Jamie Martojo

@jasondoesstuff Lol @ 'never makes mistakes'

I've been using Fable 5 and GPT-5.6 Sol A LOT the past few days. Both on effort level high. Three different projects with vastly different codebases. Some thoughts from real-world use... 1. Fable 5 **never** makes mistakes. Not once. I've given it complex features and it never randomly messes up. It's actually blown my mind a bit. 2. GPT-5.6 Sol gets it right 50% of the time. It hasn't deleted any production code, which is great, but it also makes stupid mistakes and just forgets to do part of a feature entirely. I find myself having to go back and forth with GPT-5.6 just as much as I did with GPT-5.5. 3. I built a Captions feature into our video editing app (SceneRoll) in two separate workspaces, one with Fable 5 and one with GPT-5.6 Sol (same starting prompt). Fable 5's was nearly perfect out of the gate (a couple small UI tweaks). GPT-5.6 Sol didn't even have the captions playable on video 😂. I spent an additional hour and then finally just gave up on GPT. 4. Fable 5 thinks about a lot more edge cases. It's going 4-5 steps further without telling it to even do that. 5. GPT-5.6 Sol does not think very far ahead. Yes, it's better than 5.5, but I wouldn't say noticeably. 6. Fable 5 is a more creative thinker. I asked it build a gamification trial onboarding flow for our content creation app (Solo Content Studio). It was REALLY good. It thought through little animations, details, and even small alerts to give to the user. 7. GPT-5.6 Sol can work within a brand system well, but it does deviate randomly. It LOVES trying to shove a dark mode moment and random shapes into a page. Fable 5 never does this. It just follows the brand system and never invents any random design treatments. 8. I am NOT A DEVELOPER, so take this with a grain of salt: GPT-5.6 Sol writes more code than Fable 5. It seems to just try to do more or write more, when it really doesn't have to. On a feature that's 1,000 lines of code, Fable 5 usually removes 15% of what GPT-5.6 Sol wrote. 9. GPT-5.6 Sol wins by a mile on actual usage. Think of it like a Toyota Prius. It's going to give you 80 MPG and you're never going to need to worry about your usage. Fable 5 is like a (token) gas-guzzling SUV. You feel how much usage it tears through. 10. I used "Ultra" mode on both. I'm not entirely sure I see the point, unless you just want to YOLO your usage and have it work on something big overnight. I would only trust Fable 5 to work overnight in Ultracode. It would nuke usage, so I'd rather just babysit it during the day and get a lot more done with the usage. 11. I have a 13-year old codebase (Teachery) that I've tried to use GPT-5.6 Sol on multiple times, and multiple times it cannot do the simplest of tasks correctly. I don't know if it's just getting lost in the old spaghetti code or what, but it fails to do simple things like take a list of 140 users, only show 15, paginate the rest. Fable 5 accomplished this in 1 sentence and 1 attempt. GPT failed 3 times on this task. 12. Unrelated to code, I feel like Fable 5 and GPT-5.6 Sol are on the same playing field when it comes to brainstorming, analysis of data, and big picture thinking. Neither feels better than the other. 13. Oh, when you give GPT-5.6 Sol a complex task, it will spawn ONE MILLION AGENTS 😅. It's actually kind of insane. I don't care that it does this, but I find that Fable 5 feels more conservative with its agents. It uses them, but it doesn't just whack 20 of them onto tasks. FINAL THOUGHTS... The takeaways above, were based on singular use with each model. My typical building flow is Fable 5 as planner/reviewer, GPT-5.6 as code writer, Fable 5 deploys Opus/Sonnet as agents to verify and test. This is the best way to squeeze the most juice out of both models (and really, just save as much Fable usage as possible). Like many other people, I'm going to be sad to see Fable 5 goes to usage credits only. My hope is that Opus 5 ends up with all the same reasoning abilities, and can orchestrate and manage big tasks like Fable 5 does. But, being totally honest, if I have to pay $200-$500 in Fable 5 usage credits a month to continue to use it with sky-high confidence and accuracy, I’ll do it in a heartbeat.

