Tom Panos

The First 11: A Jordan Legacy Explore the legendary first 11 Air Jordan sneakers from 1985-1995. A stunning showcase of Michael Jordan's iconic footwear legacy.

#Google Gemini 3.0 looks like it will drop on the 18th, but you can try it early by accessing the latest checkpoint by using the mobile app in canvas mode.

This site was built in one-shot. The UI & animations are pretty flawless. Impressive stuff.

See it here → jordanhistory.netlify.app

#AI

I’ll absolutely give it a try. Thanks for the recommendation.

I think my overall opinion is that the ease of access to creating something that looks “good enough” will have a larger negative impact vs a positive one unfortunately.

That said, there will always be those who use it effectively.

I completely agree. I’m not an artist. I’m actually writing a follow up to add a bit more detail, but you make a good point.

Personally, I think if you are yourself an artist and you decide to use AI as a tool that’s much different, and not really what I was trying to focus on.

Why AI Art & Media Is Useless As someone who works in AI and genuinely believes in the value and power of LLMs to make professionals more useful and valuable, I can confidently say that I hate everything about AI image/video/music...

Check out my recent article on why #AI art is entirely useless and serves to only further denigrate the real value of using ai as a tool for professional (not creative) work.

Read it here → tom-panos.com/all-posts/bl...

(No, I didn't generate that AI art myself, I found it at slop central, X)

Google Workspace Flows - About Google Workspace Flows Google Workspace Flows Google Workspace Flows was developed to automate tasks between Workspace apps. It connects directly with Gemini as your AI agent for research, analysis and content generation. H...

#Google just released a SUPER useful tool called Agents & Flows.

I just built a flow that reviews upcoming meetings. If there is a Meet or Zoom link, it pulls together a summary of all my communication w/ the attendees. Meeting prep done.

Learn more → sites.google.com/view/workspace-flows/about

Here's an example of the #ManusWideResearch feature I used for an article I want to write.

I used it to analyze this image & create a table of credible sources all the transactions.

I use the @notion.com connector to add it to a page in my notes database.

See it here → tinyurl.com/246dys6t

#AI

Manus | Wide Research: Beyond Context Window Deploy hundreds of AI agents in parallel and overcome context window limitations with parallel multi-agent architecture.

If you're a researcher, or do research frequently, this is my favorite feature. It's effective and pretty much set and forget.

It's called "Wide Research" and it does about the best job (maybe #Claude) outperforms it slightly) collecting sources.

manus.im/features/wid...

Manus | From prompt to web apps Build intelligent, full-stack web apps from a single prompt. No coding, no drag-and-drop — just describe what you want, and Manus does the rest.

For anyone looking for a solid #AI #Agent that can execute genuinely useful tasks with super high consistency, check out Manus @manusaihq.bsky.social.

I've tested most major platforms. For most, I think it will deliver the most useful outputs.
→ manus.im/features/web... (not an affiliate link)

I could genuinely see #OpenAI taking the company public sometime in the next few years.

It would boost their value, attract investment, and solve their core issue: over $1 trillion in obligations with insufficient revenue.

#AI #AIBubble

Yep, just noticed "Oracle" is for some reason "Ontcle" in the visual. Great work #Gemini.

#Nvidia #OpenAI, #Oracle and the rest of the companies that make up the #AI Bubble are starting to look more and more like a Ponzi scheme. I created this sophisticated visual using reporting on every transaction I could find.

→ Full outline/sources here: tom-panos.notion.site/ai-ponzi-sch...

Hey Sam, you also can’t declare nearly a trillion $$ in future revenue as though it’s certainly going to happen…

If we can’t control MechaHitler, how will we steer AGI? It's Chana and Aric again. Firstly, we cannot believe the response to our last video. Thank you all so much for creating an initial swell of views and support, for your feedback and comments and subsc...

We will only ever be as safe as xAI (the least safe AI lab), which should terrify everyone.

Here's a great article on the topic by @80000hours.bsky.social

→ tinyurl.com/24qulc9h

While I'm sure most people remember the #Grok "MechaHitler" debacle. I don't think everyone truly grasps the existential horror that is #xAI.

#OpenAI has been at it 10+ years & #Anthropic, 4+ (CEO was prev VP, Research @ OAI).

XAI: 2 years - 0 published research - Owns the most AI GPUs on earth.

@zohrankmamdani.bsky.social‘s campaign game is so many levels above any other politician in America right now it’s not even close.

#Mamdani2025

Six days left and a whole lotta doors to knock.

Let's get it: zohranfornyc.com/gotv

Well who could’ve seen this coming…

This man is solely an accelerating us towards the AGI apocalypse.

This is an absolute masterclass is how to create a political ad.

My god, the irony of the billionaire-backed candidate using bottom of the barrel AI to generate racist slop ads.

Meanwhile, @zohrankmamdani.bsky.social is out here making the best ads i've seen in years.

Great to hear about this, just found a park near my place where we can go and see it!

As Toni Morrison said, "the very serious function of racism is distraction. It keeps you from doing your work." But she also reminded us never to lose our capacity to be shocked.

Part of my conversation on @ayman.msnbc.com earlier today.

KFC website in 2002

KFC website in 2002

#WebDesignHistory

The richest man on earth owns X.

The second richest man on earth is about to acquire TikTok and his family could soon own both Paramount and Warner Bros.

The third richest man owns Facebook, Instagram, and WhatsApp.

The fourth richest man owns The Washington Post.

See the problem here?

The #FBI director, Kash Patel, is what I call an "under the bus guy."

In negotiation it's helpful to have this guy, who you can throw under the bus if shit goes off the rails. Generally, the guy knows he's potentially under the bus.

Kash has no idea, but he's that guy for sure. He's gone soon.

YouTube video by Vince Vintage The Disturbing Truth Behind Catch Me If You Can

This incredible video by @vincevintage.bsky.social just shattered my lifelong, and pretty positive, opinion of Frank Abagnale (the Catch Me If You Can guy).

An excellent watch → tinyurl.com/23e6kb3u

#cons #scams #fraudprevention #stopscams

We're quickly headed to a point in time where literally every prominent republican has some type of racist text message scandal.

#OpenAI just released a new agentic browser (I know — what a shocker). I’ve been using Comet by #Perplexity for a while now and have actually found it quite useful.

I’m testing both browsers against eachother using the same prompts and will post the results soon.

Early favorite: #Atlas - its good.

Who Profited from the recent Tariff Announcement? Someone in the #Trump admin?

I've identified the 10 individuals most likely to be responsible for the massive #Bitcoin & #Ethereum shorts that made them nearly $200M in 24 hours.

Full report here → trumps-insider-tariff-trade.netlify.app

Just think about the irony of Trump literally begging for the Nobel Peace Prize and compare that with Bob Dylan, who didn't plan to even show up to the ceremony. Trump may have the words vibes of any human man. Ever.

Sorry my friend, you are just 100% provably wrong.

LLMs aren’t just word suggestion tools, and their output isn’t just random (RNG). They use neural nets to learn deep patterns, so their “emergent” behaviors come from complex data. Emergent behavior comes from training...not anything like RNG.

a man in a suit and tie is standing in front of a trash can and saying `` i understand ... nothing . '' ALT: a man in a suit and tie is standing in front of a trash can and saying `` i understand ... nothing . ''

It's very scary just how many people think that LLMs are no different than any other software/app.

1. They aren't "programmed," they are trained.
2. Fact: We don't understand the emergent behavior of LLM's
3. "Human" sourced data is nearly gone. Future llms are trained on synthetic data.