Get the FREE Learn AI guide: https://aiwithkyle.com/learnai

Build your AI skills with the AI Confidence Accelerator: https://aiwithkyle.com/confidence

Turn an idea into a small app or website. I walk through my tech stack and the concepts you will meet as you build, then show how I keep up with AI.

Part 3 of the Chat, Delegate, Build series, adapted from my full How to Learn AI course.

00:00 Start building with AI

04:58 My tech stack

08:04 Concepts to understand

11:16 How to keep up with AI

13:32 Give new information a home

16:23 Inside AI Confidence Accelerator

18:48 Get the free guide

Get the free illustrated guide and first-product worksheet: https://aiwithkyle.com/mini/sell-your-expertise

Build your confidence using AI at work: https://aiwithkyle.com/confidence

Coaches, consultants and freelancers can turn their expertise into AI products that help clients get work done. I walk through four ways to do that: skills, plugins, an MCP server and a SaaS product. Using a personal branding coach as the example, I explain what the customer gets, what you have to maintain and where your own coaching or hands-on help fits.

We start with getting your process out of your head and into a skill, including the questions you ask, examples you approve and mistakes you catch. Then I show how plugins bundle that work, how MCP connects a client's AI to your service, and what changes when you run the whole software product yourself. I also cover the custom GPT transition and why the simplest option that solves your customer's problem is often the sensible place to start. The prices I mention are examples for explaining the offer structure.

*RESOURCES*

AI Skills 101: https://www.youtube.com/watch?v=w49OfWDTTDo

Custom GPT migration guide: https://learn.chatgpt.com/docs/migrate-custom-gpts

*CHAPTERS*

0:00 Why Would Clients Pay You When They Have AI?

2:21 The Personal Branding Coach Example

3:17 Turn Information Into Action

4:36 The Custom GPT Retirement

6:08 Four Ways to Package Your Expertise

7:58 What an AI Skill Actually Contains

8:45 Get Your Process Out of Your Head

11:10 How Customers Use Your Skill

13:49 Bundle Your Work Into a Plugin

16:36 MCP: Connect Clients to Your Service

20:01 SaaS: Run the Whole Product

23:46 Why People Pay for Something They Could Build

25:05 Products, Group Coaching and One-to-One Help

27:17 Where to Start

*KEY POINTS*

• Capture your process, examples and judgment in a reusable skill.

• Bundle related skills and tool connections in a plugin.

• Use a hosted service when you need to control access and maintain the software.

• Account for onboarding, support, billing and AI costs before choosing SaaS.

• Offer personal feedback and implementation alongside the product.

*SOURCES AND TOOLS MENTIONED*

Agent Skills: https://agentskills.io/home

OpenAI plugin architecture: https://developers.openai.com/plugins/concepts/plugins

MCP servers: https://developers.openai.com/plugins/concepts/mcp-server

OpenAI plugin directory guidelines: https://developers.openai.com/plugins/app-guidelines

Claude plugins: https://claude.com/blog/build-plugins-for-claude

Blotato: https://www.blotato.com/

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/

#AIForBusiness #AISkills #AIWithKyle

Get the FREE Learn AI guide: https://aiwithkyle.com/learnai

Build your AI skills with the AI Confidence Accelerator: https://aiwithkyle.com/confidence

Hand a complete piece of work to AI. Learn the 4S approach to delegation, then use agents, skills and connections to get the work done.

Part 2 of the Chat, Delegate, Build series, adapted from my full How to Learn AI course.

00:00 Getting AI to do the work

07:15 Delegate with 4S

10:21 Agents, skills and connections

Get the FREE Learn AI guide: https://aiwithkyle.com/learnai

Build your AI skills with the AI Confidence Accelerator: https://aiwithkyle.com/confidence

Choose your AI assistant, build a daily habit, use voice and files, and get better answers with RISEN. We also cover the concepts behind the tools.

Part 1 of the Chat, Delegate, Build series, adapted from my full How to Learn AI course.

00:00 How to learn AI

01:18 Learn by doing

02:51 Chat, delegate, build

04:25 The chat checklist

10:40 Communicate with RISEN

14:45 Concepts to understand

20:07 AI Confidence Accelerator

Get the FREE Learn AI guide: https://aiwithkyle.com/learnai

Build your AI skills with the AI Confidence Accelerator: https://aiwithkyle.com/confidence

If I was starting to learn AI today, this is the path I would take. We work through three stages: chat, delegate and build.

I cover better conversations with AI, RISEN, Projects, voice and files, delegating real work with 4S, Skills and connections, and building your first small app. You will also see how I organise new AI information so I can actually use it.

