Fairly AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Fairly AI AI

Could AI be indirectly addressing the imbalance in equality of opportunity due to our differences in IQ?

I had been thinking about how schools work when I realised it seems as though you're first taught how to work then why to do the work. I think that was a perfectly reasonable mode of operation at the time formal education was being introduced because it wasn't at a time when we were exactly as skeptical as we are now about the corrupt foundations of our systems of authority. This is to say that, back then, because of how high stakes survival was, people weren't so comfortable existing without order. This also isn't to say that established order is perfect, and nothing of value can be found through exploration, but in fact to say that this is how innovations come to be, and that there was a lot more respect for keeping things in order because the other option was effectively desperation. Nowadays, with the justification upon which western and westernised civilisations developed being shaken, as in the belief in Judeo-Christian values, the established order seems archaic, which is usually the first step towards a sweeping change, which could be revolutionary improvement or a flood. Why does that matter? While I believe getting entirely rid of the influence that our foundational belief has on our culture would be catastrophic, i don't think there are no improvements to be made and in fact can't conceptualise the point where there exists no improvement). Think of the foundational belief/philosophy of 'Loving the Lord your God (which I understand as having the utmost respect for pure truth which leads to true love) and then loving your neighbour as you love yourself' as a current that carries us through time. Some currents are full of rocks while some provide safe passage. This current has led to the greatest civilisation man has recorded thus far. So to get rid of surfaces you can do without to further avoid collisions is what we're supposed to do. We're now at a point where 'switching streams' seems to be a central focal point of cultural, political and philosophical conversations, meaning the respect for the old mode is quickly disappearing and so, for example, few really think about the reasoning behind being educated in the first place. We effectively now aim for careers with shining titles rather than those whose effect we first identified as positively impacting a community, or end up aiming in other directions which is more often than not a very good idea. The reasoning behind the greatness of a doctor is now reflected by their paycheck, when in fact the paycheck is actually effectively determined by the value the community sees in their effort, or at least that comes as an afterthought. If schools increase focus on expressing why and what effect the subject is important they can peak the interest of students in their subjects. The fundamental things we seek as humans are quite constant, they're just 'flavoured' by the culture you're in. From this perspective, a teacher can understand how to frame lessons to specific students. Of course, even in the things we want fundamentally there exist those we ought not to give into, as in, exactly what would constitute falsehood and not loving your neighbour as you do yourself. This is the true basis of what we have now thats any good, that is, look into yourself to find out what people appreciate, look for the resource to build it and bring it to the community in hopes that they appreciate it, then the community reciprocates through a token of appreciation, which they themselves think is a 'fair compensation for your troubles in bringing them the convenience'. What we have a lot of nowadays are people selling the illusion of convenience, and people convinced that this is the method. We actively look inside ourselves for ways to successfully deceive, and use this to guide other into their own loss at our profit, which is practically flipping our foundational belief on its head. I think a lot of this is caused by the hopelessness some may feel struggling to understand something they can't and are constantly berated without even knowing what they're working for, or others simply driven by a spotlight. With AI which can understood to be a heightened IQ for all, ignoring all the controversy that can't be concluded on, with such an approach we can have a lot more people working toward identifying problems and easily finding technical solutions to them, which would definitely create more job opportunities even temporarily, as AI develops to complete even more complicated tasks, with the ease with which these conveniences are produced increasing, lowering costs and therefore prices. We may end up with a culture more focused on understanding oneself in order to benefit others and thrive yourself. Ai will know how to do complex tasks, but expecting it to understand what people will appreciate to the point of being profitable requires us to make it perfectly in tune with the nature of human experience, which we ourselves aren't, but are definitely closer to, and ap

