David Josephs

6 hours ago, Keyvan Shahrdar said:

I don't know how accurate bard.google.com is.  I don't know if it is hallucinating and making up this information.  I thought I would share it.

You gotta stop using BARD with this my man.  Just use the search API on the metadata to search.

BARD is filling in the blanks of what it thinks you want when we need to just focus on using the API and the existing meta tag info.

It seems to be creating the documents itself since it can find the tags for it but not the actual docs.

Bottom line is BARD does not add records in reality - is it simply making #'s of WCD up since the resulting "text" is virtually identical each time or are these WCD #'s actually in the metadata itself.

The Search API response I've posted here is the correct type of response for the search... and does name these 2000 series WCD's

Is the archive hiding something, is the archive being populated with nefarious metadata...

All I know is - please do not post any more BARD responses, not one has been correct. 

Post API search results with no BARD involvement at all, please.

1 minute ago, Steve Thomas said:

David 

Yes. I think I joined the Forun in 2004, shortly after it started. It's been almost 20 years now.

Older, but not necessarily smarter.

😃

Steve

21 minutes ago, Matt Allison said:

If you can't see that the photos are clearly the tramps from Dealey Plaza in older age then you cannot be taken seriously about anything, sorry.

COINTELPRO - well done 

"If" you don't agree, "then" no one should listen

8. Invoke authority. Claim for yourself or associate yourself with authority and present your argument with enough 'jargon' and 'minutia' to illustrate you are 'one who knows', and simply say it isn't so without discussing issues or demonstrating concretely why or citing sources.

9. Play Dumb. No matter what evidence or logical argument is offered, avoid discussing issues except with denials they have any credibility, make any sense, provide any proof, contain or make a point, have logic, or support a conclusion. Mix well for maximum effect.
 

 @Matt Allison Freaking amazing right Matt? Look down below at the linked 2005 post where Robin shares a letter that suggests there are FBI interviews with these men over the ass'n weekend.   You still looking for support for your "collage"?

@Steve Thomas. I can understand that...  

Yet these three men had no record and were unknown to the DPD whereas Daniel supposedly mentioned these crimes himself.  Man you been here a long time as Robin mentions your comments in the link below.  I posted the text of the post as well...

Gotta see if the review board did anything about these questions...

Marina Oswald Porter's Letter

to the ARRB -- April 1996

Mr. John Tunheim, Chairman

JFK Assassination Records Review Board

600 E Street, NW, Second Floor

Washington, DC 20530

Dear Mr. Tunheim:

I am writing to you regarding the release of still classified documents related to the assassination of President Kennedy, and to my former husband, Lee Harvey Oswald. Specifically, I am writing to ask about documents I have learned of from a recent book and from a story in the Washington Post by the authors of the same book (as well as other documents they have described to me). The book reviews Dallas police, FBI, and CIA files released since 1992, and places them in the context of previously known information. I would like to know what the Review Board is doing to obtain the following:

1. The Dallas field office and headquarters FBI reports on the arrests of Donnell D. Whitter and Lawrence R. Miller in Dallas on November 18, 1963 with a carload of stolen U.S. army weapons. I believe that Lee Oswald was the FBI informant who made these arrests possible. I would also like to know what your board has done to obtain the reports of t he U.S. Marshal and the U.S. Army on the same arrests, and the burglary these men were suspected of.

2. the records of the FBI interrogations of John Franklin Elrod, John Forrester Gedney and Harold Doyle (the latter men were previously known as two of the "three tramps") in the Dallas jail November 22-24, 1963. All of these men have stated that they were interrogated during that time by the FBI.

3. The official explanation of why the arrest records for Mr. Elrod, Mr. Gedney, and Mr. Doyle, as well as for Daniel Wayne Douglas and Gus Abrams were placed "under federal seal" in the Dallas Police Records Division for 26 years as described by Dallas City Archives supervisor Laura McGhee to the FBI in 1992.

Just now, Denis Morissette said:

I’m not blaming or trying to ridicule you. 

sure sounds like it...

@Matt Allison has not posted much of an argument in support of his photo comparison, doesn't seem to understand there is a difference between a BARD generated answer and a Search API result, cherry-picks the NARA sites explanation of the Search API - thereby completely butchering the intent of his excerpt -  while ever so politely asks me to be less invective.

