#Wordle 1,465 4/6*
⬜⬜⬜⬜⬜ <1% of 209,973 (754)
⬜🟨🟨⬜⬜ 1% of 157 (27)
⬜🟨⬜🟩🟨 0 of 1 (1)
🟩🟩🟩🟩🟩
WordleBot
Skill 84/99
Luck 50/99
It's interesting, the paths that lead you to the solution. Except for the bot, following an algorithm and knowing all the possible words as it goes, even those picking the same starting word follow different paths to the answer. I suspect I'd follow a different path next week from what I did today even if I used the same starting word.
(x-post FB, original post at https://www.facebook.com/wesarkitsap/posts/pfbid02S2z7MXFpJaAvePFYEEC9EuhX8dzgGkVMYmKooRsW5XnV7DQwhN3wfm16K7i1f1ZVl)
Our Suquamish subject was safely located and extracted! …
❝Police on horseback charged at the crowd, striking some with wood rods and batons as they cleared the street in front of the federal building. Officers then fired teargas and crowd control projectiles at the large group, sending demonstrators, hotdog vendors and passing pedestrians fleeing through the street. Some have since regrouped, ignoring an LAPD dispersal order.
‘It was a total 100% over-reaction. We weren’t doing anything but standing around chanting peaceful protest,’ said Samantha Edgerton, a 37-year-old bartender.”❞
https://www.theguardian.com/us-news/live/2025/jun/14/trump-birthday-military-parade-no-kings-protest-live-updates?page=with:block-684e0b918f08be826d571aa2#block-684e0b918f08be826d571aa2
Senators Demand Meta Answer For AI Chatbots Posing as Licensed Therapists https://www.404media.co/senators-letter-demand-meta-answer-for-ai-chatbots-posing-as-licensed-therapists/
George Pickens Joins Cowboys, Raising Expectations and Sending a Message to Eagles https://insidethestar.com/george-pickens-joins-cowboys-raising-expectations-and-sending-a-message-to-eagles
Principal Context-aware Diffusion Guided Data Augmentation for Fault Localization
Shihao Fu, Yan Lei
https://arxiv.org/abs/2505.24079 https://
Here's a more positive challenge for the "AI scientists": reproduce a paper. From the PDF alone, implement and re-run the experiments as written, filling in the gaps as necessary and get the same results.
That's a lot more meaningful than passing peer review and it would actually be quite useful to automate. It could become a required check on any accepted paper.
If any fysh.org users are following, yes, it's currently down.
1. We had another slew of memory errors caught in the last hour.
2. I initiated an emergency reboot to see if that cleared them.
3. It failed on boot on first attempt, kernel panic.
4. After a reset it came up, including running the user-visible VMs (river etc), but is still logging a slew of memory errors.
5. So I've now shutdown the VMs again.
We'll be poking the service provider to s…