00:00 How to learn AI

01:18 Learn by doing

02:51 Chat, delegate, build

04:25 Chat: the practical checklist

10:40 Communicate with RISEN

14:45 Chat: concepts to understand

20:07 AI Confidence Accelerator

20:54 Delegate: the practical checklist

28:17 Delegate with 4S

31:23 Agents, skills and connections

35:37 Build: the practical checklist

40:42 My tech stack

43:48 Build: concepts to understand

47:00 How to keep up with AI

49:16 Give new information a home

52:07 Inside AI Confidence Accelerator

54:32 Get the free guide

Want to build with AI but have no idea where to start? https://aiwithkyle.com/join-freedom-os

Join my private WhatsApp group: https://aiwithkyle.com/ai-with-kyle-group-chat

Media/Sponsorship Inquiries: https://aiwithkyle.com/media-kit

Get my daily AI newsletter: https://aiwithkyle.com/

Claude Opus 5.5 and GPT-6 Sol arrived within an hour of each other, and the comparisons are everywhere. Opus leads the early Artificial Analysis benchmarks. Sol and Luna make everyday AI work much cheaper. Which improvement will you actually notice in your business?

I walk through the published results, the six-out-of-ten benchmark lead, professional work scores and the demos people are sharing. Then we look at API prices, cost per task and why maximum reasoning effort can burn through your budget without improving the answer you need.

I also show how I compare models using a repeatable task from my own work, what Simon Willison's pelican test reveals, and where these models are available at launch. These are first impressions from published tests, other people's examples and my own early experience. Use a task you know well before moving your whole workflow.

The 50% price reduction discussed here is API pricing versus the previous Sol and Luna generations. Subscription prices and usage allowances are separate.

*RESOURCES*

Build one AI brain across your tools: https://www.youtube.com/watch?v=LqkglPK9SUE

My Instinct walkthrough: https://www.youtube.com/watch?v=y9UD7Vs3APE

*CHAPTERS*

00:00 Claude vs GPT: what changed

02:48 Opus 5.5 benchmarks

04:55 What the demos actually show

08:39 GPT-6 Sol and Luna

10:08 The API price gap

13:46 Test models on your own work

15:51 Effort levels and the pelican test

17:27 Update your prompts

18:48 Where to get the new models

20:42 Which model should you use?

*KEY POINTS*

- Opus 5.5 leads the early Artificial Analysis Intelligence Index.

- Sol and Luna cut API prices by 50% versus their previous generations.

- Compare completed-task cost as well as token prices.

- Try several effort levels on a repeatable task from your work.

- Check model access and refresh old prompts before switching.

*SOURCES AND TOOLS MENTIONED*

Anthropic announcement: https://www.anthropic.com/claude-opus-5-5

OpenAI announcement: https://openai.com/index/introducing-gpt-6-sol-and-luna/

Artificial Analysis, Opus results: https://x.com/ArtificialAnlys/status/2102438210798514391

Artificial Analysis, Sol and Luna: https://x.com/ArtificialAnlys/status/2102462962758033624

Matthew Berman's Unreal demo: https://x.com/MatthewBerman/status/2102483668468195539

Ethan Mollick's shader demo: https://x.com/emollick/status/2102441628661080384

AI/ML API visual comparison: https://x.com/aimlapi/status/2102515672710533417

Simon Willison's tests: https://simonwillison.net/2026/Sep/22/opus-and-sol-and-luna/

Lance Martin's prompt audit: https://x.com/RLanceMartin/status/2102575471502528989

Model availability: https://learn.chatgpt.com/docs/models#recommended-models

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

Website: https://aiwithkyle.com

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/

#Claude #ChatGPT #AI

Get your Instinct invite: https://aiwithkyle.com/instinct

Instinct is an AI assistant you can message through WhatsApp or iMessage, and it’s free right now. I’ve been using it to work through the jobs I keep putting off: sorting priorities, organising projects, bookkeeping and the small bits of admin that pile up. Here’s what I like about it, how it responds, and where I’d start if you’ve never used an AI assistant before.

I also walk through other people’s examples, including getting footage from the US Open, negotiating a lower internet bill and finding an apartment. Those are their results, alongside my own experience. Then I look at Instinct Files and the account access this kind of assistant can use. You can start with one ordinary job and limited access; forwarding a single email is an option before connecting your whole inbox. I’m not sponsored by or affiliated with Instinct. Access is invite-only, and its current free pricing may change.

*RESOURCES*

Instinct: https://instinct.com/

Instinct privacy policy: https://instinct.com/privacy-policy

*CHAPTERS*

00:00 Instinct, an AI assistant in WhatsApp

02:20 Get an invite

03:32 How Instinct works

06:33 An AI that knows when to stop talking

07:43 Projects, priorities and bookkeeping

11:17 What other people use it for

16:20 Instinct Files

17:48 Accounts, data and privacy

20:55 Start with less access

22:07 Try one job

*KEY POINTS*

- Message Instinct through WhatsApp or iMessage.