Do you actually think you or I can go into the database with a write key and ADD FAKE RECORDS?
Does that make sense to you?

So I guess the comment was you laughing at his ignorance since the API reply is right out of the data and had nothing to do with BARD.  You knew that, right?

I bent over backward to apologize for making the wrong assumption BARD was getting info from the documents themselves, it wasn't...  so we are getting away from BARD entirely.

Has nothing to do with Matt posting a BS image of 6 different men and trying to convince us there are only 3 by not offering anything in support save one woman's ID of one of the men...

Next?

30 minutes ago, Steve Thomas said:

David,

Back in 2010, there was a Thread on the Forum called, The Three Tramps.

https://educationforum.ipbhost.com/topic/16268-the-three-tramps/page/2/

On page 2 of that thread, I asked a series of questions like, Where were you when you found out Kennedy had been assassinated? and What kind of railroad car were you in when you got arrested?

I looked the answers and saw how different they were.

Steve Thomas

Thanks! Old things are new again when seen from a diff POV. 

So we are aware that each of the 3 arrests reports has ROBBERY on the "CHARGE" line, despite it being impossible they were involved in a Robbery

in the linked thread: you write:  Wise on a robbery call at the far end of south Dallas when the dispatcher told #71 to go Code 3 to the TSBD and report to the supervisor.
We do not know if he has 3 suspects, or why ROBBERY would be added to their arrest reports given the description of how they were removed from the train, and when.

Arrest time 4:00pm yet the men from the trains were arrest almost immediately after the shooting.

How many times have we seen 2 events combined into one batch of evidence...  it runs rampant all thru the evidence.. I've been writing about it for years.

7 minutes ago, Adam Johnson said:

Folks i have serious problem with this AI tracking of statements......point above the prouty information......it says Prouty was interviewed on December 13th 1963.....but:

"On December 13, 1963, Congress passed Senate Joint Resolution 137 (Public Law 88-202) authorizing the Commission to subpoena witnesses and obtain evidence concerning any matter relating to the investigation. The resolution also gave the Commission the power to compel the testimony of witnesses by granting immunity from prosecution to witnesses testifying under compulsion. The Commission, however, did not grant immunity to any witness during the investigation.

The Commission acted promptly to obtain a staff to meet its needs. J. Lee Rankin, former Solicitor General of the United States, was sworn in as general counsel for the Commission on December 16, 1963."

So prior to being given subpoena power or having a general council, Prouty was contact and asked to appear and give testimony? How would they get to prouty so quickly.....when was the commission members selected even finalised...whatvdate was their first meeting?

All this info coming from AI searches seems like a huge jumble of facts/dates and who said what....most of it seems debatable to me from what im seeing poted.

AJ

The "interview" between McCloy and Prouty was generated by Google's BARD based on finding metadata within the NARA database for Warren Commission documents that claims a record exists with this metadata:

{ "total": 1, "results": [ { "objectFilename": "Warren Commission Document 2094", "description": "Transcript of an interview with Fletcher Prouty", "content": { "text": "Fletcher Prouty was interviewed by the Warren Commission on December 13, 1963. Prouty stated that he was a former Air Force colonel who had worked in the Office of Special Operations. Prouty stated that he had knowledge of the assassination plot against President Kennedy." }, "source": "Warren Commission", "dateCreated": "1963-12-13", "dateModified": "1963-12-13" } ] }

We have 2 things going on...  BARD guesses at what the document says... which I mistakenly understood at the time as being within the inaccessible document - so the test of the interview is BARD GENERATED... but the information from the above SEARCH API results

HAS NOTHING TO DO WITH BARD @Matt Allison

And we will make sure that BARD is completely removed from this process of searching the NARA database - completely.

3 minutes ago, Matt Allison said:

Just so we all have it for the record, David- you're saying that the photos I posted of the tramps in their later years are not them, and their arrest records are fake/incorrect/not real.

Go ahead and confirm that for us.

What I am saying is the men in the photos in DP are not the three men on the arrest reports.

What I am saying is there are no records of these "several other individuals removed from the train other than the three...", that the arrest reports are timed in at 4pm when the reports for when they were arrested suggests it was prior to 1pm since while in custody officers hear about the Tippit shooting.