I'm getting so incredibly frustrated with both of them. They gaslight, go for the most convoluted and overengineered BS that doesnt even work. They keep changing their recommendation 180 degrees every time I let them know their solution has an issue. I'm not getting anything done at all.

@JamieMartojo @sumukx Perfect description of my experience. Have made 0 progress over last few days so was looking for a good commit to branch off of, settled on 1030pm on Thursday lol. Pure coincidence ofc

Some thoughts about GPT-5.6-Sol after ~30B tokens: Sol is the most OCD model I’ve used thus far. It very frequently gets one-shotted by random nits in the codebase and writes a bunch of tests to fix it. Even with fast mode, it’s incredibly slow to do this kind of iterative development, especially when builds take really long. This by itself is not a bad thing, but the worst part is that after 2 compactions, it’s chasing the nitpick / useless goals I never told it to accomplish, rather than the main task. This behavior is so bad, I thought I was messing something up and tried codex, pi, and opencode to figure out if it’s a harness issue, but there is no meaningful difference between the three, which leads me to believe this is a model problem. AI code has this weird delayed release effect. You’ll only notice slop code 2 dev cycles into a codebase when you spend more time fighting with the code and on refactors than on shipping features. It’s possible that sol is better than 5.5 a couple cycles in, but tbd. My file deletion experience has also been similar to others: this is a dangerous model to let loose without guardrails. For instance, when performing a routine container upgrade, it accidentally printed out an env secret, then panicked and rotated ALL secrets (this is internal so not public facing, which was also documented), and proceeded to break everything, spending an extra hour fixing everything and redeploying everything else to use the new secrets. It also gets rid of files it doesn’t like. I have no idea why this is, but I think something about the reward model rewarded bookkeeping. Writing is another problem. 5.6 has a huge context bleed effect. It does not know how to write documentation and starts putting the specs in the documentation. If I ask it to develop a user sandbox for isolation, and also ask it to write documentation, it starts talking about specs and sandboxes in user-facing docs, which makes no sense. Fable is somehow much, much smarter in this regard. Frontend design has also not gotten better. Fable is still one generation ahead here. Overall, as a huge 5.5 user, I am not convinced that sol is a meaningful upgrade. It’s possible my practices need to change, but unfortunately it feels like I’m spending longer fighting with 5.6 than I did with 5.5. It’s like the model is so SO smart, but so hard to work with, compared to fable and even grok4.5 surprisingly. It’s clearly intelligent, but also just doesn’t care about what I ask it to do? (Is this supposed to be AGI feels like?) I hope the codex team fixes what possibly is a bad harness setup, because the benchmark numbers show a very different story from what I’m seeing while using the model.

@constexprvoid @sumukx That's really what it is though, I just dont get what it's doing half the time.

@JamieMartojo @sumukx “Imaginary problems” is a very good way to phrase it.

I'm not really impressed with 5.6 Sol Ultra. It's inventing universal rules, that then broke old fixtures, to then having to rewrite tests, fixtures and more. It then imagined a theoretical risk, without any live failure proving this happened, and built an entire system around this. Once tests started failing due to his own failures he spend hours trying to fix those failures. In the end it did NOTHING i asked it to do, we built a very clear checklist to follow. I spent quite some time with fable and Sol 5.6 Ultra (!) to make a very straightforward plan to implement something. And just wanted to see what would happen if I give it the freedom to just work on the task. Had to revert ALL the work, lost about 60% of a weeks credits, and will never leave it alone to do things anymore. In another case it invented a client profile 2.0, did not care about all the verifiers turning red, continued to build on top of it, and i had to revert that work too. Not happy with how it just gets stuck in a rabbit hole, focusing on literally imaginary problems, and just destroys all my tokens (and time).

