4.2022

Sheesh – still don’t feel particularly good

10:00 Dentist

Large Language Models Are Human-Level Prompt Engineers

  • By conditioning on natural language instructions, large language models (LLMs) have displayed impressive capabilities as general-purpose computers. However, task performance depends significantly on the quality of the prompt used to steer the model, and most effective prompts have been handcrafted by humans. Inspired by classical program synthesis and the human approach to prompt engineering, we propose Automatic Prompt Engineer (APE) for automatic instruction generation and selection. In our method, we treat the instruction as the “program,” optimized by searching over a pool of instruction candidates proposed by an LLM in order to maximize a chosen score function. To evaluate the quality of the selected instruction, we evaluate the zero-shot performance of another LLM following the selected instruction. Experiments on 24 NLP tasks show that our automatically generated instructions outperform the prior LLM baseline by a large margin and achieve better or comparable performance to the instructions generated by human annotators on 19/24 tasks. We conduct extensive qualitative and quantitative analyses to explore the performance of APE. We show that APE-engineered prompts can be applied to steer models toward truthfulness and/or informativeness, as well as to improve few-shot learning performance by simply prepending them to standard in-context learning prompts. Please check out our webpage at this https URL.

How Online Mobs Act Like Flocks Of Birds

  • A growing body of research suggests human behavior on social media is strikingly similar to collective behavior in nature.

Book

  • Done rolling in current edits
  • Review and sign contract
  • Spend some time working on better terrain. Done!

SBIRs

Phil 11.3.2022

Had a reaction to the latest booster. I feel like an elephant sat on me

Good thread on online radicalization of a primed subject

Graphika leverages AI to reveal and study online communities. We are the best in the world at analyzing how online social networks form, evolve, and are manipulated.

SBIRs

  • Twitter dev conference was canceled. Trying to get my funds back
    • Credit for SW
    • Can move the hotel into Dec/Jan
  • Working on MORS paper

GPT Agents

  • Half a meeting last night. Time zone issues. We might look at the changes in right-wing and left-wing interactions on Twitter pre and post-Musk
  • Adding spreadsheet output to tweet counts and Wikipedia counts – Done!

Book

  • Rolling in changes
  • Promised to have the contract back by Friday COB

Phil 11.2.2022

I think the quality of Twitter is dropping

SBIRs

  • One of the things to add as suggestions is a model-training facility with dedicated staff. The facility exists to train up to very large models that are resilient to attack (think of a GPT-3 ensemble), and staffed with people who study how models fail. The facility also trains faulty models (mode collapse, overfitting, etc) that can be invisibly swapped in for verified (whatever that means) models so that AI pilots can learn to recognize degraded model behavior. Lots of simulators that allow users to be trained in high-stress situations to adapt to failing models.
  • Since the facility trains many models, it will be possible to train meta models that can understand which hyperparameters and data sets produce effective models, and how to degrade them. This will be extremely valuable as AI/ML continue to move into more roles that were previously occupied by highly trained and/or experienced people.
  • Find chess paper that shows AI/human tams out-perform AI-only

Book

GPT Agents

  • More documentation
  • Need to figure out some keywords for watching Twitter pre/post Musk
  • 4:00 Meeting

Phil 11.1.2022

Took some much needed PTO

Tasks

  • 1:45 booster
  • Vote!

SBIRs

  • Write paper
  • 9:15 standup
  • 10:00 Weekly meeting

Book

  • 4:00 Meeting

GPT Agents

  • Documentation

Phil 10.27.2022

Homicide is a leading cause of death for pregnant women in US

  • Women in the US are more likely to be murdered during pregnancy or soon after childbirth than to die from the three leading obstetric causes of maternal mortality (hypertensive disorders, haemorrhage, or sepsis).1 These pregnancy associated homicides are preventable, and most are linked to the lethal combination of intimate partner violence and firearms. Preventing men’s violence towards women, including gun violence, could save the lives of hundreds of women and their unborn children in the US every year

Book

  • Rolling in more edits
  • “In the meantime, some great news—the project is now officially approved! I’m now just waiting for a draft of the publishing agreement, so as soon as that’s ready I will send it over.”

GPT Agents

  • Finish db access and build a view to see the text and meta info
  • Here’s a view that I created to link multiple rows of key/values to a root result. Really proud of this:
create or replace view test_view as
    select tt.*, ttd_k.value as keyword, ttd_c.value as created, ttd_l.value as location, ttd_p.value as probability
    FROM table_text tt
    inner join table_text_data ttd_c on tt.id = ttd_c.text_id and ttd_c.name = 'created'
    inner join table_text_data ttd_k on tt.id = ttd_k.text_id and ttd_k.name = 'keyword'
    inner join table_text_data ttd_l on tt.id = ttd_l.text_id and ttd_l.name = 'location'
    inner join table_text_data ttd_p on tt.id = ttd_p.text_id and ttd_p.name = 'probability';

