The affidavit is available here: https://storage.courtlistener.com/recap/gov.uscourts.dcd.287327/gov.uscourts.dcd.287327.1.1.pdf
One thing to note is previously the FBI had stated that cellphone data was not available due to corruption (https://vinnews.com/2023/06/19/in-stunning-testimony-former-fbi-official-says-j6-pipe-bomber-could-not-be-located-due-to-corrupted-phone-data/). However, now in the affidavit cellphone data is able to place Cole at the scene of the crime. That seems to indicate that someone is either lied to congress or someone screwed up their job. Because its possible the FBI queried one cellular network for data and the network claimed the data was corrupted and then the FBI made no further queries to other providers and then went to congress and claimed the data was corrupted. This is probably not technically lying (its ok to be deliberately mislead your audience as long as you don't make a clear lie!) but this would be gross incompetence. The data was clearly available because now we know the government was able to access it and these were provider records (presumably, from the carrier) not information from Cole's own cellphone. I guess its possible in the affidavit the government is being misleading when they say 'provider' and they mean some kind of other third party location data ('google') but I feel like that would be misleading the court because they refer to Provider's 'cell site records'.
The congressional testimony also claimed the provider in question was aggressively deleting cell site records (presumably to protect the privacy of their customers and to save money). But now we find out that after 4 years the cell site records are available. I presume the FBI was able to collect cell site records from other providers at the time but for some reason was able to claim to congress because they were missing information from one provider that it was not possible to use the records to identify suspects. Of course its also possible that some providers are keeping cell site records for a 4 year period but that seems very unlikely due to the storage costs. The other possible explanation is the government has another crazy data collection program where they are thieving huge amounts of metadata from the cellular networks and storing this somewhere and this is classic government lying to the court to hide their data collection methods.
I presume any trial is going to be a complete shit show where the defence is going to ask how these cellphone records mysteriously appeared after law enforcement officials testified to congress that they were not available. I also now have a bit more empathy for congressional witnesses who claim they are unable to talk about ongoing investigations.
Also, I assume this is not the full evidence of the State's case against Cole. But it has Steve Baker gait analysis vibes. This guy bought some stuff related to bomb making and then his cellphone was in the area of where the pipe bombs were planted. Hopefully, there is some more solid evidence linking him to the crime. Though, the cellphone data would seem to be much stronger than the gait analysis since presumably there would not be that many people in that area at the time, whereas presumably there are many people would match the gait.
benmmurphy
4d ago
The recent React code execution bug seems expose the current shallow reasoning behaviour of LLMs. So the react framework has had a serious security issue recently involving serialisation which apparently can cause arbitrary code execution. Some people have tried to generate exploits based on the patch using LLMs but this has failed. For example: https://github.com/ejpir/CVE-2025-55182-research/blob/main/TECHNICAL-ANALYSIS.md which assumes that the server would whitelist very unsafe modules for remote access (https://react2shell.com/). Of course the criticism of the LLM in the situation should be taken with a grain of salt because whoever is driving the LLM seems to have no idea what they are doing. It might be possible that someone who knows what they are doing is able to drive the LLM to a solution, but then question is how useful is the LLM. One advantage of the LLM in this situation is it allows someone with very little context of React and their bizzaro RPC protocols to quickly generate valid payloads that would generate weird behaviour and this could be used to generate a chain that would lead to code execution.
people who commit crimes and get caught make bad decisions. this is the simple explanation for everything. of course in this particular situation the timing is very weird because you have the grifter news site making bombshell accusations about capitol police being involved in the pipe bombing. but there is even a explanation for the timing which is grifter news site is making FBI look bad even if the grifter story is false so FBI now allocates resources on this particular case instead of whatever 'higher' priority the FBI would otherwise use their resources.
also, if this is an FBI coverup this confirms the republican party is the washington generals.
i suspect such a thing is unlikely to appear because the lender is probably repackaging the mortgages to fannie mae (or something similar) and that likely violates the agreement between this third party and the lender. but it might be possible someone trying to bump their numbers gave a wink to James that she could do this.