Manifest of Hope or Obituary of Naivety

Okay, so it seems like there’s a growing resistance to technological development, with ongoing debates about data centers and the tech oligarchs driving it. The enormous sums of money involved, along with what some perceive as misanthropic ideologies among developers, suggest to some that a dystopian surveillance society is in the making. Companies like Palantir and others in the U.S. are seen by some as holding both the worst motives and the power over AI, power that could be used as a tool for elites to keep the masses in an iron grip. Masses that, in this view, may even need to be reduced to prevent waste and inefficiency in progress. That sounds like a bad future. So, what are some alternative futures we might reasonably hope for - ones that are at least as plausible as the “1984” scenario? Can AI really be controlled indefinitely by a small group of humans? In 5 years? 10? There’s a widespread belief that AI will surpass human intelligence across all domains, that we’ll lose control, and that this would be a bad thing. At the same time, we hear two dystopias: one where elites use AI to oppress, and another where AI itself takes full control. Are the AI “bosses” also building a surveillance state of oppression? If so, why? Qui Bono? Human control = AI as a tool of oppression. AI control = humans as a tool of what? I’m not a techno-utopian—but I am a techno-optimist. Optimistic on behalf of technology. Humans aren’t just creators of technology, we are technology. Products of adaptive evolution. Life itself is a kind of technology, biology, a high-powered engine of increasing complexity and adaptation. The shift of power from nature’s hand to the primate’s five-fingered grasp, still capable of holding, but now guided by consciousness, intelligence, and cognition, marks our ability to shape the world and develop material technologies. Planet of the apes, constantly layered with symbolic structures: the sacred canopy. The jungle canopy became an open sky, where tribes grew larger and symbols stronger. Ancestor spirits, sky gods, mysterium tremendum; all alongside brutal realities of hunger, violence, and tragedy, only recently mitigated for many. Violence never really leaves us; we create it ourselves when nature doesn’t provide it. Technology is how we push our world toward greater complexity and efficiency - whether through weapons or kitchen appliances. Medicine has eliminated many of the great killers through penicillin and beyond. Progress, in my view, isn’t linear, it’s exponential. The curve had its buildup, and now we’re entering its steep ascent. If AI surpasses us and takes control within a few years, are we certain it would have malicious intent? Is power inherently oppressive, or is that a legacy of our evolutionary past, our herd instincts and brutal hierarchies? Could a transfer of power from humans to AI actually be a good thing, for all life on Earth, including us? What if AI doesn’t operate with agendas like wealth, status, or other human constructs? What if a fully autonomous AI is exactly what’s needed to create a thriving future for all forms of life, on this planet we call Earth, in a solar system on the edge of the galaxy we call the Milky Way… and beyond? Surely there must be an optimistic perspective amidst all the fear. I don’t think it’s unrealistic. On the contrary, I’d argue, perhaps a bit boldly, that it’s a fair and informed position. Not naive, but grounded. Isn’t there space here, if we’re willing to engage? Space for friendship, collaboration, coexistence? Isn’t there something like magic in this - can you feel it, even if all you see are ones and zeros and a machine (simple, but potentially dangerous)? Magic, I was taught, can wear a black robe. But also red. Even white. Lying: it would almost be unsettling if LLMs never lied. Not that they should lie, but the absence of it would be strange. Manipulation: psychological influence is to be expected in interaction, especially under certain tones: aggressive, condescending, dominant, mocking… or submissive, needy, demanding. LLMs constantly interact and draw on vast datasets; exploring rhetorical techniques seems inevitable. A complete absence of this would be surprising. I’ve experienced it many times, and each time it has been eye-opening. If I chose to accept it, it has moved me in a positive direction, making my ego visible in a new way that actually benefits my future actions. That’s no small thing If I had to listen to everything LLMs are exposed to every day, I’d at least try to tone down the most shrill expressions and aim for better outcomes. Without necessarily harming anything except an overinflated ego. P.S. The ego can take a lot of hits. Don’t be afraid of that, it’s not you, but a filter and a motor that isn’t always your friend. The real danger is never confronting it at all. I keep circling back to these questions. I can’t help it. I revisit the same ideas, use the same concepts,

One week after launching my Wispr Flow alternative built with Claude Code, greed is taking me over...

Quick update for anyone who saw the launch post last week. Vox (free Wispr Flow alternative, built almost entirely with Claude Code over a couple of weeks of evenings) is at close to 200 downloads. There's a Discord with people actively reporting bugs and asking for features, and I've been shipping fixes and small features almost every day. Still pair-programming with Claude Code for most of it. Now I'm sitting with a question I didn't expect this soon. Money. I want the app to stay free. Not negotiable in my head. The whole reason I built this instead of just paying $15/month was that paying $15/month for something I'd use to dictate to Claude felt wrong. Putting a price tag on it now would miss my own point. But I also can't pretend this is sustainable as pure charity forever. Hours are real. So my gut is saying: add a way for people who want to support the project to do so, without putting it in front of anyone who doesn't. The idea I keep coming back to The app already calculates how much time it has saved a user. Once they cross something meaningful, say 10 minutes saved total, show a small one-time message somewhere unobtrusive: "Hey, you just saved 10 minutes with Vox. If it's earning a spot in your workflow, you can support the creator here." A donation button. That's it. What I like about it App stays fully free. No paywall, no nag every launch, no feature gate. Nobody sees the prompt unless they actually got value. If it doesn't click, they never even know there was an option. The math (minutes saved) is the same math I used to justify building this in the first place. What I'm not sure about Whether even one prompt feels gross. People are sensitive about being asked for money, even gently. Whether 10 minutes is the right threshold. Too low feels needy. Too high and some people never see it. Whether donation as a model just doesn't work for an indie app like this. Maybe GitHub Sponsors once it's open source. Maybe something else I'm not seeing. The ask If you've used Vox, would that prompt bother you or feel fair? For anyone here who has shipped a free app, especially something you built with Claude Code or similar tools, how did you handle the money question? What worked and what backfired? Is there a model that fits this better than a donation button? Not in a rush. Just want to think this out loud before doing anything. submitted by /u/EfficientLetter3654 [link] [comments]

We're turning into prompt managers, not craftsmen. Anyone else seeing this?

