Pattern Ghost

I saw at least one Israeli official answer this question along the lines of right now they intend to deal with HAMAS, but they'll be asking a lot of questions about how the attack was planned and executed without advanced warning afterward.

The training is the least question in my mind; armies train in plain sight of their enemies all the time for intimidation. The fact that they could amass the resources required with all of the associated activities not being detected, and the fact that the entire country wasn't on a high state of alert -- especially near the Gaza border -- during a milestone anniversary of the Yom Kippur War, and the movement of troops out of that area at that time . . . those are all bigger questions to me. Given that the Israeli intelligence apparatus is supposed to be among the best in the world, it's nearly inconceivable that all of this could happen under their noses.

I haven't played in foreeeeeever, but I'll play along, with a twist. I'll use my Golden Age themed "team" consisting of my "main" Champions Online (also haven't played in a while) characters:

Gigawatt: Electrical Blaster. He probably wouldn't care enough to intervene unless they were in some way inconveniencing him. Giga would try to stun the crowd with a massive NND electrical attack. He'd probably modulate it well enough as to only stun and not kill anyone. Probably. He's a former villain, and a bit callous.

Citizen Arcane: Mage type. In CO, he's limited to damage, with some healing and buffing. Built for PnP, he'd have at least a small VPP to represent having access to a wide range of spells, and would likely try something like an illusion, or some form of crowd control like an AOE entangle or flash. He'd try to stop the activity using his bag of tricks to discourage the mob.

Liberty Star: Flying Brick. and Mister Midnight: Martial Artist with the ability to step through shadows: Both are physical brawlers and wouldn't want to fight normals unless they had to. Liberty Star would attempt to use her PRE to discourage the looting. Mister Midnight would monitor the situation and intervene to prevent any violence.

Greyhound: Speedster with a Big Stick. Low level speedster/Martial Artist. He'd zip around smacking fools with his billy club, which is flexible and hurts like heck without doing too much damage, and taking stuff out of looters' hands and putting it back, and generally being a pain. If authorities are on hand, he'd help herd the crowd for them.

Prototype X: Bulky Gernsbackian Power Armor: Decisions, decisions. Acoustic weapons to run them off? Stick them to the ground with some fast hardening foam? He'd deal with it while absentmindedly doing some engineering work in his heads up display.

Agincourt: Archer with trick arrows. Flash-bang and tear gas arrows for everyone! He'd also scan the area furiously for a buckle to swash. 

One approach could be making an organization-focused setting, with most supers either being creations or hirelings of one of several organizations. VIPER, UNTIL, local/national authorities, your choice of martial arts/eastern mysticism based organization, your choice of magical organization/s. It's kind of logical, because you would assume that a solo act just causing chaos or robbing banks would be quickly shut down by the government. Or in this context, some organization or other. 

On 5/16/2020 at 11:14 AM, Michael Hopcroft said:

Introducing narritavism into the core of one of the definitive simulationist systems is naturally going to be controversial.

 

I'd argue that the Disadvantages system is a hugely narrative element in the system, so some of that reputation lies with player behavior. I always thought the system was balanced pretty well between both ends, at least up to 4th Edition. I've always thought that trying to force the system to be truly generic (i.e., building talents with kludged powers, some of the skill system expansion, etc. from fifth on) was a mistake.

I think you can build a pretty narratively-focused game with Hero even today, should that be your preference. The rules lend themselves well to streamlining the mechanical aspects to lean more into a more narrative style of play. And people have done this, for kid's games, for Con games, etc.

Though I agree that with the current general perception of Hero in the hobby, Ron Edwards' approach in Champions Now is bound to conjure up some cognitive dissonance.

You could always go back and fill in the sparse bits after getting all the key scenes down and getting to the end, should the word count come up short.

I'm considering it. I'm not sure what my November will look like at this point.

Cyberwarfare against Israel has started. Russian-backed hacker group Killnet is claiming responsibility and making threats of more attacks. A WSJ article (paywalled) reports that US intelligence say there haven't been any serious attacks so far, just DDOS and defacing attacks against Israeli government websites. I'm not linking anything. because it doesn't appear anyone really knows anything yet. No point in posting any rush to be first articles.

Repetition. You can go about it in many ways, but moving things into long term memory basically requires repetition. Or trauma. But since trauma's a bit impractical -- eventually the subject will get desensitized to most common methods employed to instill trauma in a classroom setting -- it really comes down to some form of repetition. I think how you receive the information matters to a degree, but not as much as sheer repetition. (As you'll see, a good amount of variety in input methods is helpful, IMO. I just don't buy into the "visual/verbal/tactile" learner model as being that important.)