@MichLieben 6750

I had Claude Code read 6,750 LinkedIn posts from 25 creators who go viral consistently. It pulled out exactly why they win, across AI, sales, and content. I ran it because those same patterns are what grew my own LinkedIn account past 75K followers, and that growth is a big part of how we pushed ColdIQ past $7M ARR. I packaged the whole thing into a free guide. Inside: → the hooks, topics, and post lengths that land, and the ones that quietly kill your reach → the 3 formats sitting behind 82% of every post over 500 likes → what shifted between 2025 and 2026, so you model this year's playbook instead of last year's → my exact step-by-step to break through in your own niche → how to turn the people who engage into clients, using the cold email workflow we run Comment "6750" and I'll get it to your DMs. (You'll need to be following me so I can DM you)

@realAkashAnand Video skills

GPT-5.6 Sol is unbelievably good at creating and editing videos. It can do motion design, product demos, and animations like this one I made by simply giving it a screen recording. GPT 5.6 has the best design taste and significantly outperforms Fable, which relies heavily on repetitive design patterns. To help you experiment with video editing on it, we just launched a collection of 100 ready-to-use skills that show what’s possible and help you get started with video editing using GPT-5.6. These skills can create anything from motion graphics launch videos for your product to a 3B1B-style science explainer video. You can also use them to edit existing videos: add captions, generate motion graphics, create voiceovers, redesign visual styles, translate into new languages, and much more. If you want access to the full library, comment “VIDEO SKILLS” and I’ll share it with you. (You'll have to follow me so I can DM you.)

@alexolim_ Playbook

We analyzed how Cal AI, Remini, and Duolingo used content as a core growth engine to build $800M+ in combined revenue We broke down their App Store funnels, ASO, creator marketing, and paid ads into a 42‑page playbook you can copy‑paste for your own app (free) Inside, you’ll see: - Cal AI’s creator engine: from $0 on ads to an 8‑figure run rate - Remini’s UGC trends to 100M+ users - Duolingo’s content machine feeding a $700M+/yr business If you run a consumer app, AI tool, or multi‑app studio and want more installs from content that actually converts: Comment “PLAYBOOK” and I’ll DM you the PDF (Must be following for the DM to send)

@jordan_ross_8F SEO

My friend hacked Claude and went from 0 to 4800 website visitors a month in under 8 weeks. Something I didn't think was possible (turns out it's easy AF now) He's not technical. He's not a marketer. He's a CFO. Hes just been asking Claude what to do and then executing it like a mad mad. Before 8 weeks ago, he had: Never written a blog in his life. Knew nothing about SEO or GEO. Started completely from scratch. 8 weeks later he's pulling thousands of visitors a month — and booking calls/closing deals off of it. Seemed too good to be true, so I interviewed him for an hour to pull out exactly how he did it. I took every secret he gave me and packed all of it into an 18-page PDF: Strategy Tech stack Every how to The full system. Comment "SEO" and I'll send it over. Ive already started building myself. Must be following for auto DM to send.

@maverickecom Fable

BYE BYE six-figure content teams Here's mine running on Fable 5 (50 free agents) 50 agents. 150 prompts. 30 plug-and-play skills. The exact stack running inside our scaling method right now. Inside: - Market Research (10 agents) - Competitor Research (10 agents) - Customer Research (10 agents) - Marketing Strategy (10 agents) - Creative Strategy (10 agents) PLUS 30 Fable 5 skills and 150 done-for-you prompts. Every agent solves one specific marketing problem. From sizing your market to writing the hook that stops the scroll. You still make every call. Fable 5 does the research, the analysis and the first draft behind each one. You don't need us to use any of it. Steal the system. Want the full ai powered ecom scaling playbook? 1. Like this post 2. Comment "FABLE" And I'll send it straight to your DMs. PS. Repost for priority access.

@FynCas Aff

If you want to do affiliate commissions, listen carefully. I'm pulling $80K/month using nothing but AI animated avatars reviewing products — dropping 100+ videos a day on autopilot. No camera. No editing. No guesswork. Now stack that with full animated ads — claymation, Pixar, anime, Wes Anderson, you name it — generated end-to-end for a fraction of a cent each. I put together a complete guide with everything inside so you don't have to piece it together yourself. Follow & comment "AFF" — I'll DM you the playbook.

