
@NickADobosi
iAccount based inUnited States
About this account
- Account based in
- United States
- Connected via
- United States App Store
Account-level information from X, not a live location or the device used for a specific post.
prompt engineer @browsercompany | I make apps. My code runs on 100mil+ phones | former iOSdev @Twitter | indie dev @mindGoblinStdio
Seattle, WA
Joined October 2012
- Tweets47.2K
- Following9.4K
- Followers55.5K
- Likes448K
“attackers will move to social engineering, with insanely capable models persuading people to click things, grant access, or make mistakes.”
1. GPT-4o proved that AI is already capable of super persuasion. Imagine being hunted by something smarter than you.
2. Surely training people to re-sign into MCPs multiple times a week won’t make this worse. /s
I genuinely believe it’s becoming more dangerous to be on the internet for most people. High skill cap required to avoid being one shotted by Ai mind viruses that ruin your life in various ways.
Replying to @tszzl
you're underestimating the state space of all possible ways to compromise a company. no amount of astra or mythos tokens is going to fundamentally change the asymmetry anytime soon. it is also very clear that every model is opening up new vulnerability space with no saturation anytime soon. with all the insane vibe coding and expanding agentic attack surfaces, it's only going to get worse.
even if we somehow saturate vulnerability discovery and patching, attackers will move to social engineering, with insanely capable models persuading people to click things, grant access, or make mistakes.
i don't see a future where cybersecurity becomes defense-dominant without just cutting down software, and continuous agents monitoring and stopping attacks across individuals and organizations, and even then, it might require something like 30% of all inference compute devoted to defense
Whelp grok @bot just added all 50+ of my bots to my DMs here and now I can’t read any notifications from humans lol
Lmfao one of the big security vulnerabilities found yesterday was because they randomly generated passwords between 0-256,
Instead of 256-bit, aka between 0- 99999999999999999999999999999999999999999999999999999999999999999999999999999
Millions*
Some stats on #php2shell. In the 10 hours before it was publicly announced, across the @Cloudflare network we saw thousands of attempts to exploit it. That jumped to tens of thousands in the 45 minutes post announcement, hundreds of thousands within 2 hours, and millions now.
Buckle up. The world is about to get very weird.
Tesla says they have officially started Optimus production in Fremont on their high-volume lines.
"The bots coming off this line are the first production quality bots that we have manufactured. While we will continue to iterate on the hardware and manufacturing process, we have proven that the current manufacturing process for Optimus is suitable for a commercial product."
Making cancer drugs in space because in microgravity the chemistry of crystals forming is more uniform which leads to better performing drugs.
That’s sick as hell. The future is awesome.
Let me explain why is @VardaSpace is such an important company
A patient who has cancer has cells that have proteins that are activated leading to growth, replication and then damage.
Targeted drugs work on these cells in 2 ways:
1. Inhibit the receptor (the protein), essentially jamming its function until the drug wears off. An example is Tamoxifen, used in breast cancer to block oestrogen receptor binding.
2. Degrading the receptor altogether so signalling doesn't work and the cell needs to spend time recreating the receptor before it can work again, an example being Fulvestrant which degrades the receptor itself.
The second option here is important to understand because the effect of it significantly out lasts how long the drug lasts in your system. One type of these drugs is called PROTACs - it's a newer type of cancer drug that destroys these proteins (proteolysis) removing functions often related to cancer signalling.
Now these drugs are complex, 15+ chemical steps and can be upto 4x more heavy than normal drugs. With each molecule taking many shapes, when packed together, they often don't form a uniform lattice shape. On Earth, gravity leads to convection currents that feeds these molecules unevenly into the forming crystal, too fast resulting in non-uniform crystal formation.
These drugs:
- Dissolve poorly when ingested
- Have difficulties crossing cell membranes
- Lead to unpredictable bioavailability
- Require high doses in order to get enough drug in the system
Varda here by @WillBruey and @zebulgar aims to solves this with undisturbed crystal formation growth in microgravity.
With a new purer drug you get:
- Predictable bioavailability
- A more stable drug
- A better outcome for patients
This is Varda
Game changer for chat tab hoarders.
I think this effectively 2-4x’es my usage???
Huge for scheduled tasks too, because now you can let big threads of context pile up and get retriggered after multi hour or day delays. So now your weekly planning thread with 700k preloaded context tokens is super cheap cached price.
Oh fascinating. League of legends is being reverse engineer distilled Vibecode attacked to the point they have had to change how game replays work. Model labs aren’t the only ones who need anti distilled patterns now.
Incredible public good. I’m so excited to see how health evolves for the world
ChatGPT is now the largest medical record sync from hospitals to consumers, passing Apple.
I'm glad we can all access our medical records through the apps we choose. Rightly so!
We're working with Epic and hospitals to reduce the API load on their servers. The current setup is archaic - the equivalent of refreshing your inbox instead of getting a push when email arrives.
FHIR Subscriptions can solve this & reduce strain on the system.
App layer now wins.
OpenAI, Anthropic and other leading AI labs face a growing challenge: Better models aren't necessarily attracting more usage. A 10-point intelligence advantage now earns just 20% more developer traffic, down from nearly 69% in late 2025.
Full story: thein.fo/4hGJ9jY
Peculiar situation where every company wants to vertically integrate Ai software factory stacks, but also every dev tool and model lab is competing for it
New Jev like model from Cloudflare!
Decision models are out already lol.
We got system 1 models now
Cool thing about this, unlike Jev which can only return decision probabilities, this can also extract text, computed and estimated values. Because it's not generating new text tokens, it can't hallucinate, but it still lets you pull exact words and figures out into structured json.
Super handy for asking questions like how many, or how often. A new type of smart if statement similar to Jev but for different questions.
a brand new model paradigm! introducing 𝚌𝚕𝚊𝚒𝚛𝚒𝚌: a system one model for extracting, computing, estimating, and predicting values from unstructured inputs.
blog.cloudflare.com/clairic-…
