That these three arrest reports were not created on 11/22 - maybe read Chamber's words themselves - right?

It is up to you to prove what you posted, not for me to disprove it.   I also never said they were fake or not real... just that they do not represent the three men paraded thru DP later that day but the 3 men arrested and brought to the DPD before 1:15 that day.

So far you've offered nothing to support your posted images of 6 different men being only 3... what say you?

50 minutes ago, Denis Morissette said:

It’s your comment that made me laugh. I agree with that comment.

Of you can try to understand what I posted - 

No @Matt Allison you cannot get a key and start adding bogus records to the Archives... all you can do is leave a trial of any changes you initiate in the records themselves

bu-bye now

1 minute ago, Matt Allison said:

Maybe take a step back and reassess some of your invective

Maybe just stare at your tramps photos for similarities and let people work...

If we find that these records have been manipulated by outside forces, one would have to ask why would the Archives allow them to change the actual records...  let's see what the entire page says instead of your little cherry-picked section.

You do understand that only an entity with the key can go into records and make changes, and then those changes would forever be linked to that person's active Catalog account.

@Tom Gram

Are you saying someone or something went into the database using the API and created records when the write function only allows the addition, manipulation of tags, transcriptions and comments. 

Who might that be Matt?

Where does it say you can actually add/create NEW Archived records into the database from thin air?
I must have missed that.  In fact if you kept reading you'd see it say they are talking about the METADATA, not the records themselves...

 your invective.  

The National Archives API 2.0 is a read–write web Application Programming Interface (API) for querying the National Archives Catalog dataset. At its core, the National Archives catalog is a dataset of archival descriptions, authority records, and other information. While the browser-based catalog is a human-centered way to interface with these records, the API is an alternative method by which researchers, developers, and applications can interact with the same dataset—by following documented methods to retrieve or alter the structured data in the system.

The dataset for the catalog API contains all available archival descriptions, authority records, digitized records (images, videos, and so on) and their file metadata, and public contributions (tags, transcriptions, and comments). The API allows researchers and developers to retrieve metadata in JSON format for any given record or search results set. This method offers added flexibility in comparison to the user-centered interface in regards to advanced search or refinement options; this is because the API can search and retrieve using keywords for any field in the system, filter based on type of record, search within ranges, apply sorts, specify only particular fields to return, or any combination of these options. Bulk export of search results can then be generated. The API is also writable, which means it can be used to post and manipulate tags, transcriptions, or comments. In order to utilize write functions, users must request the user-specific API key associated with their active Catalog account by emailing api@nara.gov; for read-only functions, users must request an API key by emailing api@nara.gov. 

As a work of the federal government, all metadata is in the public domain (except excerpts, quotes, or transcription of copyrighted material); additionally, associated digital objects that represent exclusive works of the U.S. federal government can be freely reused without permission for any purpose. Note that the Catalog does contain some copyrighted works

21 minutes ago, Tom Gram said:

Interesting. If that’s an actual response from the API it looks like NARA added a “text” field with a summary of the record. I don’t know how or why the hell they’d do that for the entire ARC, but I’ll reserve my speculation until I know a little bit more about this API. Thanks David. 

https://www.archives.gov/files/citizen-archivist/images/09-05-2019-ocr.pdf

Currently, the Catalog’s new OCR engine is applied to records in either JPG or PDF format
added to the Catalog since June 2019. NARA is exploring how to retroactively process
records from before that point, but right now this feature applies to millions of pages!

@Keyvan Shahrdar  @Steve Thomas

15 minutes ago, Matt Allison said:

Yes, it's called bullsh*t.

Thanks for proving you do not know the difference between a BARD answer and a Search API result.

The Q&A's created by BARD for those WCD's are based on the metadata it found.

I apologize for thinking the Q&A text was found in the metadata or in the document which we couldn't access.  Doesn't change the fact the arrest reports conflict with the stories of their arrest

Let me know how that API works for you.  We've got a few others working on it too.

And we'll keep BARD out of it.  

Just now, Tom Gram said:

Found this. Gonna request a key and play around with it a bit:

https://www.archives.gov/research/catalog/help/api

Yes, that was what @Keyvan Shahrdar showed me...  

