How well do the security community's techniques hold up against AI-enabled cyberattacks?
We examined 832 malicious accounts and mapped their activity onto a longstanding database of tactics and techniques used by threat actors.
Here's what we learned:https://t.co/fgOqJRh2rx
Unfortunately, NONE of the coding models is reliable. At least not reliable enough to build stuff and trust that there isn't a huge bug or vulnerability hiding in there somewhere. I agree some are stronger at coding and Gemini can write incomplete code and skip stuff but it's not as bad as it's being presented all over.
My recommendations - mix the top models.
1. Use Gemini to discuss and prepare your requirements and then Claude or GPT to review and update.
2. Use Gemini to build and use Claude or GPT to review and fix bugs.
Also, consider using their harnesses.
"Wireless Fidelity" has been known for decades. There has been too many attempts to rewrite this history, though I have no idea to what end. A lot of history has been scrubbed or modified but the coiner definitely mimicked "High Fidelity (Hi-Fi)" to spawn Wireless Fidelity.
I recall the actual objective was to suggest 802.11b was a high fidelity wireless standard hence Wi-Fi. Old school ITC folks know this.
@AkanAseserm@weloveghana042 I guess you don't know there are drains along the main roads to collect water off the road. Why you don't know this is because they are covered with gaps for flowing water
@AkanAseserm@weloveghana042 The problem is not the problem, the real problem is the people who are ignorant about solving the problem but will come and argue because they have data and access to an Internet enabled device. At least use it to do some research and self enlightening
@jamessturge@ccsakuweb Let me know if you face any challenges. Remember you must sign up to opencode to get access to their models including the free ones. I found Mimo-V2.5 and Minimax-M3 to be fantastic. They are free at the moment.
@jamessturge@ccsakuweb find the path to \chatLanguageModels.json (or ask the AI if it knows where it's located) then find the opencode docs url (https://t.co/RiSjO5EWJ3). Then prompt your current Github Copilot AI inside VS Code to "add all the models opencode models to your settings.
Google Play Console Verification is the weirdest verification experience I have ever come across. Almost feels like a scam.
You first pay before they request verification, they then reject the ID verification, and finally state you can only retry an unspecified number of times.
What I expected:
1. If you are a Google ecosystem user for years, that should count.
2. Driver's license, Passport and National ID verification should be automated. The tech already exists and works fine.
3. Don't ask people to pay before you request verification. Make sure all is sorted before demanding payment.
4. Explicitly state what exactly went wrong with their verification process so that they can fix it.
5. Don't limit how many times they can try otherwise don't take their money first.
6. Refund if they absolutely fail verification.
7. Prompt support. Maybe AI driven for non-critical actions e.g. Analyze verification details before they are submitted to save people the headache and time wasting.
@GooglePlay@Google
Does anyone ever realize that if you got locked out of your gmail. You'd essentially not be able to function.
And over time all our accounts are interlinked in a way which means it's all a house of cards.