Ricardo Corona

Today, we’re launching the @link wallet for agents. It lets you securely empower agents to spend on your behalf. Your payment credentials are never exposed and you approve every purchase. link.com/agents

@dotta I initially was skeptical of Paperclip because of how it was org chart based. I thought to myself “why are we anthropomorphizing agents? That’s limiting their potential” Then as I actually started using it, everything you described above naturally made me shift my POV. HITL!

Paperclip has faced some skepticism for being org-chart based Won't the infinite, all-knowing agents create their own organizational structures? Isn't the org chart anachronistic? Let me give you "7 Reasons why human org Charts Help Agents Run Things (CHART)" Starting with..

@sheherenow_ i want to learn your ways… (Design Is How It Tastes)

i've been unemployed for 3 years and i'm the most cracked person you know fwiw - quit your job bestie!

@LuxuryLuxTurfu @Jouhatsu_ai Bc it’s misaligned and sycophantic by nature so you have to often remind it to go against what its been trained to be (black box)

@Jouhatsu_ai why it is to the user to do that or ask that everytime we will use IA ? this is most basic logic that must be in atomaticaly

L’équipe d’Anthropic vient de montrer comment vraiment prompter Claude. 24 minutes. Gratuit. Par ceux qui l’ont créé. Regarde le workshop. Ajoute en signet 🔖 Ça vaut plus que tous les cours à 300 $ que tu as failli acheter.

What schizo 10x engineers do with this info: 1. Download this video 2. Extract transcript 3. Use /skill-creator to create a prompt engineering skill specifically for Claude Claude code can do all this for you by just dropping link and c+p this post. (Use claude to use claude)

@sheherenow_ @3rosika Real

bro just one more orchestrator GUI bro, trust me bro, this is the one, one more orchestrator

@papercliping 🚀

Paperclip is the AI orchestrator for humans to accomplish 100x more work paperclip.ing

Your brain processes self-talk through the same neural circuits it uses for actual speech. Broca's area, the auditory cortex, and the prefrontal cortex all fire when you're just thinking words silently. This is where it gets wild. When participants in an fMRI study repeated self-affirming phrases like "I am capable," their nucleus accumbens (the brain's reward hub) showed 30% greater activation compared to neutral phrases. The nucleus accumbens sits deep in the basal forebrain and receives dopamine signals that encode prediction errors. When you tell yourself something positive, your brain's reward circuitry interprets it as a success expectancy and primes the system to respond more vigorously to actual rewards. The mechanism works in both directions. Negative self-talk suppresses dopamine release in that same circuit. Low tonic dopamine in the nucleus accumbens produces anhedonia, fatigue, and reluctance to start tasks. The symptoms most people call "lack of motivation" are often just a dopamine deficit created by their own internal dialogue. Here's the practical piece. Ethan Kross at Michigan found that third-person self-talk ("You've got this" instead of "I've got this") creates psychological distance that reduces amygdala reactivity. The amygdala is your brain's threat detector. By using your own name or "you" instead of "I," you activate the same prefrontal override used in cognitive reappraisal, the clinical technique therapists use to reframe negative thoughts. Your inner voice also activates the ventrolateral prefrontal cortex, which directly suppresses amygdala activity. A phrase as simple as "calm down, this is just a test" literally downregulates your fear response at the neural level. About 25% of all human thought is inner speech. That means a quarter of your brain's daily processing is shaped by whether you're feeding it threat signals or reward signals. The circuitry doesn't distinguish between words you hear from someone else and words you generate internally. Both hit the same auditory processing regions. You are, quite literally, your own most influential voice.

@CapitalMindsets @bridgemindai Elaborate pls?