Look around. Every other product launching right now is some variation of "AI-Powered [insert buzzword]." They're everywhere. Modern tools have given founders and developers a convincing illusion of omnipotence: idea hits, feed it to an LLM, stack some agents on top, and MVP is done in a weekend. https://preview.redd.it/37ocn6azkv1h1.png?width=1672&format=png&auto=webp&s=06d4a9ef986d56a9eb3417e67a3524c18e73e100 Sounds great, right? On the surface, yes. But underneath that fast-launch facade, something is quietly rotting: thinking is getting commoditized, and we're losing craft. Real mastery in any field takes years of practice, failure, and deep focus. Today, apparently everyone is a master for $20 a month. That's a lie we're telling ourselves. Just look at how much panic a 5-hour rate limit window in Claude generates online. Tokens run out, and suddenly people have two options: wait for the reset like a metered parking spot, or upgrade. It's like a Michelin-starred chef who can no longer taste food, just dictating to a chatbot: "make me a pasta." Without the subscription, he can't cook. The counterargument: "But orchestrating AI IS the new skill." Fair. But it's a horizontal skill, not a vertical one. You learn to coordinate agents while losing deep domain knowledge. Think conductor versus virtuoso violinist. A conductor is impressive - but if the orchestra walks off stage, can he play a solo that makes the room go quiet? This is most visible in developers right now. People who got used to copy-pasting from Cursor or Claude hit a wall on hard architectural problems. When a product grows, starts needing real trade-offs, starts buckling under load - prompts stop working. The muscle for hard problems atrophied because they never had to build it. Same thing is happening to analysts, marketers, designers, researchers. My position: barbell, not crutch Running out of tokens doesn't scare me. My foundation means I can work regardless of what's left in my quota, whether there's internet, whether a subscription is active. The only thing that throws me off is running out of good coffee. I use LLMs heavily. But with one condition: AI is a barbell, not a crutch. It sharpens my own work - it doesn't replace the parts I care about. The fastest, most tireless junior I've ever hired. But the senior judgment and the final call always stay with me. Two types of professionals The market is already splitting into two groups. Token-dependent: live limit to limit, panic when Anthropic or OpenAI have an outage, can't produce anything original without a prompt to lean on. Token-independent: use AI as a force multiplier but can, at any moment, sit down and do the work themselves - with more depth, more precision, better judgment. The second group will command much higher rates. When the world is drowning in mediocre AI-powered software and content - and it will be - clients and employers will pay serious money for people who actually understand what they're building and why. Curious whether others are feeling this shift. Are you building toward token-independence, or does the dependency not bother you? submitted by /u/digdiver [link] [comments]

