sharing my ai journey here ex cmo @LeonardoAi (0→1M in 5 months, acq. by @canva) ex growth @delphi_labs grew fb pages to millions now i advise & build
Joined March 2012
- Tweets14.4K
- Following984
- Followers67.5K
- Likes93.8K
Pinned Tweet
most people use AI by telling it what to do
but... some of the most powerful prompts I've found don't tell AI to do anything
they make it question what it's already doing
one simple question can completely change how an AI approaches a problem
I've been collecting these prompts for a while
and I'll be sharing them here, one at a time
bookmark this thread, I'll keep adding to it
Midas 👑 retweeted
ChatGPT and Codex are helping engineers build for the physical world.
We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
Midas 👑 retweeted
Day 14 of 30 with @boardyai. Samsung Portugal's CMO once sat across from me in Lisbon and asked "but what is Discord? I've never heard of it."
This was around 2022. My BD guy, one of the co-founders of Comic Con Portugal, had got me in-person meetings with the CMOs of Samsung, Asus and the like. A week after ours, she called him back. Her son had just told her Discord is what he uses every day to play Fortnite with his friends.
I told that story yesterday on a one-hour call with someone @boardyai connected me with, who builds community intelligence tooling for studios and brands. We ended up on how badly big companies run their own Discord communities.
In that meeting, the person who signs off on the budget had never heard of the platform, and the one who used it every day was her teenage son lol. A lot of brands still set up their servers that way. Somebody gets hired as community manager, and the title doesn't come with the authority to change anything about it...
Midas 👑 retweeted
As AI becomes more capable at finding vulnerabilities, defenders need to put it to work before attackers do.
We’re planning our next internal security sprint at OpenAI and partnering with Cathedral Systems, whose researchers will use our models alongside our teams to challenge our defenses. They’ll bring differentiated expertise and approaches to our ongoing work to find and fix weaknesses with AI.
We’re also raising bug bounty rewards, including more than quadrupling rewards for critical findings. Exceptional findings will be eligible for up to $172,000, including bonuses.
We’ll keep advancing our models, using them to strengthen our own systems, and expanding access for trusted defenders.
openai.com/index/preparing-o…
Midas 👑 retweeted
Fellow resetters, heads up that we're working on expanding the family of frontier models that power dots, including GPT-6 Astra and GPT-6.1 Terra and some others in the future. Terra brings speed and a nice personality upgrade, and will help us bring dots to more people.
Midas 👑 retweeted
We know some Git-fu. Excited the team is finally able to talk about it a bit. Massive props to @tnm and team
Midas 👑 retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Midas 👑 retweeted
So @Ubuntu still hasn't patched this open hole on any Ubuntu server on the public internet that lets anyone take it over
We might just have to switch all our servers over to Debian in this new post-AGI time of rapid cybersecurity hacks which require just as rapid patching
Slowpokes will be left behind
chalmovsky.com/posts/wtf-ubu…
This is why I said you want your VPS ports all closed
Then only let it be accessible inside your Tailscale for SSH and via a Cloudflare Tunnel to receive web traffic
Any Linux box can be taken over with just a single TCP port exposed to the internet now
A stop gap is disabling MPTCP on your VPS now but even better is not having any public ports exposed!
Midas 👑 retweeted
@MachgenAI is coming to London 🇬🇧
We're teaming up with @WonderStudiosX and @LocalGlobe on a hackathon for creators and developers building with diffusion models.
Sign up and come build with us 👇
The frontier of generative video just moved up a gear. 🚀
@localglobevc x @MachgenAI x @WonderStudiosX are co-hosting a hands-on hackathon at Wonder Studios’ East London HQ to explore what happens when video generation runs in real-time without gaps, opening up brand-new possibilities for interactive media and gaming.
What to expect:
✨ Model access & credits provided by @MachgenAI
✨ Showcase by @JustinHackney33 (Co-Founder/CCO at Wonder)
✨ Open-ended building & team prototyping
If you're a builder, engineer, or creative technologist working at the intersection of AI and media, apply to join the room!
Link in thread!
Midas 👑 retweeted
If you've got some extra tokens to burn, check out my 'duo' skill.
It gives 2 models the same task. They both work blind, and neither sees the other's work until they've finished.
Then the main agent checks what each found and merges the results.
Been using this a lot more since I've been trying to actually put my GPT subs to use.
Linked below if you want to check it out.
Midas 👑 retweeted
I asked Claude: "What types of businesses outside of the labs will get stronger as raw intelligence explodes?"
Honestly, spot on response.
Midas 👑 retweeted
With five full years in the books, today is day one of year six for me at @OpenAI. I joined as the first sales person and a founding member of the GTM team.
OAI was a ~150 person research lab with an API and a bold mission.
While Davinci may be the real OG, InstructGPT is my first true love.
Here’s to five more years.