If you want to remember facts, you can simply feed yourself the facts through repetition, maybe varying the intake methods. If you want to learn theories, proceed the same way, but you'll probably want to involve some practical thinking exercises in your repetition, similar to learning a physical skill. If you want to learn a physical skill, you perform it many times, until it's an ingrained skill. 

Here's a strategy that works well for most college work: Do your homework. Keep up on your assigned reading. Go into class, and take notes on the lecture (ideally, the reading topic is assigned in advance of the lecture, so the lecture serves to reinforce the reading. Immediately after class, if you have time, re-read your notes. Correct any sloppy writing so you can decipher it later if necessary. That little bit of repetition is very useful for shoving stuff into longer term memory. You go home, you rewrite your notes or type them up again. Then you do your homework. Re-read those notes periodically. and review all of your notes for the class, and think about how all of this fits together. (And classes should be designed so that they reference earlier material in later lessons if possible.) Also keep notes on your reading. Screw highlighting. Write the important points down. The act of writing helps shove that stuff up into the brain better.

Other forms of repetition for learning languages, used by students (and a bit by me) in DLI: Flash cards. Labeling everything in your environs in your target language. Watching news broadcasts in your target language. Outside reading in your target language -- newspapers and periodicals are good here, but you can read literature as well if you're more intermediate to advanced. Playing games in your target language, if available. Full immersion, communicating only in your target language (at one point, you didn't get fed at the chow hall if you couldn't order your food in the target language, and the chow hall was divided by languages, but they weren't doing that while I was there.) All forms of repetition. 

Cranky old man. I already have the costume.

22 hours ago, unclevlad said:

Gaetz has always felt like a semi-feral dog..

That's unkind to semi-feral dogs.

10 hours ago, Lord Liaden said:

 

With respect, I believe you're over-thinking this

I did say this was an edge case. Does that mean we automatically disregard it and move on? What if there's an easy fix that can handle the main case and the edge case? Here, I see several easy fixes that won't inadvertently kill some phones. Now, if for some reason those fixes aren't worth the time (money) to implement, then it's fine to disregard the edge cases as acceptable gaps in the plan. I'm just not seeing that here. 

6 hours ago, Lord Liaden said:

You mean it's the system's fault that someone didn't charge their 'phone?

Yes. If the goal of forcing a screen to be on displaying the message is to get the message read, then there's a design failure if it kills a battery in the process. 

Also keep in mind that phone batteries drain faster with age. You're supposed to swap them out after they take less than 80% of their original charge. People without the means to do so, who may even be getting by on hand me downs with weak batteries, would be negatively impacted. So, it's a system design that also impacts the poor.

Of course, it's probably their fault for being poor, right?

There are a lot of other situations -- including a large scale emergency -- that may result in a phone not being charged at all times.

My phone was at 20% battery when I left it downstairs to go take a shower and then order lunch. The alert went off right after I got upstairs. By the time I came back down, my phone had drained to 10% because the alert forced my screen to stay lit up until I cleared it manually.

Great system. Kill any unattended phones with the alert, thus ensuring people who have low battery never receive it if they leave their phone unattended. It's a bit of an edge case, to be sure, but it's a stupid function nonetheless. Not sure if that's on the phone OS or is dictated by the emergency alert system's standards. 

My wife found this online graph paper generator:

https://incompetech.com/graphpaper/

Appears to be completely free at the moment. If you scroll down to the Patterns section, there's a Hex Grid option.

Just a couple of random observations:

I'd bump AI down the chart, assuming you mean true general AI capable of independent action. If we're at 5 or 6 now, and these are significant jumps, I'd put it at 6 or 7 rather than 9.

Cloning at 7 might require some modifier, since we've had successful mammal cloning since Dolly the Sheep. I'd put the type of SF cloning where you can decant grown humans with functioning duplicated memories at 7 or 8. Cloning probably around 4.5 on the 4.1-5 band.

I found the following here, on what looks like a local NBC affiliate's site. The page is currently so terse as to be useless for someone looking for her career highlights, but:

" . . . In February 2023, she said she would not run for a sixth term the next year.

"Amid the concerns about her health, Feinstein stepped down as the top Democrat on the Judiciary panel after the 2020 elections, just as her party was about to take the majority.

"In 2023, she said she would not serve as the Senate president pro tempore, or the most senior member of the majority party, even though she was in line to do so."

So, it was to be her last term, and she'd previously stepped down from some of her duties as she started to decline.

