Fable 5 is back.....run these prompts before July 12th

0:00 Okay. Today is July 1st, 2026, and Fable 5 is finally coming back. >> Yeah, it's July 2nd. We're a day late. >> Okay, but here's the thing. You only have 7 days. >> Actually, 6 days. >> 6 days to use the most powerful AI the world has ever seen. Start the countdown. It's all you got. >> Now, if you don't know what Fable 5 is, >> >> it was such a good AI model that the government took it away from us 3 days after it launched.

0:23 >> They were like, "Wow, too good, too smart, too dangerous." >> That was back on June 12th. And now here we are, July 1st. >> July 2nd. >> We have 7 days, 6 days to ask the most intelligent AI in the world questions. >> What are we going to ask it? What prompts are we going to use? >> I don't know. >> That's a lot of pressure. >> That's why I texted Daniel Mearsler and said, "Dude, please help." He's the best AI prompter I know, creator of Fabric and Pi, now called Life OS. We jumped on a quick call, and we talked about the prompts that you should use when Fable 5 comes back online.

0:51 >> So, you really have to prioritize. Like I'm trying to get I'm getting ready to schedule my sleep patterns to make sure I could use my my window, my Fable window. Do you know what I mean? >> Now, what happens after 6 days? >> Well, Anthropic changes the rules. Right now you can use Fable 5 as part of your plan, up to 50% of your weekly usage. >> Once you go past that or July 7th, you're going to be using your >> What do they call it?

1:13 >> usage credits. >> Here's the prompts, and we're doing basically three things. First, let's optimize all your stuff. Make sure it all works well. >> And then let's optimize your life. >> What the heck are you doing? >> And then, you know, if we have time, let's answer life's biggest questions. Why not? >> So, you got to look at these prompts and come up with your own and just be like, "Really, if I mean, just take it down to one. If I only had one question that I can ask a super intelligent human, what would it be? Or a super intelligent an a super intelligent alien who just showed up." And it's like, I can answer anything from like physics, like wormholes, like whatever it is.

1:47 Like, you don't want to spend that on, you know, "What's the next plot of the next Marvel movie?" or something, you know? >> And the thing with Fable is that it's so smart, you want to throw your hardest problems at it. Your biggest things. Big think. Dale's first thing is to start with harness optimization. >> Yeah, so I've got some examples up on the blog, but I'm trying to capture like the the main idea. I think the main idea is you want it to not only improve like a subsystem or like an execution of a system, but you want to figure out what systems actually improve other systems. So, it's like you want you want to improve the deepest possible layer of like everything you're doing. So, for me a harness is like the number one thing to improve because it governs how you all your AI works, right? So, if you improve your AI harness, >> >> especially the kind of system that I have where you've got like all these interconnected hooks and prompts and system prompt and skills and stuff like that.

2:46 >> Now, what's a harness? Think Hermes or Open Claw. For Daniel, that's PAI or Life OS now. And then we have security. This might sound boring, but this is your one chance to have everything you do looked at by the smartest model we'd ever seen. Take advantage of it. >> See here, security upgrades. I think this is a big one, prompt injection handling. So, that's a great example of where if I had a super alien, I would be like, "Hey, we haven't solved this yet. What do you think is the best way?" Oh, one one last prompt that I recommend, especially for people who have been building and deploying with AI, is go and find every single thing that we have built. Like, we've deployed probably across Vercel and Cloudflare, like all these different places. Make a list of them and make sure we are doing security testing on them. We can't have open ports. We can't have open auth. We can't We have to test those APIs.

3:36 So, have it build like a continuous scanning system for everything you have deployed. Like attack surface management for for all your infrastructure. >> And this is probably my favorite one, your overall life work and optimization. Do a self-model audit. What are you actually doing? Are you stressed out? Do you have no direction? This is your chance to have something good very smart look at all your stuff. >> So I I think this is a huge one for someone who's not done like a T loss or something like that. Just be like >> >> you know, I'm always jumping around.

4:08 I think this might be one of the biggest prompts actually that that's like super useful for people. It especially people who have been sort of freaking out about AI and they've just done a million projects and they've tried a bunch of product ideas and some of them worked and some of them didn't >> >> and they're feeling like massive anxiety. I think one good prompt would be like, "Hey, go look at everything that we've done together."

4:31 Right? Look at all of our history. Look at all the different blog posts. What whatever I put out online. And like come up with a theme. Like what am I actually shooting at, right? And again, maybe it doesn't apply to us because we think about this all the time. But for a lot of people >> >> they're just they have AI anxiety which you've done videos on quite a bit, but it's like they're not focused. So Fable, help me focus like what I should be doing. Like do some of these things that I'm doing do they rhyme? Are they things that you think the world needs?

5:03 Or are they about to be like super easy and trivial trivial and I should just go and do something else? So I think that's a really cool prompt. >> We didn't talk a lot about this, but it's in the blog and I just wanted to talk about it here. Tell me which parts of my work and life are about to go obsolete and which are about to 10x based on where frontier AI is going. And then give me some direction. What would you would you recommend using like Opus 4.8 to prepare prompts for Fable?

5:30 >> not really. I I actually recommend writing these out yourself. yeah, this is the type of thing I mean, you could have a conversation. Okay. So yeah, let me change my answer actually. I think I think if you go into chat mode or if you have a harness or whatever, >> >> or if you're using an Entropic app, or it doesn't actually matter >> >> what interface you're using, but I would say go into chat mode, maybe on a drive or a walk, and just talk through the things that matter to you.

6:03 Right? I'm really concerned like none of my products are working. You know, I'm trying to get out of this corporate job. I'm trying to transition to working for myself. Or I've got all these relationship problems, or I'm taking care of loved ones, or I'm trying to start this business and it's like it's not working. Or my whole AI stack is ruined and like I just don't know what I'm doing. Or what whatever it is, talk through all of those, and then literally explain, "By the way, a new version of you is coming out soon. I'm going to have access.

6:32 What should I ask you if you were smarter, but I could only use you for a short amount of time given everything that I just told you?" And then this one's kind of crazy and just fun. Daniel's trying to answer life's biggest questions. He even created this website with questions that might just be too hard for humans. I'm actually getting ready to build this into my harness as well. It's like, "What are the things that are actually too smart for humans?"

6:59 And then have all my different AIs, including the new latest and greatest, grinding against these questions and seeing like how we're making progress. Another way to think of this is like an alien intelligence shows up and they're like the best architects in the like the galaxy, and you're like, "Hey, can you You can only do a few things. >> >> Can you improve our universities, our roads, our global like logistical systems, you know, our textbooks, our scientific method?" Those are the core things we want them to improve so that if they leave, we still get all these benefits for a very long time. So it's not about just giving it bigger tasks. It's about doing deeper work. Work that will pay off even when the the window is gone, right? And at that point you don't want to pay those crazy API prices. So, what's your question? What are you going to ask Fable 5? Let me know in the comments below.

7:53 Hey, you made it to the end of the video and at the end of my videos I like to pray for you, my audience. Let's do it real quick. God, I thank you for the person on the other side of the screen. I thank you that they are interested in AI and that they want to improve their lives and I pray that through this video, through what they're about to figure out with AI, that you'll speak to them and through them and give them clarity with the tools that you've given us.

8:15 Lord, give them so much clarity right now. Let them be so excited about their lives and what they're doing. Fill them with joy. I ask this in your name, Jesus. Amen. Thanks. I'll catch you guys next time.