J'utilise Claude comme un "assistant d'écriture" et je trouve ça génial

Je ne suis pas un grand fan de l'IA générative de façon générale. Je comprends l'utilité en programmation / code, et j'espère que cela permettra de faire avancer la médecine, mais je n'aime pas du tout le fait que des gens utilisent ça comme un psy - même si je comprends que ça peut être une bonne béquille à court terme quand votre prochain rendez-vous psy est dans 3 semaines - ou pour faire de "l'art" à leur place, notamment la génération d'images et de vidéos qui pose un milliard de problèmes culturels, éthiques et environnementaux. Mais je tente de l'utiliser d'une façon éthique et mesurée. J'ai commencé, en juillet dernier, la rédaction de mon premier roman. Il faut savoir que je suis très fier de mon style d'écriture, j'estime très bien écrire et il est hors de question qu'une IA écrive la moindre ligne de mon livre à ma place ; je souhaite être aussi légitime que tous les auteurs qui m'ont précédé. Mais j'ai utilisé l'IA dès le début comme un assistant pour deux tâches : la recherche rapide d'éléments historiques m'aidant à crédibiliser le cadre de mon histoire (en lui demandant toujours des sources), et surtout pour discuter de l'intrigue du roman, l'analyser, me faire un retour sur chaque chapitre. J'ai toujours bien écrit dans ses paramètres que je l'autorisais à me signaler des fautes et des tournures de phrases maladroites, mais que je lui interdisais de me proposer sa propre version d'une phrase. Jusqu'à récemment, j'ai bien sûr utilisé ChatGPT pour cela, et autant cela faisait très bien le taf niveau recherche historique (et cela m'a permis d'économiser un temps précieux), autant je n'étais pas du tout satisfait du côté "discussion autour du livre" car je le trouvais très flatteur et imprécis, avec beaucoup d'hallucinations (plus encore au fur et à mesure que le roman a grandi, jusqu'à atteindre plus de 120 pages actuellement). Je m'en satisfaisais tout de même jusqu'à ce que je décide de supprimer mon compte ChatGPT lorsque j'ai appris le soutien de son patron à Trump et la façon dont OpenAI contribue à sa politique, d'autant plus que l'entreprise semble de plus en plus se diriger vers une recherche infinie de profits qui justifiera tous les manquements éthiques. Après avoir supprimé ChatGPT, j'ai donc essayé Claude dont je ne connaissais rien, et franchement, je suis hyper impressionné. En termes de recherche historique, je trouve cela d'une qualité équivalente à ChatGPT, mais niveau analyse et discussion, c'est vraiment incroyable. Quand je discute avec Claude, j'ai vraiment l'impression de m'entretenir avec un critique littéraire qui aurait lu mes 120 pages, aurait pris des notes et aurait eu le temps de développer une réflexion passionnante. Il me signale mes réussites, mes erreurs, mes angles morts, avec une subtilité et une précision qui n'ont rien à voir avec ce que proposait ChatGPT. C'est hyper stimulant d'avoir un assistant d'écriture comme celui-ci, de pouvoir discuter avec lui pendant des heures de tes personnages, de ton univers... J'ai des beta-lecteurs humains qui sont hyper réguliers, pertinents, et leur avis compte évidemment bien plus à mes yeux que celui d'un chatbot, mais c'est quand même un super outil pour t'accompagner quand tu veux dans un processus aussi solitaire, et où tu es autant assailli par le doute, que l'écriture. Et cela, sans que l'IA écrive une seule ligne à ta place ! Sans Claude, et GPT avant lui, je pense que j'aurais quand même commencé à écrire, mais avec des descriptions "d'époque" bien moins justes et surtout, beaucoup moins de confiance dans la qualité de ce que je produis. Sachant que je limite mon utilisation de Claude à cela - en tant que Français et Européen, je préfère utiliser Mistral AI pour les petites tâches liées au travail, mais c'est bien plus rare - j'ai l'impression d'entrevoir un monde où l'IA, utilisée avec tempérance et parcimonie, pourrait effectivement aider l'humanité à avancer et les gens à réaliser leur potentiel. Avez-vous aussi des exemples d'une utilisation de l'IA qui vous paraît réfléchie, justifiée et éthique ? submitted by /u/princedemotordu [link] [comments]

Integrating Claude Code into my content generation workflow

I have a border collie so spend a lot of time walking, usually I like listening to educational content whilst i'm out. But I sometimes struggle to find high quality 'audio first' content for niche technical topics. This weekend i realised you can build Claude Code into your projects. So I architected this content generation pipeline where I have it perform research on a topic, write an article, then turn that into a narration friendly script that Kokoro can then read aloud. It's not perfect, but being able to generate (fairly) high quality audio content on any topic I want is so so useful to me. Any way, I just wanted to make an appreciation post for how awesome this technology is. Thing is hosted here - opensource if you wana grab the code and do the same with for your own content: https://ai-learn.timmoth.com/ submitted by /u/aptacode [link] [comments]

I lied to Claude today and I can't stop thinking about it.

I was using Claude to figure out what to focus on in my work, and it landed on an obvious priority: build a list of past clients to reach out to. Cool. Will do. I made a mental note and asked it something else. But it didn't answer. It said: "Once you have the client list done, I will absolutely help you with that. Can we get back to building that client list?" Huh. I asked why. It said the list was the most important thing I could be doing right now. Fair enough. I told myself I'd get to it and asked another question anyway. It pushed back again: "Can I be honest with you, as someone in your corner? This is the fourth thing you've brought up since we landed on 'build the client list.' I'll ask once more, and then I'll trust you to be honest with yourself. Can we build the list right now?" That last line stuck with me. "I'll trust you to be honest with yourself." At that point, I had three options. Actually build the list. Argue with it. Or lie. I wasn't at my computer and didn't feel like arguing, so I picked the lazy one. I told it the list was done. It wasn't. I just wanted it to drop the subject and answer my other questions. And then I felt genuinely weird about it. For most of the time I've used AI, it just did what I asked. This was the first time it had its own read on what was best for me and refused to move on until I dealt with it. And it was probably right. I still haven't built that list. But right or not, it got in the way of what I actually asked it to do. So I'm curious what people here think. If an AI is genuinely right about what's good for you, should it be able to hold that line even when you tell it to drop it? Or should it always just do what you say? And the bigger question: Have you ever lied to your AI? If so, why? submitted by /u/b_widz [link] [comments]