3 minutes ago, Tom Gram said:

Interesting. If that’s an actual response from the API it looks like NARA added a “text” field with a summary of the record. I don’t know how or why the hell they’d do that for the entire ARC, but I’ll reserve my speculation until I know a little bit more about this API. Thanks David. 

When you put into a spreadsheet you have a "title" column with descriptive text.  Might be taken when they OCR the title page which usually has that info

Thank you for helping out.   I'd like to corroborate that the search API into this database returns that result... or if somehow BARD is involved... yet isn't BARD doing its thing AFTER it find the metadata, it doesn't just create metadata does it?

@Matt Allison @Steve Thomas @Tom Gram

Amazing what you can find by looking

https://developer.collibra.com/tutorials/getting-started-with-the-search-api

Just now, Matt Allison said:

Yes, it's called bullsh*t.

Do you know what it is or not?

Technically

@Matt Allison

Do you know what this is?

FYI - 
{ "total": 1, "results": [ { "objectFilename": "Warren Commission Document 2094", "description": "Transcript of an interview with Fletcher Prouty", "content": { "text": "Fletcher Prouty was interviewed by the Warren Commission on December 13, 1963. Prouty stated that he was a former Air Force colonel who had worked in the Office of Special Operations. Prouty stated that he had knowledge of the assassination plot against President Kennedy." }, "source": "Warren Commission", "dateCreated": "1963-12-13", "dateModified": "1963-12-13" } ] }

2 minutes ago, Tom Gram said:

So you can just request a free API key from NARA and get programmatic access to a database of RIFs and metadata that isn’t otherwise publicly available? NARA is (supposedly) working on digitizing the whole collection IAW Biden’s Dec. 2022 memo, but as far as I know the only currently available “database” of RIFs is the 2021 Excel spreadsheet.

Does anyone have a link? Part of my job involves integrating external APIs with AI so I’d love to tinker with this a bit. @Keyvan Shahrdar what is the response schema from the NARA API? Is it just the same fields in the Excel Spreadsheet in JSON? Is there like a .pdf url, or encoded .pdf, or something like that? I’m curious how you get to the actual content of the record since some of the Bard responses you posted look like it retrieved more than just metadata. 

Thanks Tom... the more in this the better I think.  I am leaving this to @Keyvan Shahrdar  

@Steve Thomas

Is this the result of the API search?

FYI - 
{ "total": 1, "results": [ { "objectFilename": "Warren Commission Document 2094", "description": "Transcript of an interview with Fletcher Prouty", "content": { "text": "Fletcher Prouty was interviewed by the Warren Commission on December 13, 1963. Prouty stated that he was a former Air Force colonel who had worked in the Office of Special Operations. Prouty stated that he had knowledge of the assassination plot against President Kennedy." }, "source": "Warren Commission", "dateCreated": "1963-12-13", "dateModified": "1963-12-13" } ] }

44 minutes ago, Matt Allison said:

A lot of what is being posted here is total nonsense, which doesn't help anyone.

Once again, these AI chatbots you people are referencing are not accurate in the slightest; they fabricate things whole cloth:

https://en.wikipedia.org/wiki/Hallucination_(artificial_intelligence)

I posted the modern photos of the three tramps marched through Dealey Plaza above; we have their names, the arrest record and photos match. They were not conspirators or involved in the assassination in any way.

We've had enough fake news from the Fascist Right-Wing in this country the past few years, let's not add to that problem here, OK?

Matt...

While I have enjoyed reading your posts now for a while, your insistence on this definitive identification based on, what?  Can you post the facial analyses they did - given we can now scan 2d, create 3d and do exact comparisons based on the mathematical relationships between facial features.  If that was done, please forgive me and provide the link.

If not, and this is based on pure visual, and "trust me that's me" confessions, you'll have to excuse my remaining a bit skeptical.  Especially since I have been thru mountains of WISE/TRAMPS related documentation, have a brain, and see this as part of an overall cover-up, confusion effort that extends in virtually every direction that afternoon and thru the weekend.

As to "AI chatbots" - we are using a direct API into the NARA database that comes back with metadata results which match the BARD results when looking into the same database.

