
🔥 Building @aranx_com 🤖 ex AI & Computer Vision @JohnDeere ⚡️ 3-time founder | @joinstationf S17 Also 🇫🇷👶🏼👶🏼🎾✈️🥐
Joined June 2008
- Tweets1.6K
- Following931
- Followers432
- Likes3.6K
turns out software engineers will not be replaced by AI
they’ll have full-time jobs to roll out patches 250 times per day, every day of the year
Announcing 3 Critical vulns for hundreds of millions of IoT, Firewalls, OT, and Military devices including: Palo Alto Networks, Comcast, D-Link, HP, Canon, Honeywell, GE, Oracle, Siemens, Rockwell, Digi, Lockheed Martin, Boeing...
The vulns are with GoAhead Web Server which is a backend library used by these devices. There's a patch available from GoAhead but that doesn't mean there's a patch available for the devices yet -- that's up to the manufacturer.
- TLS traffic over DHE ciphers have a 1-byte secret, "04", meaning anyone can decrypt connections to these devices and steal login credentials and session tokens if the client negotiated with DHE ciphers like the latest TLS_DHE_RSA_WITH_CHACHA20_POLY1305_SHA256, which is extremely likely.
- LibreSSL mTLS has been completely bypassed, meaning anyone can connect over mTLS. If you have one of these devices listening on the internet and are expecting mTLS to protect it from attackers, take it offline or patch asap.
- OpenSSL Certificate Revocation Lists (CRLs) are not enforced, meaning if a user device is stolen or compromised, and you revoke that mTLS cert, nothing actually happens. There are a lot of other attacks this can be used for.
These are critical because these types of devices run our world. Water systems, electricity, military, firewalls, critical infrastructure.
foxio.io/blog/choose-your-ba…
SEO is dead
Btw Google is sending 70x more traffic to websites than ChatGPT right now:
1. Google: 23.96%
2. Bing: 1.13%
3. Yahoo: 0.76%
4. Yandex: 0.39%
5. ChatGPT: 0.34%
6. DuckDuckGo: 0.23%
7. Ecosia: 0.05%
8. Meta AI: 0.05%
8. Baidu: 0.03%
10. Brave: 0.02%
11. Gemini: 0.02%
12. Claude: 0.01%
13. Perplexity: 0.01%
Google is still #1 priority.
If you use AI-generated icons/illustrations in your app, you don't have to pay for both the light AND dark theme versions.
There is this nice little CSS trick that makes the light version of the illustration also work in dark mode.
(The video below shows the assets generated for one of our ecom customers that sells skincare products)
So why do we need it and how do we use it?
Well, when you connect your website to Aranx, one of the many things the platform does is research the categories of content that would produce high-opportunity (high intent / low competition) SEO content for your blog.
Each of those categories is called a Content Strategy and is used by Aranx to research high-opportunity keywords and AI researched terms.
We've learned that adding an illustration makes it much clearer to understand for our users.
But since it is an image with a transparent background, it just doesn't work well when you use the dark theme.
The obvious fix is to generate a second image for dark mode.
BUT:
1. It doubles the AI cost.
2. Worse, it’s a different drawing. One prompt never gives the same image twice, so your light and dark cards don't 100% match.
But there is a simple trick that solves it with 1 line of CSS:
filter: invert(1) hue-rotate(180deg);
In Tailwind
dark:invert dark:hue-rotate-180
How does it works?
invert() flips lightness, so dark ink on white becomes light ink on black. It also flips hue. Blue turns orange. Coral turns teal.
• hue-rotate(180deg) spins the hue back.
• Lightness stays flipped.
• The colors return.
• Transparent pixels stay transparent.
• Only the lines flip, and the card color shows through.
With gpt-image-2.5 that’s one parameter:
background: "transparent"
Just make sure to generate a PNG.
Light mode: $0.12
Dark mode: $0
Every dark card matches its light twin exactly, because it’s the same file.
this is where you see that frontier LLMs have become good enough for pretty much everything the 99% needs.
what comes next: massive price drops and a commoditization of intelligence.
the annoying side effect of OpenAI changing its subscription model is that it will give Anthropic every reason to nerf Opus 5.5 and Claude subs to the ground
I’ll explain the new Pro 200 plan differently, before I start live tweeting from DevDay on things that are going out!
Today we are going to ship a number of things that increase what you can do across the Plus and Pro plans. A lot of compute is online for this increase. As we increase the floor, we are changing the relative difference between plans to be
Plus = 1X
Pro 100 = 5X
Pro 200 = 10X
and we are reopening subscriptions for Pro 200 (we had paused it). If you have an existing plan you will keep the 20X multiplier for a bit and also receive a lot of additional credits because we know changes are hard even if it means that everyone will get more in the end.
Interesting to notice that 𝕏 uses all 3
i find it funny how endlessly people debate business models when historically there have basically only been ~three ways to make money.
- charge ppl.
- show them ads.
- take a cut of a transaction.
almost everything else is just a more elaborate wrapper around one of these. so, why are there courses on this?
The moat of having agentic AI built into your app is starting to fade.
LLMs are about to become as ubiquitous as electricity or water.
And flow everywhere into the apps you make.
Yes, you just heard that right.
Got @ChatGPT Plus or Pro? That subscription now covers your AI usage in Notion.
Just connect your plan, pick a GPT model, and go.
Readers added context they thought people might want to know
This feature requires a Notion Business or Enterprise workspace plan (in addition to ChatGPT Plus/Pro).
notion.com/help/use-your-…
notion.com/pricing
I just made a 20 minute-podcast for less than $2 and the results were insane.
And nobody could tell it was not my voice but AI.
It’s finally the time for Audio to become a first-class citizen in the AI revolution.
And it only took 3 simple steps:
- made an @ElevenLabs voice with a random 30s recording.
- Opus 5.5 deep researched data on an interesting topic with @firecrawl + wrote an elevenlabs V4 script.
- sent everything to elevenlabs V4
The results are simply on another dimension compared to V3.
Unbelievable.
Introducing Eleven v4 and Eleven v4 Turbo, our fastest and most emotive voice models yet.
Ranked #1 by Artificial Analysis.
