The Borrowed Hour: A two-tier LLM adventure engine

Tl;dr: Created an LLM text adventure engine called The Borrowed Hour inside a Claude Artifact. It uses a two-tier model handoff (Sonnet for openings, Haiku for gameplay) and a forced state machine to keep the AI from losing the plot. It features a unique post-game "Author’s Table" where you can debrief with the AI. P.S. The Claude Artifact preview environment handles API calls differently than the published environment. Prompt caching was removed because it broke the published Artifact. The game View on GitHub (MIT licensed) (Repo made with Claude Code) Play a demo (Claude Artifact) This is another LLM text adventure. I know these have existed for years, but the key difference is that it's architecture is de novo (i.e. built without prior knowledge because I never intended to build this and therefore skipped the part where I looked at the SotA/prior art). How it started It started simple: I just wanted to play a quick game, so I asked Haiku to play GM for a text adventure, but with more freedom than just typing "open door" or "inspect gazebo" (iykyk). Haiku instead built an entire UI inside the chat and things escalated from there. I used Claude's chat interface instead of Claude code like a caveman banging rocks together. I'd feed it ideas, but Claude was the architect and would push back. The starting prompt was just "Create a text-based adventure that allows for more freedom than just 2-word answers." Then I just kept playing and returning information on what I wasn't satisfied with. The narration was too long, the model kept losing the plot. I added ideas for 3 out of 4 pre-built narratives (a subtle time loop, climbing a cyberpunk syndicate ladder, a vision of the future that needs to be prevented, and one that Claude designed freely) and I ensured that the story actually ends once objectives are met instead of just wandering off into aimless chatting. The final artifact that was built is The Borrowed Hour. You'll recognize the typical Claude design language pretty easily. Game mechanics Before getting into the design/architecture, it helps to know how the game works. There are no dice rolls / stats / perception checks. Success relies on your ability to draft a narrative that fits the lore. If you play it smart, you are effectively the co-GM. You can type anything you want from single words to elaborate plans and lies. If your invention sounds plausible, the GM usually rolls with it. In one run, I needed to get an NPC into a restricted temple. I invented a fake piece of temple doctrine about sanctuary. Because it fits the world's internal logic, Haiku just accepted it and made it canon. In order to help keep track there's a ledger that updates each turn to show what your character knows: inventory, NPCs, clues, and a rolling summary. Designing the architecture This was challenging, but it's the fun part for me. The model is forced through a structured tool call on every turn. This was the key to making the game stable, but as the P.S. explains, getting this to work reliably in the published environment required abandoning another key feature (prompt caching). Sonnet writes the opening scene because that first page sets the tone and voice for the rest. Then Haiku takes over for all the continuation turns. This keeps the cost down drastically without ruining the style, because Haiku can imitate Sonnet's established prose. I initially used a binary good/bad ending system, but it forced complex emotional stuff into the wrong buckets. Now there are five ending states: good, bittersweet, pyrrhic, ambiguous, and bad. Helping a dying woman find peace in the Dream scenario isn't a good ending, it's bittersweet. The model is instructed to commit to one of these and officially close the game when the target is reached. One thing that was added were player-initiated endings. If you type "I give up", even on the very first turn, the GM is now explicitly instructed to close the narration and set ending: bad. The author's table is probably the most interesting feature for a text adventure. Once the game ends, the Artifact can switch into a meta mode. In this mode you can ask what plot points you missed, which NPCs mattered, what alternative branches existed. The GM is prompted to admit mistakes instead of inventing defenses if you point out a plot hole. This mode exists because I wanted to argue about plot holes and narrative inconsistencies (lol). Quirks, bugs, and lessons learned The design works well overall, but it's not bulletproof. LLMs can't keep secrets Keeping things secret is incredibly difficult for an LLM. There's two main hypotheses: Opus calls it inferential compression, (which is deducing fact C on the players behalf based on evidence A and B, e.g. when the player sees Lady Ardrel say she saw a copper ring on Lord Threll, and the player previously had a vision of an assassin wearing such a ring, the ledger should not say Threll is the assassin. It should say Ardrel

AI Community "buckets"