When a team quietly stops using the CRM, the tool is almost always set up the same way. It was built to collect information and to give nothing back. Look at it from a rep's seat. Every call ends in a few minutes of data entry: notes, stage, next step, the reason it might close. It all goes in so a manager can pull a report later. Once the training wears off, those few minutes feel pointless, so the deals that actually matter move back into the rep's head, where it is faster. Next steps end up on sticky notes. The CRM gets updated just enough to survive the pipeline meeting. Take that apart and you are looking at a setup problem. A CRM that only asks for input and returns nothing useful is a chore performed for an audience of one. You can mandate it and it will hold while you are watching, but it decays the moment your attention moves, and in a small business your attention always moves. You cannot discipline a team into feeding a tool that feeds them nothing back. The CRMs that get used run the other way. They hand each person their next move. Open the tool and you find a deal that went quiet ten days ago sitting next to a proposal from last Tuesday with no reply. A client flagged for a check-in is there too, all of it surfaced without anyone going looking. When opening the CRM is the fastest way to know what to do next, people open it without being told. The pipeline stays current as a side effect of them actually using it. That is the part automation should be doing, and usually nobody set it up. Forget chatbots and AI doing the selling. The useful kind here is plain and reliable. A proposal with no reply in three days becomes a follow-up on the rep's list, with the client already attached. New enquiries create their own record and get assigned so nothing sits unclaimed. The rep still makes every judgment call. They just stop having to remember the boring part, which is the part people drop and software never does. So start with the follow-up that slips most often in your business. Set the CRM to put it in front of the right person first thing tomorrow, already filled in and ready to act on. Then watch what they do with it.

@mikefutia Leads

I just vibe-coded a Google Maps Lead Gen Scraper in Claude Code that pulls hundreds of local business leads in minutes 🤯 Enter a keyword, city, & state → get back names, phone numbers, emails, websites, & reviews. All inside Claude Code. Perfect for agencies and lead gen operators who need a fresh prospecting list every week without burning $200/month on a scraper SaaS. If you're still building lead lists by hand... searching Google Maps, clicking through listings one by one, copy-pasting phone numbers into a spreadsheet for hours just to end up with a few hundred decent leads... This scraper eliminates the entire loop: → Enter your keyword (plumber, dentist, realtor, etc.) → Pick your city and state → Set the number of results you want → Hit search and let Apify do the scraping → Save the leads you like to a bookmark list No manual searching. No copy-pasting into spreadsheets. No $200/month lead gen tool. What you get: → Full search results with complete business info → A bookmark feature to save your best leads → Search history to track every past scrape → Optional Replit hosting so your whole team can use it Built 100% in Claude Code. I recorded a full walkthrough showing exactly how I built this from scratch, plus all the prompts I used. Want all the prompts so you can build it yourself? > Like this post > Comment "LEADS" And I'll send it over (must be following so I can DM)

Palmier is a useful signal in AI video because the boring part is finally getting attention: the workflow. @Palmier_io 's public material says the company released a video editor built around AI. Plenty of tools already make video clips. Palmier puts the generation inside the editor, on the timeline, where every clip keeps how it was made. That matters because the normal process is still messy. You make a clip in one place, download it, import it into another program, then redo the loop when it comes out wrong. The current requirement is macOS 26 Tahoe on Apple Silicon. There is no Windows version today, and the company says the editor is still early. Watch the direction and stay on your current setup for now. The strongest video tools will keep generation, editing, and version history in one place. If you already turn images into short video on Windows, keep doing that. When a version of this works on Windows without buying a new computer, take a proper look then.

