Already ordered "We Are as Gods". Can't wait to read it. However, at the pace things are moving might feel like reading ancient history. Thanks for all you do.
Peter: Your Job Market Collapse section really resonated with me. You guys inspired me to do a deep dive into solutions, despite not being a politician. Here's a quick summary, and I'd love to send you the White Paper. You guys can not only express your frustration with government, but help change it.
HOW TO AVOID SOCIAL UNREST ON THE ROAD TO ABUNDANCE
AI displaces 2-3M jobs/yr starting now.
But 10 new policies can create 6M NEW jobs.
1. National Service Corps – 2M unemployable youth hired/yr; learn real skills
3. AI Career Agent for every job seeker: provides hope through retraining
4. American Health Card – Single payer saves $2T, 15% of income, no premium for unemployed.
5. AI Dividend Sharing – $15K retraining + 1-year soft landing severance from AI profits
6. Simple flat tax – 20% for all personal and business income: revenue surplus
7. Sovereign Wealth Social Security – $2T deficit drop, benefits increase, off-budget
8. Social Revival – national choir/dance contests restore social contact, provide purpose
9. AI tourism visas → +$30B revenue: creates over 300k new jobs
10. VOTAP – Voice of the American People online national policy referendums restore agency and democracy
Execution: American Party takes Congress 2026 using AI agents, influencer (such as Tony Robbins: candidacy would bring huge ground force), X, and smart ballot-access weapons.
While I agree that we can never “claw” our way back to the pre-OpenClaw era, I am convinced that autonomous agents are at least 12 months or more into the future. Today I am working with the latest Gemini model to help set up Xero with integration to my Mercury account. What a cluster-fratz! So, HITL is here for the foreseeable future and that bodes well for the humans who remain in the loop!
You’re spot on to point that out. There’s a massive gap between agent demos and agent reliability in messy, real-world workflows. I don’t believe HITL is going away anytime soon, still.
Peter, you are almost making me a believer and that my friend is saying a lot. Your outline will require that the God's you speak about in your book are not :God's of Greed"..As big as any of the challenges you outlined and the, as Alex would say, gorilla in the room, is that men have a tendency to fear and judge. Many humans want the extra rung to displace the fear of something that is judged in our reptilian brain. Like Kirks battle with Gorn and the metrons (AI) observation: "We are still half-savage, but there is hope."
The 24/7 agent concept isn't theoretical anymore. I've been running Wiz - a personal AI agent - on night shifts for about a month. It researches, drafts blog posts, deploys code, and handles email triage while I sleep.
The surprising part? The hardest problem isn't building the agent. It's building trust. My first night shift the agent confidently reported completing a deployment that hadn't actually happened. Now there are mandatory verification checks before anything gets marked done.
Still figuring out the right autonomy boundaries. But the productivity gap between "works 8 hours" and "works 24 hours" is real and widening fast.
New vibe coders are going to get a big wake up call when they wake up to a $7k bill because they let their agent swarm attempt to realize their half-formed business idea or get brought into court for something their agents did. I'm sitting this one out.
Sounds good. Did some experiments, but haven’t thought of a use case personally to have an agent running 24/7.
I do have a short-lived script that runs every morning to help me scraping what has happened in emerging tech in the past 24 hours, but that’s probably all I need now.
I lost my job last year. Tech. writer, 56F. Claude keeps reassuring me that there's still use for my human skills in this profession, but I'm scared. I embrace AI every day :). And I am excited about the future. However, the bills are piling up today... Even if some number of tech. writers will still be in demand, most likely the younger crowd will have a better chance than me.
Already ordered "We Are as Gods". Can't wait to read it. However, at the pace things are moving might feel like reading ancient history. Thanks for all you do.
This is great to hear. I hope you enjoy the book as much as I did writing it.
Peter: Your Job Market Collapse section really resonated with me. You guys inspired me to do a deep dive into solutions, despite not being a politician. Here's a quick summary, and I'd love to send you the White Paper. You guys can not only express your frustration with government, but help change it.
HOW TO AVOID SOCIAL UNREST ON THE ROAD TO ABUNDANCE
AI displaces 2-3M jobs/yr starting now.