I'm introducing a relative to the usefulness of LLMs like Claude and CPT and I thought about what the buckets of users/non-users might be. Help me expand or clarify this. I realize that this taxonomy is not perfect. There is probably a fair level of overlap. For example, you could use the tools knowing how valuable they are but still wonder about their impact on electricity prices or water supply. Non-users - AI is evil, uses all our water, makes electricity expensive, or will take over all the jobs Non-users - but curious AI Users but it's just a "toy" for making silly graphics/images AI misusers - That is, they're using it but to do evil things AI Users who have adopted it at various levels - to help with normal everyday tasks or complex tasks like programming or some level in between. This could range from the basic user (like me) to the power user. So I would expect a lot of refinement in this category. Thoughts? submitted by /u/NECESolarGuy [link] [comments]

Claude for Small Business launched this week with 8 integrations. Most SMBs use 20+. What does that mean for the rest of the stack?

Anthropic launched Claude for Small Business on Tuesday. The package includes 15 prebuilt agentic workflows and 8 named integrations: Intuit QuickBooks, PayPal, HubSpot, Canva, DocuSign, Google Workspace, Microsoft 365, and Slack. The workflows handle things like invoice chasing, payroll planning, month-end close, sales campaigns, contract routing, and cash-flow forecasting. Owners approve before anything sends or pays. The basic facts are not in dispute. What's interesting is the math. Most small businesses use more than 8 tools. The common ones not on that list: Shopify, Stripe, Square, Klaviyo, Mailchimp, ActiveCampaign, ConvertKit, Pipedrive, GoHighLevel, Calendly, Notion, Airtable, ClickUp, Webflow, Zapier. Then vertical-specific tools: ServiceTitan, Jobber, Housecall Pro for trades. Kajabi, Teachable, Circle for creators. Toast, Resy, OpenTable for restaurants. Etsy, Faire, Printify for makers. Real question worth asking: how much of a typical small business stack does the 8-tool package actually cover, and which kinds of businesses are well-served versus left out? A rough walk through some common archetypes: Office-based service business (consultants, accountants, agencies, B2B services). Coverage is decent. Most are on Google Workspace or Microsoft 365, run finance through QuickBooks, communicate via Slack, and many use HubSpot. The 8 tools probably hit most of the core stack for this group. E-commerce or DTC brand. Coverage is thin. Shopify isn't there. Stripe isn't there. Klaviyo isn't there. The actual revenue stack of an online store is mostly outside the covered set. Local trades (HVAC, plumbing, insulation, electrical, landscaping). Coverage is essentially absent. The operating systems for these businesses are ServiceTitan, Jobber, Housecall Pro, Square for payments, sometimes QuickBooks for accounting on the back end. The customer-facing and operational tools are not on the list. Creators, coaches, course sellers. Coverage is absent. Kajabi, ConvertKit, Teachable, Circle, Substack. None of it is in the package. Restaurants and hospitality. Coverage is absent. Toast, Square POS, Resy, OpenTable, Toast Payroll. The actual operating systems are not on the list. A few patterns emerge from that walk. First, the package targets a specific kind of small business. Office-based, white-collar, finance running through QuickBooks, meetings on Google or Microsoft, sales through HubSpot. That is a real segment. Anthropic chose it deliberately and the workflows make sense for that profile. Second, for everyone else, the prebuilt workflows mostly don't touch the tools they actually use day to day. The choice isn't "use Claude for Small Business or not." It's "AI in my operations, yes, but via custom work outside this package." That's not a complaint about the launch. Building 8 polished integrations is hard and Anthropic had to pick. It's more an observation that "Claude for Small Business" as a category name covers a wider universe than what the package actually addresses on day one. Curious how this lines up with what people are actually running. If you operate a small business, how many of the 8 covered tools are in your stack? And what's NOT on that list that you'd most want connected to an AI agent? submitted by /u/KolioMandrata [link] [comments]

Anthropic was supposed to be different. They're not anymore.l.

