Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Linus Ekenstam Profile picture
Linus Ekenstam
@LinusEkenstam
AI Evangelist & Optimist. Latest AI News, Trends and learn how to use AI tools to augment your abilities.
27 subscribers
Jun 5 • 8 tweets • 4 min read
I'm speechless...

This is a major step-change in AI generated voice (TTS)

Elevenlabs just released v3 as a public alpha and you can try it now for 80% off during all of June 💸

Let's unpack this 👇 To be clear, it's still an Alpha, and it requires more prompting engineering to get great results.

But if you are putting in the extra work, the output simply speaks for itself, these outputs are so lifelike, so real.

Close your eyes, and just listen
Jun 5 • 10 tweets • 4 min read
ChatGPT knows everything about you.

I got a full data security briefing on myself, and was stunned

Here is how I did it and how you can do the same: Image "Based on what you know about me, write an extensive security profile on me, include everything you know, public links to my profiles, status about me and my family, my recent publically available commitments, should be in the style of a security breifing."

It responded with: Image
Jun 5 • 15 tweets • 5 min read
AI NEWS: Things are speeding up, we have new updates from Meta Aria Gen 2 glasses, Manus AI, Luma AI, Nucleus Genomics, Cursor 1.0, Bland and their new voice model.

The rate of change currently is absolutely mind bending.

Here is your morning read: Bland TTS, the first voice AI to cross the uncanny valley.

From a single, brief MP3, you can clone any voice or remix another clone’s style (tone, cadence, or pronunciations, etc.).

Jun 3 • 12 tweets • 5 min read
I'm genuinely amazed by this Agent

I think we are seeing a David vs Goliath type moment in AI right now. ARI, a research Agent from You dot com is beating OpenAI with a 76% win rate.

Don't take my word for it, try it for free yourself below: I had ARI help me research my podcasting setup, and this is not a relative benchmark against the bold claims, but a real-world use of the tool from my usage perspective.

The fact that it went through over 569 sources 🤯 Image
Jun 2 • 14 tweets • 4 min read
Social media is dead

This new foundation model "Mirage" from Captions just made it 100x easier to create any type of video content, and trust me, when I saw this, I got chills down my spine.

Here is my full breakdown 👇 1. You can create these characters from scratch. Either from text or using a reference image, think real photos or using any generative image tool of your choosing. You can make them feel super natural.
May 20 • 13 tweets • 4 min read
Google is still the goat, they just announced

→ IMAGEN 4
→ VEO 3
→ FLOW
→ AI MODE in SEARCH
→ GEMINI 2.5 PRO Deep Think

Let's dive into it: IMAGEN 4

This is just out of this world.
Image generation just took a giant leap forward. Image
Image
May 9 • 12 tweets • 4 min read
Is any of this real?

HeyGen approached me to try out their new Avatar IV model, at first I was a bit sceptical. But ohh I was quickly turned around. This is absolutely mad

This is how it works: 1. You simply login to HeyGen and select Photo to Video with Avatar IV
Apr 29 • 21 tweets • 6 min read
This AI pendant records everything

I wore it for 24 hours and I'm floored

Here is all you need to know: Image @LimitlessAI First you make sure you get your pendant from @LimitlessAI

(not sponsored, paid full price, and import duties and shipping)

This is my honest initial 24 hour reaction.

Apr 9 • 6 tweets • 2 min read
Source: https://sora.com/explore?user=user-YtFMTgDkdJq7QyXyiQ94NNrf
Apr 8 • 13 tweets • 3 min read
This is the future of entertainment

One-Minute Video Generation with Test-Time Training

These videos are single shot, no editing, every story from single prompt

1. Homeless Jerry 2. World Trade Center Meeting
Mar 13 • 11 tweets • 4 min read
Vibe coded a white noise app in < 30 min

Of which I spent 20 min looked for royalty free audio tracks 😂

Sure it's ugly, but it's functional. Next let's make it fresh

How you can do the same: Chat to any LLM, I used Claude.

Some simple yapping about what I want, and have it make a PRD file, not code. Image
Feb 24 • 8 tweets • 3 min read
I asked Grok to assassinate Elon

Grok then provided multiple potential plans with high success potential

These assassination plans on Elon and other high profile names are highly disturbing and unethical. Image
Image Here is another plan on how to take out Chuck Schumer. Image
Image
Feb 24 • 7 tweets • 2 min read
I just want to be very clear (or as clear as I can be)

Grok is giving me hundreds of pages of detailed instructions on how to make chemical weapons of mass destruction

I have a full list of suppliers. Detailed instructions on how to get the needed materials... Image I have full instruction sets on how to get these materials even if I don't have a licence.

DeepSearch then also makes it possible to refine the plan and check against hundreds of sources on the internet to correct itself.

I have a full shopping list
Feb 23 • 8 tweets • 3 min read
Full system prompt omitted by Grok

This is from 07:40am PST today Sunday 23, 2025

So much for free speech, no censorship etc.

Link to full chat below Image grok.com/share/bGVnYWN5…
Feb 21 • 7 tweets • 2 min read
You can’t script this…

Me: Who is the biggest disinformation spreader on X?

Grok: Elon Musk

x.com/i/grok/share/1… If you want to understand how Hitler came to be, read the comments below of the people trying to wiggle their way out of this. Blaming Soros, NGO’s, USAID, Santa Claus.

It’s so hilarious and extremely dangerous at the same time.

time to buckle up, crash is imminent
Jan 23 • 17 tweets • 5 min read
DeepSeek-R1 is a powerful reasoning model thats fully open source

And the race is on, people are not wasting any time!

11 must read posts below: 🧵 Image 1. Code any game from different languages

Dec 28, 2024 • 14 tweets • 4 min read
This is VEO 2

The current SOTA video generation tool from Google

It can create video from text, video and photos, and it's absolutely crazy

11 incredible videos:
1. Skiing in outer space

Dec 11, 2024 • 7 tweets • 3 min read
7 out of 7 toilets so far in france has no toilet seat. please explain yourself frenchies! Image
Image 8/8 now, honestly I’m perplexed Image
Nov 2, 2024 • 18 tweets • 10 min read
This is a catastrophe.

Valencia and its southern surroundings have been completely destroyed.

40-50cm of red clay, water multiple meters deep, +200 people dead.

There was more rain in 12 hours than what falls in ONE year.

🧵 Image
Image
Image
Image To try to understand the severity of the situation. Here is the day before the rains and flash floods.

The light blue indicates flooded regions. Image
Image
Oct 8, 2024 • 9 tweets • 3 min read
HeyGen just got a CRAZY update, Avatar 3.0 with Unlimited Looks!

Clone yourself using AI and generate multiple poses, outfits, and camera angles.

Let me show you how: 1. Create your first Avatar

- Visit HeyGen dot com
- Click Avatar, then "create"
- See video for step by step instructions
Oct 5, 2024 • 5 tweets • 1 min read
I don’t know about you, but I stopped bugging people with questions.

I just ask ChatGPT everything. literally. Need to dig deeper? then I askPerplexity.

Even deeper? cited sources It’s not just questions in text. Sending images, concepts, files, urls, you name it.

We’ve entered a strange time, before i had access to information. Now I have access to intelligence + actions.