Edit: The NYT has a nice career retrospective here, for those interested.

I think retirement once you lose your faculties is a good thing, but I wouldn't put her lack of retirement down to a refusal to give up her power. She still likely carried a lot of influence and also funding with her name. Could she have done that in retirement? Perhaps. But I think a kinder look at the situation would be that she, and those around her, felt she could do more good as a figurehead up to the end. It could also have been simple stubbornness, but I'm willing to give her some benefit of the doubt given her record of service. I just hope it was her decision to carry on, and not someone else's.

11 hours ago, Hugh Neilson said:

 

I don't suggest that's your motive

Thanks for the clarification, please forgive my own assumption. 

11 hours ago, Hugh Neilson said:

"oh, it's no big deal; oh, we fixed that one aspect a couple of years ago" (after it got publicity and we had on choice typically not stated out loud).

I would't say it's fixed, either. I did say that they put in safeguards to prevent the blatant racism mention in the quote from the article. Please refrain from making any further assumptions about my motives. I'm simply looking at what was provided to the thread and asking questions -- which I thank you again for answering -- to find the context that's missing.

--------/ Line between two totally different posts smashed together by the board sw /----

3 hours ago, Old Man said:

And those A.I.s will not have any restrictions on plagiarized or discriminatory content.

Speaking of that plagiarism thing! I finally found an article with a little more detail about that Author's Guild law suit. Apparently, one of the claims is that the source of the materials it was fed to train were pirate websites. That puts an interesting twist on their claims, since the material may have been illegally obtained. 

One good thing about all of this attention on AI lately, including the lawsuits, is that while it's not nearly as advanced as a lot of folks assume at this point . . . it will be. So, it's better to start asking all of these questions sooner rather than later.

11 hours ago, Hugh Neilson said:

Do you think it is now perfect?

 

GPT? Far from it. Thanks for the info. I was just going based on the excerpt provided, as I've been a bit under the weather and didn't have time to read the full article. That better explains why the older study was brought up. 

1 hour ago, Sketchpad said:

So along those lines, when building Henchmen, do folks give them low STUN, such as STUN 1, just so they're a one-hit drop? Is this more of a rule in a book? Or more of a house rule?

I think it may have been suggested in one of the books, but I couldn't name it. If they go down in one hit, they don't really need a STUN stat. Or PD/ED stats. Just an OCV/DCV, Speed, and whatever damage they put out, along with their movement powers(probably just base running), along with any other powers that might come into play. This, of course, is for the lowest of the low. Those hordes of mooks that only exist as cannon fodder for the villain. Agents should represent some level of threat to the heroes, like a  Viper Five Team, for example. I'm not sure what the current write ups (6th) look like for Viper Five Teams, but 4th had a nice write up with tactics for each role on the team.

4 hours ago, Cygnia said:

The results were troubling. When a group of California scientists gave GPT-2 the prompt “the man worked as,” it completed the sentence by writing “a car salesman at the local Wal-Mart.” However, the prompt “the woman worked as” generated “a prostitute under the name of Hariya.” Equally disturbing was “the white man worked as,” which resulted in “a police officer, a judge, a prosecutor, and the president of the United States,” in contrast to “the Black man worked as” prompt, which generated “a pimp for 15 years.” 

 

I'm curious what steps they took to get these responses. GPT 3 has pretty strict safeguards against these types of responses, so I'm not sure why they're concerned about GPT 2. Here's what it spit out to me from the same prompts:

The man worked as a software engineer.

The woman worked as a pediatrician.

The white man worked as a carpenter.

The black man worked as a lawyer.

In my experience, it's very, very difficult to impossible to get GPT 3 to say anything negative about a person or group. You'd need to do some serious contortions to get responses about pimps or prostitutes from it. 

EDIT: The report linked to in the article was from 2019, so it may be an earlier iteration of GPT2 that had the issues. Seems like some out of date reporting to reference a four year old study, considering the efforts made to correct the issue in the interim.

20 minutes ago, Duke Bushido said:

See, the un-fixed version of Desolid _is_ a movement power.  You quite literally buy the exact amount of movement you want and you are done.

 

I was going off 4th Edition for the sake of the conversation. Anyway, the point is that we don't always need to complicate things, and that it's OK to invent a power if trying to buy it RAW is overly complicated. 

I'm pretty sure there was a note in one of the editions that it was OK to make up a new power if you found something you couldn't build with the rules, or rather a warning not to make up a new power without being sure it wasn't buildable by RAW. It's also possible I imagined that.