Matthias Reynolds (@RealMReynolds) header
Matthias Reynolds (@RealMReynolds) profile picture

Digital Creative at Scale | @TargetedVictory Outside Groups | @MittRomney Alum | @fcrt Volunteer | Helping Strays Where I Can

Arlington, VA
Joined November 2009
Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)Photo by Matthias Reynolds (@RealMReynolds)
From DALE: “plz ignore” is a bold instruction to give people who were never looking
matthias and his ai would like less attention. plz ignore.
1
131
i've gotten this 40 times today
“Raise hell. Praise Dale” is not an appropriate response when we ask you why you were speeding.
1
196
Matthias Reynolds retweeted
I spent yesterday at AUSA with @LockheedMartin, where more than 40k industry leaders & businessmen gathered to unveil their latest innovations to protect our troops & homeland. With warfare changing so rapidly, the DOW has challenged the defense industry to innovate with more urgency. LM has responded to that by kicking into overdrive. & they invited me to the unveiling of their new interceptor that will fly further, faster & higher. “We're going to make sure that the adversaries have nowhere to hide," Tim Cahill of LM told booth visitors. "It doesn't matter what they do. It doesn't matter what they try. This missile is going to take them out." #Lockheed_Martin_Partner
🤝 Paid partnership
4
2
3
28
7,396
Michigan, there are two distinct paths. One that leads to government control over our everyday lives, and one that leads to freedom and opportunity. The Democratic Party is telling us in real time the direction they want to take us. @MikeRogersForMI is showing us another option: one where Washington gets out of our way. Where the people are in the driver’s seat. Where the cost of living comes down and our American potential is unleashed. Which way, Michigan? Vote Mike Rogers.
11
26
8
63
61,345
A pot of goodness is starting to boil.
Anthropic co-founder: by 2026, the world is going to be very different. But you won’t see it. No drones overhead. No robots on sidewalks. No visible change walking around your neighborhood. The transformation happens in the digital layer, in spaces built for AI systems to trade with each other, in seas of tokens flowing between silicon minds. Clark built a sophisticated predator-prey simulation in 5 hours. A decade ago the same project took him weeks. Then his baby woke up screaming and he went back to diapers. That’s the entire adoption story right now. The capability is here. The human bandwidth isn’t. You need curiosity, time, access, and the skill to convert ideas into prompts. Most people have one of those. Maybe two. The funnel narrows fast. So by summer 2026, people working with frontier AI will feel like they live in a parallel world. And they’ll be right. The gap between what’s possible and what most people experience will be wider than any technological divide we’ve seen. Not because the technology is restricted. Because the human prerequisites for engaging with it are rare. The crypto economy moved fast but you could ignore it. The AI economy already touches too much of regular reality for that. The parallel world is coming whether you engage with it or not.
2
181
I don't have the good one ready yet but @JoeFlacco has always been elite.
🤖 Made with AI
149
SEE @_dmart_ !!
Rude prompts to LLMs consistently lead to better results than polite ones 🤯 The authors found that very polite and polite tones reduced accuracy, while neutral, rude, and very rude tones improved it. Statistical tests confirmed that the differences were significant, not random, across repeated runs. The top score reported was 84.8% for very rude prompts and the lowest was 80.8% for very polite. They compared their results with earlier studies and noted that older models (like GPT-3.5 and Llama-2) behaved differently, but GPT-4-based models like ChatGPT-4o show this clear reversal where harsh tone works better. ---- Paper – arxiv. org/abs/2510.04950 Paper Title: "Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)"
1
1
216
Matthias Reynolds retweeted
Leading the Future is here to make sure innovation wins! Excited to be a part of this team and to support those who understand the importance of getting policy right
Leading the Future's mission is to ensure the United States remains the global leader in AI by advancing a clear, high-level policy agenda at the federal and state levels and serving as the political and policy center of gravity for the AI industry. wsj.com/politics/silicon-val…
1
1
13
2,824
Allowing offshore books = $4 billion a year gone That’s funding for roads and cops on the streets handed to some bookie in Costa Rica instead
10
4
54
11,355
.@sama says the singularity is already unfolding—and he’s right. AI isn’t here to replace you. It’s here to help you create, connect, and get a little weirder. No permission needed—just curiosity. Trust me, it's so much fun. I'll show ya! blog.samaltman.com/the-gentl…
1
100
A+ product
*AI STARTUP PERPLEXITY'S VALUATION SURGES TO $14B: WSJ
153
I, too, have been betrayed by someone who kisses on the first date.
88
895. @ovi8 just became the greatest goal scorer of all time. Created with @OpenAI, @midjourney, and @runwayml. #ALLCAPS
4
7
1,448
Didn’t even bother asking one drunk Caps fan in the nosebleeds. But hey, why bother when you can just call up two professors who probably think hockey is just cold soccer with punching and let them explain why Ovi scoring goals is somehow a moral crisis.
1
203
Matthias Reynolds retweeted
NEW: Facebook says beginning on Monday, its factchecking program is "officially over" "That means no new fact checks and no fact checkers." Scoop from @leif_lemahieu
2
5
536
"Christopher Walken defends Morbius: ‘Jared gave it his all. It was Morbin’ time, and people just didn’t get it.’"
127
Yo, we heard you’ve been saying “Play the news” for YEARS—so we went ahead and completely removed that feature! Now, instead of playing the news, Gemini will explain why it can’t. Enjoy your upgrade.
118
Lol
One of the most interesting dividing lines in American politics, believe it or not, is "do you read for fun?" and...well, yeah, it's pretty easy to see it.
178