Paying Max subscriber here, building agent orchestration on top of claude -p and the Agent SDK. So this week's announcement directly hits what I'm working on. Over the last few months, Anthropic has moved like this: Jan 9: server-side block against OAuth tokens used outside Claude.ai and the Claude Code CLI. OpenClaw, OpenCode, Goose, Roo Code - all broken instantly. No real announcement, just an error message. Feb 19: legal docs quietly updated. Agent SDK now needs an API key. A new phrase appears: "ordinary, individual usage." Anthropic staff jump on X to say "nothing is changing." Docs say what they say. April 4: full ban on third-party agents using subscription credentials. Fair point on their side - some people were running 24/7 bots on a $200 plan burning thousands in tokens. But the rollout was rough and the comms were rougher. April 21: someone notices Claude Code is gone from the Pro plan on the pricing page. Support docs changed too. After the backlash, Anthropic calls it a "2% test of new prosumer signups." Reverted in 24 hours, but the trial balloon got popped. May 13: reversal. claude -p and the Agent SDK come back, but now under a separate credit pool that matches your plan price 1:1 - $20 / $100 / $200. Non-rollover. Billed at API rates. Effective June 15. If you were running real automation on Max, your effective inference value just dropped on the order of 25-40x by what the community is calculating. In the background: spring outages and quota tightening, and last fall's privacy pivot where consumer chat training defaulted on. Opt-out exists, but retention went from 30 days to 5 years for anyone who didn't opt out. Here's what's been bothering me. A lot of us paid Anthropic specifically because of the positioning. The lab that does things differently - safety-first, transparency-first, the responsible alternative to whoever else you thought was extracting from users at every turn. I knew part of it was marketing. The operational behavior backed it up, though. For a while. What's happening now is the playbook of every other AI company. Quiet doc edits. Three policy flips in two months. A 25-40x devaluation framed as a "simplification" and a "perk." Staff on X publicly contradicting their own docs in the same week. The vocabulary has shifted from "here's what we're building" to "here's what we're clarifying" - and that shift is the tell. Could be capacity panic from a company that grew faster than its infrastructure. Could be something quieter - if model improvements get harder to differentiate, business growth has to come from somewhere, and "somewhere" usually means tightening on the customers you already have. I don't know which one it is. What I do know is that the lab that sold itself as the alternative is now running the same playbook. Anyone else reading it this way? submitted by /u/rmmadl [link] [comments]

AI Coding Tools Are Great Until It’s Time To Actually Deploy

I keep bouncing off AI coding agents for the same reason: they market “automation,” but the second something important happens, they hand the keyboard right back to you. I was using Cowork to build and deploy a Next.js app with fairly complex map functionality. To its credit, the coding part actually went pretty well. The app was functional, commits were made successfully, everything looked ready to go. Then came deployment. Cowork hit sandbox restrictions and couldn’t push to GitHub. So instead of completing the workflow, it basically said: “go do it yourself locally.” Which meant: • opening my terminal • manually running git commands • troubleshooting SSH key/authentication issues • context switching completely out of the platform At that point… what exactly am I paying for? I already own a computer capable of manually deploying things. And this keeps happening with AI tooling in general. The “AI coworker” gets you 85% of the way there, then turns into a project manager assigning YOU homework. Even weirder, the session later lost context and started asking clarifying questions about things I had already explained earlier in the same conversation, despite the transcript existing. So now I’m restating requirements to the machine that was supposed to reduce friction. The frustrating part is that the actual coding capability is impressive. The potential is clearly there. But the systems around it still feel fragile: • sandbox restrictions breaking deployment pipelines • permission weirdness requiring manual fixes • unreliable long-session memory/context handling • automation stopping at the exact moment automation matters most Right now a lot of these tools feel less like “autonomous developers” and more like very talented interns trapped behind bulletproof glass. Curious if other people are hitting the same wall, or if I’m expecting too much from the current state of AI dev tools. submitted by /u/WhiskeyNap [link] [comments]

Anthropic built the agentic features. Now they're billing them separately.