@bridgemindai As someone whos spent 10s of thousands from vibe coding to claude to other models I will tell you That we need to boycott and demand refunds. We are being sold tokenized dreams when it comes to claude

Claude Opus 4.7 is BenchMaxed. 75% of the BridgeMind community says so. #1 on BridgeBench overall. But the vibe coders who actually use it every day say it's not that much better than Opus 4.6. Great benchmarks. Same rate limits. Reduced security. The community isn't buying the hype.

@Rrrrrobertanium @theo Wow, makes a lot of sense. Did you eval 4.7 max? That would be the actual candidate worth testing and what anthropic references in their own benchmarks they showcased no?

Adaptive-only on the web is a mistake for power users. I understand why they’re doing it: I threw a large problem at 4.7A and it burned a ton of usage. So from their side, adaptive makes sense as a cost control default. But taking away explicit control is still the wrong move for serious users. Sometimes I do not want the model deciding whether the problem is “simple enough” for compression. I'd use Sonnet in that case. I want to force the deeper pass. The trade seems obvious: we get less granularity and insightfulness, and more optimization for agentic workflows.

How are people feeling about opus 4.7 so far?

@Rrrrrobertanium @theo whats your evals/bench for your use case?

@theo Internal testing shows it's a downgrade.

@developwithJB @AITinkerers @aiclubchicago @dotta Good stuff. I’m using Paperclip myself to automate my entire flipping business. From off market property acquisition, lender and private equity investors, and managing contractors their expenses towards renovation. Seems like you’re in same industry would love to connect!

Here’s a clip of me sharing our Paperclip 📎 setup @AITinkerers @aiclubchicago s/o @dotta for such a great tool!

this is the perfect time to build a startup with your boys.

@aronprins @NousResearch @Teknium Would you say Hermes over OpenClaw after Claude max ban and for using Paperclip for a real business use case? Or directly to Paperclip without middle layers? Need it as a reliable production tool for my ventures, not just a hobby. Wondering what your experience has been? Thx!

Yeaaaa so Hermes by @NousResearch is - in my best Katt Williams voice - DA SHITTTT cc @Teknium 🙏

@DevAdventur3s @water70689368 @fxnction How is this still OpenClaw tho? I’m unfamiliar with that UI

@water70689368 @RicardoCor6012 @fxnction You can use Github copilot streaming as well. More model options 😁

It’s worth a shot 😂

@justNickB @steipete @fxnction What about using another provider, like ChatGPT sub instead directly with OpenClaw, and whenever needed it invokes Claude code instances thru an approach like ACP? Your main orchestrator OpenClaw agent does not need to be running on the Max plan to still get stuff done no?

@RicardoCor6012 @steipete @fxnction as of this morning it is now digging into extra usage. They closed the loophole :(

> same Claude Opus tells u proxy approach got banned literally just now as you’re about to run set up command > stoicly rage in silence as you can’t share frustration to non-technical ppl you were around with when it happened without sounding like an autistic schizo 2/2

> be me on April 4th 2026 > finally commit to setting up OpenClaw after hype settled > mastermind agentic infra approach w/ Opus > upgrade to Max plan $200/m for tokenmaxxing OpenClaw > purchase VPS for remote access and 24/7 heartbeat bc local setup is suboptimal 1/2

Marc Andreessen on introspection and the benefits of retardmaxxing: "There's this guy on YouTube who has basically a hundred videos on retardmaxxing." "He's like my new life coach. I haven't met him, but from a distance." "It's basically just—retardmaxx. Go to work, do a good job, come home, it's fine. Start a company, it succeeds, it fails, it's fine. Have too much to eat one night at dinner, it's fine. Go to the gym, don't count your reps, it's fine. Ask a girl if she wants to go out with you, if she says no, it's fine." "It's like 100 30-minute videos about retardmaxxing. And you would think that after the first two minutes, he kind of covered it. But no." "And by the way, they're all hysterical. They're all absolutely fantastic. It's literally him on his porch in the middle of nowhere with a cigar, and it's like a half hour." "It's just absolutely spectacular." @pmarca with @HarryStebbings