SBIRs

  • 9:15 standup
  • 11:30 Touch point
  • Work on paper. Reading the DSIAC-BCO-2022-216 report, which is a lot less than I thought it would be considering the page count, but has some good stuff in it.
  • Add example paragraph to rationale section
  • Add mute() method and flag – done. Also added the publish
  • Make publish work with base classes

Phil 10.26.2022

To Fight Misinformation, We Need to Teach That Science Is Dynamic

  • Science is a social process, and teaching students how researchers work in tandem to develop facts will make them less likely to be duped by falsehoods

GPT Agents

  • Model Explorer
    • Make it so that seed re-use happens only on new text – Done
    • Add ‘%’ to percentage formatting – Done
    • add parsing of probes to split on commas – Done
    • Get experiment loading working – Done
    • Wire up to db – Not done
      • Created table_text and table_text_data in the gpt_experiments schema
  • 4:00 Meeting today – discuss AI combat paper?

Book

  • Responded to Katy’s email with ideas for website
  • Continue with edits

SBIRs

  • Got a good first pass on Rationale
  • Working on Piloting AI paper
  • Found a good resource with stopkillerrobots.org

Phil 10.25.2022

https://www.nytimes.com/2022/10/24/us/politics/russia-dirty-bomb-west-ukraine.html
Tried the keyword explorer out on this

SBIRs

  • Sprint planning meeting – done. Started on taks
  • 3:00 pm MCWL meeting. Delayed until Friday

GPT Agents

  • Finish putting in the counters and wire up the model parameters
    • Counters are working and now use regexes
    • Wired up parameters
    • Wired up “save experiment” as json. Need to load back in now
  • Start documentation
  • Added these and other issues to the GitHub project
{
    "probe_str": "]][[text: ",
    "description": "This is a description",
    "max_len": 128,
    "top_k": 50,
    "top_p": 0.95,
    "num_sequences": 10,
    "batch_size": 1,
    "seed_flag": false,
    "model_path": "D:/Development/models/ivermectin_paxlovid"
}

Book

  • Start folding in new edits
  • Write up website thoughts for Katy

Phil 10.24.2022

SBIRs

  • 9:00 Demos
  • 2:00 MDA Meeting
  • Wrote up request to get hours for MORS paper
  • Wrote up some scenarios for Rukan

GPT Agents

  • Cannot figure out how to select and highlight text so I’m dropping the HTML frame
  • Adding the percentage tracking

Book

  • 4:30 Meeting with Brenda

Phil 10.21.2022

Call BSO about ticket for Saturday

Book

  • Finish up Money section. Maybe introduce the idea of self-grounded spaces? Either add a section to Belief is a Place, and continue here, or introduce here and continue on with the rest of the book? Not sure

SBIRs

  • Add method that handles a particular class and produces a tab in a spreadsheet
  • Add overridable “publish” method in BaseController?
  • Comment changes in SharedObjects and BaseController
  • Document for MDA Q3
  • Slides for sprint demos
  • 1:00 Meeting

Phil 10.20.2022

The Slippery Slope: How Small Ethical Transgressions Pave the Way for Larger Future Transgressions

  • Many recent corporate scandals have been described as resulting from a slippery slope in which a series of small infractions gradually increased over time (e.g., McLean & Elkind, 2003). However, behavioral ethics research has rarely considered how unethical behavior unfolds over time. In this study, we draw on theories of self-regulation to examine whether individuals engage in a slippery slope of increasingly unethical behavior. First, we extend Bandura’s (1991, 1999) social-cognitive theory by demonstrating how the mechanism of moral disengagement can reduce ethicality over a series of gradually increasing indiscretions. Second, we draw from recent research connecting regulatory focus theory and behavioral ethics (Gino & Margolis, 2011) to demonstrate that inducing a prevention focus moderates this mediated relationship by reducing one’s propensity to slide down the slippery slope. We find support for the developed model across 4 multiround studies. (PsycINFO Database Record (c) 2014 APA, all rights reserved).

SBIRs

  • Working on showing controller commands, states, and responses – done!
  • Add SharedObject “queries” that create spreadsheets for specified (maybe an array?) types. Got the basics working. Need to break out a method to handle a type and a writer

Book

  • 10:00 Meeting! We are going ahead on the book! Submission of the full book at the beginning of December, and a decision 1-2 weeks after that!
  • Continue to roll in changes

Phil 10.19.2022

GPT Agents

  • Use the GoogleExplorer as a template for prompt interactions, and try some repeat interactions! Parsed output can easily be rendered as html which should be a nice touch. The text is the main piece, and the meta attributes are <ul> elements
  • Add param list for GPT generation
  • Add buttons to save current output to db
  • Add textarea for description
  • 4:00 Meeting
  • Progress for today!

SBIRs

  • More writing

Book

  • Reply to Katy, and try to set up a meeting? Done. 10:00am tomorrow!
  • Fold in more changes
  • I realize that fiat money is also a self-grounded belief space. I think that means that money, organized religion, and constitutional governments are all related. They are also distinctly different from ungrounded belief spaces. Added a note to the text