Initial thaughts after doing some experiments with Jev:
1. X is 80% filled with garbage usecases and lies just to ride the hype of Jev to go viral or promote their product. Game is game, i respect that.
2. the 20% of x that actually experiment with Jev is a goldmine, I believe jev looks flashy as demo but the strength of Jev is to focus on boring stuff. More on that in 3
3. JEV can be extremely powerful in classification of Boring tasks / System 1 tasks, which Jev is btw mainly for.
Couple Examples:
- Classify Invoices
- Classify support ticket and decide on urgency, on who should grab it
- Classify customer Intend through out a periode of time to decide e.g on churn , purchase intend etc
- Structured Voice Commands and decision making like for robotics or Home assitance.
- Realtime Entertainment Apps and Games
I am personally very excited about the last one.
2
103
Higgsfield is now on VidGuy!
You can now combine the power of @higgsfield with the abilities of VidGuy.
- deploy 50+ managed TikTok profiles
- create AI Influencers
- generate viral content formats
- post, analyse and iterate
Everything on Autopilot!
You simply evaluate the progress. Vidguy does the rest.
2
81
Mark is on his X Villian Arc lool
"Meta delayed shipping Muse for several months to focus on safety and security. We didn't call for everyone else to do this before we would. We just did it as part of our day-to-day work because it was clearly the right thing for people and for us. I'm proud of the security foundations we've built."
Last month I wrote about how we can build a positive and safe future for everyone: meta.com/thefutureisforevery…
Every lab has the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens.
The reality is:
- People won't want to use agents that are misaligned with them and that don't do what they ask, so labs have a strong natural incentive to make their models more aligned.
There is a lot of debate about slowing progress on capabilities until alignment catches up. My view is that trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models. Any lab that doesn't focus on alignment will fall behind.
- Labs face significant liability if their models cause harm, so they have a strong incentive to prevent this as well.
Meta delayed shipping Muse for several months to focus on safety and security. We didn't call for everyone else to do this before we would. We just did it as part of our day-to-day work because it was clearly the right thing for people and for us. I'm proud of the security foundations we've built.
- Engaging independent evaluators and advisors is industry best practice. MSL already does this today in several areas because it helps produce better work. Other labs can just do this too. In general, it would be helpful for there to be a larger and more diverse ecosystem of evaluators.
- Committing the significant majority of compute towards serving people rather than racing towards recursive self-improvement is one of the best ways to ensure we develop this technology safely. Meta has made this commitment and other labs can do this as well.
I believe the key to building a positive future for everyone is maintaining the right balance of power. This is within our power to do.
98
IPOS are in shambles
What type of magic is this!
Super exciting time for real time Experiences.
Minimax H3 MAX turbo and jev = Consumer apps ?
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
1
1
231
If you are not H3 Fast Maxxing through @reactorworld , what are you doing then?
If your are not H3 Turbo Max maxxing through @vidguyai
What are you doing then?
1
1
90
Wild!
Introducing the world’s first AI that turns TikTok viewers into customers.
What it does for you:
- Access 1000+ hosted & warmed up accounts
- AI Influencer vault with 100% character consistency
- Hyperrealistic slideshows that go viral + convert
- Clip, Edit & distribute existing content automatically
Everything. On. Autopilot.
Try it today for free!
92
Reach and distribution is defintiely the moat, but I think many people underestimate that @marclou has built a very convincing narrative for his "Product"
Untill now, everyone in the indie hacker space who jumped on this trend after @jonathan_wilke kicked things of with outbid, sold the idea of "hypothetical reach and attention"
"Your Brand on my laptop"
"Your brand on my car"
"Your brand on my wife " lol
But where does the car go? What city? Does it park in front of Apple store or in your garage?
Marc took the idea and attached tangible hypothetical attention.
1. Him as a distribution cahnnel
2. Hyrox event as a distribution channel
3. His audience and Hyrox audience as Reach
No matter what you think about hyrox, the sport attracts above average successful people in life, this might be not only my opinion when looking at @romanbuildsaas actually paying 8K for marcs chest (Never skip chest day guys)
Of course roman is a master marketer, so he will milk this for x and linkedin etc, but his company is not charity either, so every dollar spend needs to perform for a vc backed company with immense pressure to grow
I learned a lot by watching the events happening. Thanks!
Congratz to marc for the good execution!
being early means nothing if nobody is watching
you can have the idea first
someone bigger copies it
and makes the money
distribution is everything
1
148
This absolutley true.
The new H3 Max by @fal is an absolute game changer.
I believe it enables very new category of products. We are working on couple ideas as we speak.
As someone who is actively building in this Space, I have to say it is still to expensive to make sense except for "branded Streams"
Doing the math : 10c per second x 60 x 60 = 360 USD per Hour of stream.
However streaming Revenue or Reach can make this worth alot.
Examples:
- Sponsored Livestream
- Tiktok Live Shopping
- Ebay Live Shopping
- Guerrilia Launch Campaigns
- AI Companion Products
Excited to Showcase something very soon. :)
Today is a very historical moment for AI video generation
You can now generate AI video faster than you can watch it
Before it'd take let's say 2-5 minutes to generate 15 seconds of video
@fal made a post-trained Minimax H3 variant called Max which is 50x faster than the original but still maintains quality
It generates 15 seconds of video in 9 seconds!
That means you can now do new things like build a perpetual livestream with it that never ends!
1
2
246
When I traveled to Shenzhen this year, I discovered Bilibili for the first time. My VPN did not work so I went into Bilibili rabbit hole.
The Culture there is so different. It might be an undervalued attention for western Creators in the early months of global release.
I am sure Gary Vee will say the sentence above very soon.
He better quote tweet me. 😇
🇺🇸🇨🇳🔥Apple has approves the 'BiliBili App' on the App Store (Globally) and may fully rolled out as of September, 2026 after PlayStore, and PC.
- Chinese YouTube (BiliBili) will soon add millions of users and hiring community managers from US, Europe, Africa and Asia.
2
222