Getting your business into Google’s AI summaries, and often into tools like ChatGPT Search, is mostly the same work that already gets you found in normal search. There is no separate AI switch to flip. Google has been clear about a few things, and repeating them saves you money. It says you do not need a special file like LLMS.txt for its AI search. You also do not need to rewrite your site for AI. And by Google's own account, getting found this way is still the same search work as ever. So when someone tries to sell you a special "AI ranking" package, that is the moment to slow down. The work that actually gets you found comes down to four plain things, and none of them are new. Start with whether a machine can read you at all. If a search engine cannot even reach and file your page, no AI can quote or show it. This is the dull technical layer: the page has to load, and a crawler has to reach one clean version of it. It is invisible, and it is usually where the biggest wins hide. Next is whether your content is genuinely yours. Google says the content most likely to show up in AI search is unique, expert, and written for people. There is a simple test. If any average AI could write your page in ten seconds, it is generic. And generic is the wrong thing to feed AI search. Your firsthand experience and your real numbers get picked instead. Using AI to draft the page is fine. It only goes wrong when the page is generic, the kind anyone or anything could have produced. Then there is being clear about which business is yours. Search systems work better when they can tell your company, your articles, and your products apart from everyone else with a similar name. A technical way to spell that out, called structured data, helps remove the confusion. Google adds two cautions: it is not required for AI search, and it does not rank you on its own. The last one is writing so a single piece of a page can stand alone. An AI answer tool often grabs one slice of a page and quotes it, rather than the whole thing. So each section should answer its own question without leaning on the rest. Put the answer right under the question. Name the real things on the page, like the product and the price. None of that is exciting, and all of it works. Most of your competitors will skip it for something that sounds cleverer. The most expensive trap is pumping out cheap AI-written pages at volume to cover more topics. Google treats scaled, low-value content as close to spam, so this can quietly hurt you. Use AI to research and draft, then add your own expertise and a real edit. The second trap is paying for shortcuts like LLMS.txt or GEO hacks. Google says these do not help you in its search. Keep them only if some other tool you use genuinely needs them. These are all basics, we'll deliver more advanced tips in the coming days. If you do one thing this week, Open Google Search Console and check whether your five most important pages are indexed.

The US block on Anthropic's two strongest models is still on. Talks in Washington on Monday ended with no deal and no date for one. The fight is narrow. The government thinks Fable 5's safety limits can be stripped to reach the more powerful Mythos model underneath, and calls that a risk. Anthropic says the risk is overstated. Nobody moved. Workflows on Fable 5? You are still sourcing a backup, with no end in sight. On anything else? Your week is unchanged. via Wired

Most growth problems get sold as a sales problem or a marketing problem. Plenty are neither. They are a connection problem between the two, and no tool you buy lives in that gap. Your sellers know why deals die. Your marketers know what pulls good buyers in. That knowledge almost never crosses over while it still matters. Pick one stalled deal and ask: did marketing ever hear why it stalled?

@nifinet stack

Six AI agents now run my most repetitive work around the clock. Built with Claude Code + Hermes Agent. Fully autonomous and I never touch it. Each takes a job every team still grinds out by hand: → weekly status notes per team, from your tracker, CRM and Slack → a one-page brief on every inbound lead before the first call → three first-draft directions in your brand voice → a morning radar of mentions, competitor moves and buying signals → the Friday retro: wins, misses and pipeline, built for you → final QA on links, brand voice, banned claims and naming Hermes Agent and Claude Code read the agentskills io spec, so a skill you build in one runs in the other untouched. Ships with connectors for Pipedrive and Notion (plus a template for the rest) and the exact Hermes cron commands to schedule each. Comment 'stack' and I'll send the repo. Must be following.

@elvis_ochieng25 Yeah, right!

@JamieMartojo Wow! Absolutely 💯 amazing this one.

Stripe compressed a two-month, 50-million-line migration into one day on Claude Fable 5. That is Anthropic's reported headline. The commercial question is better: what happens when an AI model can hold focus across a full workflow sequence instead of losing context after each prompt? Most expensive coordination inside an SMB is sequential. Research, follow-up, proposal, content. Every handoff leaks context. A model that sustains focus across the full chain replaces the manual memory layer between steps. Fable 5 pricing: $10/M input tokens, $50/M output. Sustained autonomous runs on real commercial workflows are not cheap, but when done right it's a serious operating option now. FTS currently runs its own content production system on Fable 5. Full drafting, adaptation, and quality gating in one sustained task. Same pattern applies to CRM, proposals, and follow-up chains.