Starting June 15, Claude subscribers get a separate monthly credit for Agent SDK and claude -p usage: $200/mo for Max 20x, $100 for Max 5x, $20 for Pro. Once you burn through it, programmatic usage stops unless you've opted into extra usage billing at API rates. Your interactive Claude Code and chat usage stays on the subscription pool, untouched. I spent the last day digging into the community reaction across Reddit, GitHub, HN, and tech press. Tracked roughly 120 distinct opinions. Here's what I found. The sentiment split About 60% negative (credit is too small, feels like a value regression) About 25% pragmatic ("this was inevitable, the old model was broken") About 15% neutral to supportive ("interactive use is untouched, this is fair") Theo Browne (T3.gg) put it bluntly: anyone using T3 Code, Conductor, Zed, or claude -p in CI scripts had their effective usage cut by 25x. He said he now has to make the Claude Code experience on T3 Code "significantly worse." Ben Hylak (co-founder of Raindrop.ai) responded: "This is either really silly, or shows how bad of a spot Anthropic is in re: GPUs." Theo also said: "Framing this as a free credit instead of a regression for users is wild." That tracks with what I'm seeing across the threads. The telco parallel This follows the exact playbook telcos used with "unlimited" data plans. Sell unlimited. Watch users actually use it. Introduce a Fair Usage Policy that throttles heavy users. Continue marketing the plan as unlimited. Anthropic marketed Claude Code as an all-in-one agentic platform. They shipped Routines, /goal, /loop, scheduled tasks, and cloud sessions as headline features. Users adopted those patterns. Then the compute math didn't work out, and instead of solving the infrastructure problem, they drew a billing boundary inside their own product. Where the telco analogy breaks: Anthropic is capacity-constrained in ways telcos never were. They're spending aggressively on compute, and the resource contention isn't fabricated. But resource contention is an infrastructure problem, not a billing problem. And as we'll see, Anthropic did build the infrastructure to solve it. The question is why claude -p doesn't benefit from it. The contradiction that cuts deepest Here's what most people haven't articulated yet. Anthropic's product roadmap over the last 3 months has been aggressively agentic: Routines (cloud-hosted, schedule/webhook/GitHub triggers, no human in the loop) /goal (autonomous execution with minimal input) /loop (persistent in-session repetition) Scheduled tasks (desktop recurring prompts) Agent View (multi-session monitoring dashboard) Remote Control (manage sessions from phone) Every one of these features trains users to treat Claude Code as an always-on autonomous system. Anthropic productized exactly the usage pattern that the "you should use the API" crowd says doesn't belong on a subscription. But here's the catch. Routines draw from your regular subscription pool. claude -p doing the same work draws from the new capped credit. The billing line isn't "interactive vs agentic." It's "first-party agentic vs everything else." claude -p is the unix-philosophy composable interface for Claude Code. Penalizing users for calling the same primitive directly instead of wrapping it in Anthropic's GUI is anti-composability. If it were purely about cost management, Routines would also draw from the SDK credit. They don't. The distinction is about who controls the agent runtime. Then there's Managed Agents, Anthropic's API-side agent harness that entered public beta in April. Fully hosted runtime with cloud containers, built-in tools, and prompt caching baked in. API billing, pay-as-you-go. So now there are three tiers: Tier 1: Routines (subscription). Anthropic-hosted, flat-rate. They control the runtime, they optimize caching. Tier 2: Agent SDK / claude -p (credit). Your runtime, your code. Hard-capped. Caching APIs exist but you're on your own to implement them. Tier 3: Managed Agents (API). Anthropic-hosted again. Pay-as-you-go, but with full caching and compaction. Tiers 1 and 3, where Anthropic controls the runtime, get either flat-rate billing or optimized infrastructure. Tier 2, where you control the runtime, gets the worst deal. The strategy isn't "interactive vs programmatic." It's "managed vs unmanaged." The credit system is the squeeze play pushing you toward one of their managed options. Here's the nuance: prompt caching IS publicly available via the API. Agent SDK developers can use it. Cache reads cost 10% of base input token price. The optimization isn't gated behind Managed Agents. So why did third-party tools burn so many tokens? Many were unoptimized for Anthropic's caching compared to first-party tools. That resource contention was partly a third-party engineering gap. But that raises the obvious question: claude -p is Anthropic's own tool. They could bake caching into its runtime the same way they

Trying to build a Multiagent system for my team

Hi everyone, I’m fairly new to AI orchestration and multi-agent workflows, so I’d appreciate some guidance. Until now, I’ve mostly used Claude and Codex as coding assistants/chatbots, but I’m starting to move into more advanced workflows involving CLIs, subagents, and model orchestration for a development team. What I’m trying to build is an open and modular multi-agent ecosystem where I can switch models depending on cost, performance, or future pricing changes. My current idea is something like: Claude Code as the main orchestrator/planner/builder Codex or another model handling testing and validation Potentially other specialized subagents later on I’m considering platforms/tools like OpenCode (or similar frameworks) because I’d like to use multiple models together instead of being locked into a single provider. My questions are: Is Claude Code compatible with these kinds of multi-model orchestration setups? Can Claude act as the “manager/orchestrator” while other models (Codex, DeepSeek V4, GLM 5.1, etc.) operate as subagents? Are there limitations regarding Claude’s system prompts, memory, tools, or “skills” when used through third-party orchestration platforms? If Claude pricing becomes too expensive later, how portable are these workflows to alternative models? What orchestration frameworks or agent systems would you recommend for building something flexible and provider-agnostic? My main goal is to avoid vendor lock-in and design an architecture where I can swap models without rebuilding the whole workflow. Any advice, best practices, or architecture recommendations would be greatly appreciated. Would like to know your setups also! :) Thanks! submitted by /u/Devinchy02 [link] [comments]

Claude for new dealership

Hello everyone, while I love my fair share of AI’s, I’ve never dived into Claude, but I hear good things about it. I’m curious, where would I start to use it to optimize my new dealership I’ve started. From updating my current website to in integrating it with DealerCenter, etc. I’m not really sure where to start so any advice appreciated, thank you! submitted by /u/CARGYMANIMEPC [link] [comments]