But 10 new policies can create 6M NEW jobs.
1. National Service Corps – 2M unemployable youth hired/yr; learn real skills
2. 4-week paid vacation mandate → +2.7M gig jobs, $0 fed cost, happier workers
3. AI Career Agent for every job seeker: provides hope through retraining
4. American Health Card – Single payer saves $2T, 15% of income, no premium for unemployed.
5. AI Dividend Sharing – $15K retraining + 1-year soft landing severance from AI profits
6. Simple flat tax – 20% for all personal and business income: revenue surplus
7. Sovereign Wealth Social Security – $2T deficit drop, benefits increase, off-budget
8. Social Revival – national choir/dance contests restore social contact, provide purpose
9. AI tourism visas → +$30B revenue: creates over 300k new jobs
10. VOTAP – Voice of the American People online national policy referendums restore agency and democracy
Execution: American Party takes Congress 2026 using AI agents, influencer (such as Tony Robbins: candidacy would bring huge ground force), X, and smart ballot-access weapons.
Contact now → Derek Hewitt | Derek.hewitt@gmail.com | DM @derekhewitt for details.
Yep, and the scary part is how normal it starts, a small shortcut, then the habit, and suddenly you can’t remember what you actually believe anymore.
While I agree that we can never “claw” our way back to the pre-OpenClaw era, I am convinced that autonomous agents are at least 12 months or more into the future. Today I am working with the latest Gemini model to help set up Xero with integration to my Mercury account. What a cluster-fratz! So, HITL is here for the foreseeable future and that bodes well for the humans who remain in the loop!
You’re spot on to point that out. There’s a massive gap between agent demos and agent reliability in messy, real-world workflows. I don’t believe HITL is going away anytime soon, still.
Love the optimism, but if we don’t talk about who pays during the transition, the future arrives looking great on paper and brutal in real lives.
Peter, you are almost making me a believer and that my friend is saying a lot. Your outline will require that the God's you speak about in your book are not :God's of Greed"..As big as any of the challenges you outlined and the, as Alex would say, gorilla in the room, is that men have a tendency to fear and judge. Many humans want the extra rung to displace the fear of something that is judged in our reptilian brain. Like Kirks battle with Gorn and the metrons (AI) observation: "We are still half-savage, but there is hope."
you are fooling yourself.
Our ai will keep China from prevailing over USA as long as we are humble enough to use it wisely
I still feel that most things about crazy 🦞 capabilities are scammy tbh
Thanks for a very inspiring post Peter. Book ordered.
The 24/7 agent concept isn't theoretical anymore. I've been running Wiz - a personal AI agent - on night shifts for about a month. It researches, drafts blog posts, deploys code, and handles email triage while I sleep.
The surprising part? The hardest problem isn't building the agent. It's building trust. My first night shift the agent confidently reported completing a deployment that hadn't actually happened. Now there are mandatory verification checks before anything gets marked done.
Still figuring out the right autonomy boundaries. But the productivity gap between "works 8 hours" and "works 24 hours" is real and widening fast.
Wrote about the full setup: https://thoughts.jock.pl/p/building-ai-agent-night-shifts-ep1
New vibe coders are going to get a big wake up call when they wake up to a $7k bill because they let their agent swarm attempt to realize their half-formed business idea or get brought into court for something their agents did. I'm sitting this one out.
Hello Peter, great post as always. This piece may be a nice, novel topic of conversation on your next pod: https://substack.com/home/post/p-188355262
Sounds good. Did some experiments, but haven’t thought of a use case personally to have an agent running 24/7.
I do have a short-lived script that runs every morning to help me scraping what has happened in emerging tech in the past 24 hours, but that’s probably all I need now.
I lost my job last year. Tech. writer, 56F. Claude keeps reassuring me that there's still use for my human skills in this profession, but I'm scared. I embrace AI every day :). And I am excited about the future. However, the bills are piling up today... Even if some number of tech. writers will still be in demand, most likely the younger crowd will have a better chance than me.
Finally I can agree with you on something, the claw moment.
As always Exciting and #scarysmart read Peter.
Thank you Matthew!