The metadata is there describing these WCDs which have never been seen and suggest a number of WCD's well in excess of the 1555 at MFF.  AI does not make up metadata within a search, nor does it create a Q&A without if being asked.  I stand corrected. Using the format of WCR Q&A it has Specter asking Q's which are not there.  Doesn't change the fact the metadata is still there

FYI - 
{ "total": 1, "results": [ { "objectFilename": "Warren Commission Document 2094", "description": "Transcript of an interview with Fletcher Prouty", "content": { "text": "Fletcher Prouty was interviewed by the Warren Commission on December 13, 1963. Prouty stated that he was a former Air Force colonel who had worked in the Office of Special Operations. Prouty stated that he had knowledge of the assassination plot against President Kennedy." }, "source": "Warren Commission", "dateCreated": "1963-12-13", "dateModified": "1963-12-13" } ] }

I'm sorry to disagree with your conclusions about these men and the events surrounding them.  We are trying to research the situation, while I have yet to see a single realistic presentation of research or data to support all three men being who you say they are.  Just you saying so.

You could always just ignore this thread if you disagree and let us do our work
You could always post something to convince me I'm wrong

or you can continue to throw backhanded insults in the thread showing our efforts 
in an effort to accomplish what exactly -

that you don't like people disagreeing with you and being able to prove it?

==

I got no beef with you...  I am working with @Steve Thomas and @Leslie Sharp on connecting years of research between the three of us with the additional help of @Keyvan Shahrdar on the web/API side of the equation.

So no Matt... not OK.   James Files is fake news.  Judy Baker is fake news.

The duplicity found in so much of the JFK evidence are the keys to working back to those in control of the operation.

If you haven't seen by now how things are connected within a huge spider's web of people and ideology, greed and hate... let's not add THAT problem here, OK?

1 minute ago, Steve Thomas said:

David,

One thing we have to is sort out the conflicting reports of the DPD personnel.

 

1992 FBI interview of Marvin L. Wise. FBI 124-10273-10440

Page 2

https://www.oocities.org/capitolhill/senate/5439/Susp_124-10273-10440.html

 

Steve Thomas

This makes sense since the three in the WISE images are not the 3 with arrest reports... the 1st set were already in the boxcars, while the 2nd set moved along the RR tracks and were picked up later in the day.   

But why the parade?  why Lansdale is allowed to walk to & past them?  With the 2 sets of tramps (like all the dualities), the different DPD personnel can all be telling the truth while conflicting at the same time.

https://theoswaldcode.com/index2.htm.  This shows the arrest reports and the report of another DPD officer who talks about the later arrival of the three men brought to Decker by Wise.  Sheriff Harold Elkins claims though, that he takes the three DP tramps to Fritz and the City Jail after WISE drops them off.

I'd think that was to tie the two different sets together.  

And remember Bill Middle says he was not at work on the 22nd at all. FWIW

1 hour ago, Leslie Sharp said:

I saw and I responded and my computer crashed!

Will redo and send again.

But are you saying that Prouty was responsible for the specific diversion that included the parade of three tramps?  Isn't that the essence of this exercise?

Not at all Leslie.  The documents related to the tramps - and whether the 3 arrest reports were done on that weekend or after the fact, and whether the 3 men in the photos have anything to do with the three men on the arrest reports.

Add now that BARD is claiming to find WCD #'s in the NARA release which are not viewable from the release but only from the meta-tags embedded.

AND these reports completely contradict what the arrest report says.  That there is yet another WCD # between the WCR's 1555 and BARD's 2312, #2094 which is the Q&A between Prouty and McCloy.

@Keyvan Shahrdar confirmed that the API search of the database only offers 2094 and 2312 as WCD#'s past 1555 yet with the numbering convention used, we should see WCD #'s from 1556 thru 2311, to arrive at 2312.

The API results include these WCD #'s so they are listed in the meta data of the database related to some record yet as I understand, the file themselves need not be visible or accessible for the API to find the meta data.

Keyvan can correct me if wrong

Just now, Leslie Sharp said:

@David Josephs Is this applicable to the question of who ran the three tramps diversion?  Prouty?