- Give it one real job you’ve been putting off.

- Use projects and daily briefs to keep track of your priorities.

- Check what access you’re granting before connecting accounts.

- Start small and check the result.

*SOURCES AND TOOLS MENTIONED*

Olivia Levine — getting US Open footage: https://x.com/oliviaalevine/status/2095705075633000815

Eli Weiss — lowering a FiOS bill: https://x.com/eliweisss/status/2099833405181436387

Tamara Winter — apartment search: https://x.com/tamarawinter/status/2099859471237361962

Instinct Files demo: https://x.com/noahrshinn/status/2101080443667767385

Forwarding an email without full inbox access: https://x.com/noahrshinn/status/2097443135177798060

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

Website: https://aiwithkyle.com

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/

#InstinctAI #AIAssistant #PracticalAI

Want to build with AI but have no idea where to start? https://aiwithkyle.com/join-freedom-os

Join my private WhatsApp group: https://aiwithkyle.com/ai-with-kyle-group-chat

Media/Sponsorship Inquiries: https://aiwithkyle.com/media-kit

Get my daily AI newsletter: https://aiwithkyle.com/

Donald Trump called Jensen Huang during the All-In Summit and called fears about AI a “hoax”. Days earlier, Anthropic’s Dario Amodei had proposed slowing the development of the most powerful AI systems. China has now responded too, and a global agreement looks a long way off.

I walk through the original call, China’s answer, Dario’s proposal, Mark Zuckerberg’s response and the safety measures OpenAI is putting forward. Dario’s own essay says full global pacing is unlikely any time soon. So what happens when the labs raise the alarm, politicians refuse to slow down, and development carries on?

My suspicion is that this outcome suits the labs rather well. That’s my interpretation of their incentives; it doesn’t establish that anyone coordinated it. I also look at the work that could still make a difference: independent evaluators, safety cases and reports of models behaving in unexpected ways. OpenAI’s published examples give us something concrete to examine. The decisions these companies make next will tell us more than their promises.

*RESOURCES*

My earlier video on the AI slowdown proposal:

https://www.youtube.com/watch?v=zPJsX12AdLI

*CHAPTERS*

00:00 Trump responds to the AI warnings

00:24 Trump’s call to Jensen Huang

02:09 What Dario proposed

03:49 China responds

06:50 Dario’s warning about a global deal

09:19 Zuckerberg’s response

14:34 Jensen and Nvidia’s incentives

16:54 What could actually change?

18:31 OpenAI’s misalignment reports

20:39 Who verifies and enforces the promises?

22:22 More on the AI pacing proposal

*KEY POINTS*

• Trump wants the US to win the AI race.

• China rejects the framing behind the proposed slowdown.

• Dario already described full global pacing as unlikely soon.

• Meta, Nvidia and the frontier labs have different commercial incentives.

• Independent evaluation and incident reporting can proceed while talks continue.

• Watch the labs’ decisions as well as their public statements.

*SOURCES AND TOOLS MENTIONED*

Trump and Jensen’s original All-In call:

https://x.com/theallinpod/status/2099621431621607829

Dario Amodei — We Must Pace the Frontier:

https://darioamodei.com/post/we-must-pace-the-frontier

China’s Foreign Ministry response:

https://cy.china-embassy.gov.cn/eng/fyrth/202609/t20260914_12021997.htm

Mark Zuckerberg’s response:

https://x.com/finkd/status/2099997096896274533

Sam Altman on safety cases:

https://x.com/sama/status/2099348812305473766

OpenAI’s model misalignment reporting framework:

https://openai.com/index/model-misalignment-reporting-framework/

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

Website: https://aiwithkyle.com

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/

Get my daily AI newsletter: https://aiwithkyle.com/

Anthropic wants to slow AI development down, and Sam Altman, Elon Musk and Demis Hassabis are backing the direction of Dario Amodei's proposal. These are usually fierce rivals. Now they're calling for a slower pace, outside evaluators and shared safety standards. I'm sceptical about what the biggest AI companies stand to gain.

I go through Amodei's essay, "We Must Pace the Frontier", and the full posts from the people supporting and challenging it. David Sacks tells the labs to slow down themselves. Cohere's Aidan Gomez calls out a possible cartel. Donald Trump wants America to stay ahead of China.

We also look at recursive self-improvement, OpenAI's growing use of coding agents, METR's role as an evaluator, and the problem of verifying an international agreement. AI safety deserves serious attention. So do the costs of delaying useful tools and making it harder for smaller companies to compete.

Recorded 14 September 2026.

*RESOURCES*

Previous video on Jacob Coxon: https://www.youtube.com/watch?v=6gcpXBLmwPE

Self-improving AI explained: https://www.youtube.com/watch?v=Os-s26O_W08

*CHAPTERS*

00:00 AI rivals suddenly agree

01:32 Dario, Sam, Musk and Demis

02:52 The Jacob Coxon connection

04:12 What does AI pacing mean?