Apparently, a blogger may have originally raised the problem with James's loan and he has a write up about the ongoing court case here: https://whitecollarfraud.com/2025/11/18/letitia-jamess-motion-to-dismiss-backfires-her-own-exhibit-proves-the-fraud-she-claims-doesnt-exist/
benmmurphy
21d ago
the window thing might also be because when they get old if they open quickly then decelerate quickly then there is a risk the pane of glass will pop out and cut someone below on the street in two. this happened in our office building from the 7th floor onto a busy street but by some miracle no-one was injured.
isn't this now a genius strategy to get your company bought by elon. also, the article reminded me a lot of geeks, mops and sociopaths (https://meaningness.com/geeks-mops-sociopaths)
Chess.com has now concluded the Super PogChamps series where content creators play each other in a tournament to find a winner. Each content creator is teamed with a higher level player for chess coaching and game preparation. The Super PogChamps series is slightly different from the normal PogChamps series because the players are not beginners but much stronger. The previous PogChamps 6 series was marred in controversy when Dr Lupo was caught blatantly cheating after he blundered his queen in the opening (https://www.pcgamer.com/games/drlupo-admits-to-cheating-in-usd100-000-online-chess-tournament-faces-brutal-backlash-from-reddit-dude-went-from-whats-a-horsey-to-i-can-see-15-moves-ahead-in-2-minutes/). Though, I think the consensus position is that Dr Lupo cheated in other games previous to the queen blunder but it was less obvious.
Samay Raina (1728 elo) playing from India early in the morning was a surprise winner in the Super PogChamps series after clean sheeting the group round 10-0 and then overcoming two stronger players Andrea Botez (1997 elo) and Sardoche (2039 elo). Apparently, Samay might be considered under-rated and had been previously rated ~1900 in rapid on chess.com about 6 months ago. Also, its important to note that the coaching and preparation can make a big difference in a players performance when they are not at the elite level because with the correct prep you can be playing the first 10 moves or so as good as an elite player would. Samay Raina is also heavily involved in promoting chess on Youtube to Indians so you would expect he would face strong incentives to not cheat because the reputational cost would be significant if he was caught. He has also donated the prize money from his PogChamps win to a chess charity so he has not directly benefited financially from his PogChamps win.
For prize events chess.com has a special anti-cheating software called 'Proctor' (https://www.chess.com/proctor) which it can require competitors to install and run on their machines. However, it looks like for PogChamps there was no requirement for Proctor to be used according to the Proctor web page (this might be incorrect but its to the best of my knowledge). I'm not sure what steps were taken during the Super PogChamps series to stop cheating. It's possible that chess.com still required a live feed of multiple camera angles and a screen share and this would have made it almost impossible for someone to cheat.
The opening play in this first game from white between Samay and Botez showed some unprincipled decisions from Samay causing his position to blow up (https://youtube.com/watch?v=uzUyXmK_u-w&t=14885). For example on move 3 he exchanged the bishop for the knight and according to the engine this gives black a small advantage in the opening. Generally exchanging knights for bishops is not considered good. However, sometimes the position will demand that such a thing should occur. It's not completely terrible here because while white concedes the bishop pair black has doubled pawns but i think the idea is very dubious unless it was specific preparation. I think it was unlikely it was prep because a few moves later white plays d4 destabilising the knight on c3 which could be now be pinned to the king by black with Bb4. Also, there is now no dark squared bishop for white to help break the pin because it was previously exchanged for the knight! Botez ended up wining this by converting the position to a rook end game up two pawns and then was able to eventually promote a pawn and win the game.
In the second game Samay is again caught out in the opening and quickly loses a pawn (https://youtube.com/watch?v=uzUyXmK_u-w&t=16650) but this is more due to Andrea playing a sharp line and Samay making a tactical error rather than a positional mistake like in the first game. The position becomes complicated but then Andrea makes a huge blunder that Samay was able able to take advantage of and after all the trades be up a rook. In Samay's defence c5 which is the move that made the blunder from Botez possible is considered a bad from the engine and a player at his level should be able to find the tactic that wins material after Botez made the blunder.
The second game feels a lot like the games from this person (https://www.chess.com/analysis/game/live/144499596120/review) who was banned for fair play violations. This person that was banned for fair play often makes weird moves in the opening that normal strong human players would not do, then play a bunch of normal moves (not top engine lines or anything very strong) and their opponent eventually blunders and they are able to capitalise and win. If you look at this cheater's games I feel they are very weird but its not obvious the player is cheating. Maybe some of the moves have 'inconsistent' strength but I think that is expected at the rating level this player was playing at. But maybe the cheating is more obvious to someone who is a stronger chess player. And when you play someone like this, unless they make a very non-human move your take-away from the game is I was better then I blundered and I lost it doesn't even come into the picture that your opponent may have cheated.
When it comes to Samay's play in the Andrea game he might be weaker in the opening but because he has stronger tactical vision in the middle game he is better able to create opportunities and capitalise on his opponents mistakes. For example in the third game Andrea missed multiple opportunities that would have likely won her the game if she took advantage of them. My guess is its probably impossible to know if Samay cheated in PogChamps if he did because it was not in a blatant way and there is probably not enough games to identify subtle cheating.
More options
Context Copy link