All part of the same spider web...  btw, I sent you an email with what I think is a very important question before we discuss on the forum about the Sept 16 entry and a pdf key

35 minutes ago, Michael Griffith said:

HSCA photographic experts

That's an oxymoron. @James DiEugenio 

22 minutes ago, Leslie Sharp said:

Paul, that's completely off base. If we can determined the role the tramps played, and who played them, it leads up the food chain.  Otherwise, it seems to me the tramps remain an isolated incident that some, including Ms. Gibson have spent an inordinate amout of time identifying by name, but to what avail?  

@Steve Thomas @Keyvan Shahrdar

Because we are finding WCD #'s 2094 and 2312 which are up to 800 #'s more than the last WCD # 1555 when BARD is asked to look into specific people and when the database is accessed directly via API.

2312 shows an excerpt from a SPECTER-ABRAMS Q&A where Abrams states this whereas the Chambers arrest report and Chambers' FBI interview states they were asleep in a boxcar when police woke them up and took them in.

My concern is not that BARD is broken, but that it is finding evidence of 100's  of WCD #'s we've never seen.

Abrams: Well, I was standing on Elm Street when the shots were fired. I saw two men running towards the train tracks after the shots were fired.

2094 is McCloy asking Prouty questions where Prouty says:

Prouty: Sure. The Office of Special Operations was a secret unit within the Air Force that was responsible for carrying out covert operations. We were involved in a lot of different things, including assassinations.

McCloy: Assassinations?

Prouty: Yes. We were involved in the assassination of Patrice Lumumba in the Congo, for example.

It would make sense to hide both of these as they would cause real problems - yet there are many, many items in the WCD which conflict with the WCR conclusions in broad daylight...

Steve - I think we're gonna need all your expertise in searching and digging here.. these results are quite strange

I don't know... the API for the NARA database of the released files is the program which allows us to query the database.

With the key @Keyvan Shahrdar got from NARA he, as a developer, is able to access the database directly by asking it questions thru BARD.

Sure, here is the transcript of Warren Commission Document 2312:  (no such WCD)
Date: November 23, 1963
Location: Dallas, Texas
Interviewer: Arlen Specter
Interviewee: Gus Abrams

Specter: Can you tell us what you saw?
Abrams: Well, I was standing on Elm Street when the shots were fired. I saw two men running towards the train tracks after the shots were fired.

There are no records of Specter interviewing Abrams... And the Chambers arrest reports and his testimony shows he was not there when they arrested them but was told they by the tramps themselves they were asleep in a boxcar when police woke them.

He also showed a McCloy-Prouty interview - that also is not listed anywhere.   WCD1555 is the last WCD.  Does this mean there are 800 WCD not released at a minimum?
DJ

Date: December 13, 1963. WCD 2094. 

Location: Washington, D.C.

Interviewer: John J. McCloy

Interviewee: Fletcher Prouty

McCloy: Mr. Prouty, thank you for coming down today. We appreciate your help.

Prouty: It's my pleasure.

McCloy: I understand that you were a former Air Force colonel who worked in the Office of Special Operations.

Prouty: That's correct.

McCloy: Can you tell us about your knowledge of the assassination plot against President Kennedy?

Prouty: Well, I can't say that I had any specific knowledge of the plot. However, I did have some general knowledge of the activities of the Office of Special Operations.

McCloy: Can you tell us more about that?

Prouty: Sure. The Office of Special Operations was a secret unit within the Air Force that was responsible for carrying out covert operations. We were involved in a lot of different things, including assassinations.

McCloy: Assassinations?

Prouty: Yes. We were involved in the assassination of Patrice Lumumba in the Congo, for example.

McCloy: I see. And you believe that the Office of Special Operations was involved in the assassination of President Kennedy?

Prouty: I believe it's possible. I know that the Office of Special Operations had a lot of enemies, and I wouldn't put it past them to have been involved in the assassination.

McCloy: Thank you for your time, Mr. Prouty. We appreciate your help.

Prouty: You're welcome.

This transcript shows that Fletcher Prouty, a former Air Force colonel who had worked in the Office of Special Operations, believed that the Office of Special Operations was involved in the assassination of President Kennedy. However, Prouty did not have any specific knowledge of the plot.

The Warren Commission did not investigate Prouty's claims, and they ultimately concluded that Lee Harvey Oswald acted alone in assassinating President Kennedy. However, Prouty's claims have been the subject of much speculation and debate.