05:21 Embedded evaluators

07:05 Pacing within democracies

08:36 Can a global slowdown work?

09:54 Recursive self-improvement

11:45 David Sacks: then slow down

16:19 Aidan Gomez and the cartel accusation

18:40 Who checks the checkers? METR

21:15 The China problem

22:28 Trump's response

24:06 Derek Thompson's four-point argument

25:19 Slowing down has a cost

27:19 Safety and commercial self-interest

*KEY POINTS*

• Rival AI labs broadly agree on pacing, but the rules are still being debated.

• Outside evaluators need access, independence and the ability to publish findings.

• Expensive compliance requirements can favour the biggest companies.

• Any international slowdown faces the problem of verifying what other countries do.

• Delaying useful AI carries costs for medicine, work and smaller businesses.

*SOURCES AND TOOLS MENTIONED*

Dario Amodei's essay: https://darioamodei.com/post/we-must-pace-the-frontier

Dario's announcement: https://x.com/DarioAmodei/status/2098773920774074715

Sam Altman's full post: https://x.com/sama/status/2099348812305473766

Sam's initial response: https://x.com/sama/status/2098811563415150910

Elon Musk: https://x.com/elonmusk/status/2098789109980332057

Demis Hassabis: https://x.com/demishassabis/status/2098909516582490602

David Sacks: https://x.com/DavidSacks/status/2098973625252708460

Aidan Gomez: https://x.com/aidangomez/status/2098828132396449939

OpenAI's research acceleration data: https://openai.com/index/research-acceleration-view-inside-openai/

METR's disclosures: https://metr.org/about

Trump's original video: https://x.com/RapidResponse47/status/2099155148576542928

Derek Thompson: https://x.com/DKThomp/status/2098819497821917480

Ethan Mollick: https://x.com/emollick/status/2099134552807522610

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

Website: https://aiwithkyle.com

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/

#Anthropic #OpenAI #ArtificialIntelligence

Want to build with AI but have no idea where to start? https://aiwithkyle.com/join-freedom-os

Join my private WhatsApp group: https://aiwithkyle.com/ai-with-kyle-group-chat

Media/Sponsorship Inquiries: https://aiwithkyle.com/media-kit

Get my daily AI newsletter: https://aiwithkyle.com/

Subscribe and turn on notifications to catch the next live stream:

https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

Jacob Coxon resigned from Anthropic and accused the leading AI labs of gambling with our lives. Then came the interviews, the donor connections people started digging into, and accusations that the whole thing was a "psyop".

I go through Coxon's original warning and the responses from Beff Jezos, Parker Thayer, Elon Musk and Bernie Sanders. We also look at recursive self-improvement, what Anthropic's own researchers have said, and who would get to define and enforce a superintelligence ban.

Recorded 11 September 2026.

*CHAPTERS*

00:00 The resignation and the accusations

01:59 What Coxon actually said

05:50 Who is Jacob Coxon?

07:43 Why people think it was organised

14:22 The interviews

17:46 The pushback

24:04 What recursive self-improvement means

25:51 What Anthropic's researchers are saying

27:52 Where the proposed solutions overlap

29:57 Who benefits from slowing down?

34:33 The commercial angle

36:57 More than one thing can be true

*SOURCES AND RESOURCES*

Jacob Coxon's original post:

https://x.com/hilbertspaess/status/2097476196791709843

Parker Thayer:

https://x.com/ParkerThayer/status/2097759699626328575

Beff Jezos:

https://x.com/beffjezos/status/2097818903321993315

Elon Musk on the post's reach:

https://x.com/elonmusk/status/2097908044344746406

Bernie Sanders:

https://x.com/SenSanders/status/2098078825909219721

Evan Hubinger:

https://x.com/EvanHub/status/2097497037956891126

https://x.com/EvanHub/status/2097528891846074828

Anthropic on alignment and security:

https://www.anthropic.com/news/improving-alignment-security-efforts

Sanders and Casar's proposed legislation:

https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-artificial-superintelligence-and-temporarily-pause-advanced-ai-development/

Anthropic's policy proposals:

https://www.anthropic.com/news/2028-ai-leadership

*LISTEN TO AI WITH KYLE*

Spotify: https://open.spotify.com/show/6pnwfIJvxfC8jlX891XBiE

Apple Podcasts: https://podcasts.apple.com/us/podcast/ai-with-kyle-daily-ai-news-and-updates/id1825779580

*CONNECT WITH KYLE*

TikTok: https://www.tiktok.com/@iamkylebalmer

Instagram: https://www.instagram.com/iamkylebalmer

LinkedIn: https://www.linkedin.com/in/kylebalmer/