publidave Profile Banner
Publidave Profile
Publidave

@publidave

Followers
153
Following
121
Media
84
Statuses
1K

0.5X engineer, interested in many things and least bad at infosec

UK
Joined December 2012
Don't wanna be here? Send us removal request.
@publidave
Publidave
2 years
The DJ/tie combination certainly is a vibe @dieworkwear
@claudcockerell
Claudia Cockerell
2 years
Farage teased a 'big announcement' in the coming weeks, but said he didn't want to make it on his birthday: 'Why spoil a lovely booze up?'
0
0
0
@publidave
Publidave
2 years
0
0
0
@publidave
Publidave
2 years
This is a smart move by MS: they used OpenAIs lead as the engine get ahead in the product space, now helping to create open source competitors so they aren’t married to openAI. Willing to bet MS R&D folk who have played with Gpt can do fun things with good open source models
@alex_valaitis
Alex Valaitis
2 years
🚨BREAKING: Zuck just announced that Meta will be open sourcing Llama 2 with Microsoft. This is a massive announcement for a few reasons... 1) This could kill a number of the open source LLM startups. Mosaic, Red Pajama, etc. are in major trouble. The $1.2 billion @MosaicML
0
0
1
@publidave
Publidave
2 years
Llama2 looks interesting, but starting to feel chat will not be the space LLMs compete in - opportunities to be the go to model for specific tasks: summarisation, categorisation, option creation etc. gimme a 7bn model that can summarise but can’t hold a conversation
0
0
0
@publidave
Publidave
2 years
steel made prior to atomic blasts is highly valuable, will datasets made prior to LLM emergence become similarly so
1
0
2
@publidave
Publidave
3 years
In sentiment analysis, #chatGPT 4 can spot sarcasm while 3.5 couldn’t and a general sentiment analyser was fooled by it
0
0
1
@publidave
Publidave
3 years
While asking about something else, ChatGPT just casually dropped in that in WSL you can do: ‘Linux command blah blah’ | clip.exe and it copies the output of the command to windows clipboard 🤯
1
0
3
@publidave
Publidave
3 years
I’m torn as this video on one hand is exciting as a techie, on the other hand I have a dread it’s going destroy human communication. Business will just be ai’s exchanging bullshit to each other while Meet deepfakes me so I look like I’m not hungover in bed
@frantzfries
Chris Frantz
3 years
Google hits back with a full integration of AI across Google Workspace. Includes generating images, presentations, emails, docs and more all within Google Workspace. Video overview is slick
0
0
1
@publidave
Publidave
3 years
‘Juan of the dead’ might be the most underrated #zombie movie ever
0
0
0
@publidave
Publidave
3 years
This is probably fine
@emollick
Ethan Mollick
3 years
AI has been destroying humans at adversarial games, like Chess or Go, for a bit. But now it is successfully outperforming humans at Diplomacy, the classic multiplayer game that requires natural language negotiations over chat, as well as strategic lying. https://t.co/e6V7pHQnkO
1
1
4
@publidave
Publidave
4 years
This is a great point that I wish more pentesters were aware of
@kimmydotzip
𝚔𝚒𝚖𝚖𝚢.𝚣𝚒𝚙 🅅 👊🇺🇸🔥
4 years
@AlyssaM_InfoSec When I ran appsec at a bank, our 3rd party mobile app pen test (well-known company) had "screen shots are allowed" as a finding. It was either "fix" it or deal with a policy exception with many approvals and annual review. We disabled it.
0
0
1
@publidave
Publidave
4 years
What’s impressive is this isn’t even the first time this has happened to Mazdas,
99percentinvisible.org
It’s the crossover event you’ve all been waiting for: Reply All‘s Super Tech Support takes on an annoyingly specific technology problem involving 99% Invisible. Ben loves podcasts, but he has a...
@mattblaze
matt blaze
4 years
This appears to have been triggered by a file naming error for an image (as in a picture, not a disk image). But FM subcarriers are also used for other things, including real-time traffic and wx. I wonder what intentionally malicious data could do... https://t.co/gKMCcDeu4L
0
0
0
@publidave
Publidave
4 years
Little surprised @Etsy are paying to promote this sort of thing @evacide
2
6
49
@nils
nils
4 years
Twitter: NFT profile pics me: document.querySelectorAll("div[style*='hex-hw']").forEach((e)=>{e.parentNode.parentNode.parentNode.parentNode.parentNode.parentNode.parentNode.parentNode.remove()})
0
3
11
@publidave
Publidave
4 years
This article is written for engineers but applies well to securitists, definitely seen security teams fall into the traps it describes and get left out strategy discussions
Tweet card summary image
reforge.com
To make more successful technical calls & advance careers, engineers actually need to develop better strategic decision making skills.
0
0
0
@publidave
Publidave
4 years
New paper on phishing training effectiveness: arguably made employees more vulnerable, but reporting buttons work https://t.co/ubt5ne5BY8
0
0
2
@publidave
Publidave
4 years
@citizenlab
The Citizen Lab
4 years
NEW REPORT Pegasus vs. Predator: Dissident’s Doubly-Infected iPhone Reveals Cytrox Mercenary Spyware https://t.co/58G8tpWXkG
0
0
0
@publidave
Publidave
4 years
This PhD thesis on satellite security is a great read. Come for the plaintext, stay for the study on previous incidents https://t.co/jT19gQwLFM
0
0
0
@publidave
Publidave
4 years
Takes a few mins to get going but there’s some really good advice in here on where to look for bugs regardless of target. Great interview
@dragosr
dragosr
4 years
Exploit Hall of Fame - James Forshaw (@tiraniddo) & Steampunk Cyberspider Tour for Lighting Talks Dec 10
0
0
0
@nojonesuk
Nick Jones
4 years
This eBook on #attackdetection in the #cloud is built on the work that @ajpc500 and I have done over the last few years in that space, and distills a lot of the content from our previous conference talks into something a little easier to digest and reference. Hope it’s helpful!
@FSecure_Consult
F-Secure Consulting
4 years
Learn how to develop and measure #attackdetection efficacy in the #cloud. Our latest eBook from @ajpc500 and @nojonesuk explains the 5-phase process developed by our team over a 2-year period >> https://t.co/Ozla68li9H #Clouddetection #InfoSecurity #CyberSecurity
1
5
19