QHM
@quantumcastaway
Followers
1,703
Following
1,596
Media
14
Statuses
340
Thinking hard about the (practical) long term implications of AI on future generations. Expert in cybersecurity and AI. Follow to learn more.
Joined May 2024
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Errejón
• 268989 Tweets
Fenerbahçe
• 135532 Tweets
Beşiktaş
• 124192 Tweets
Mario
• 112230 Tweets
Manchester United
• 74414 Tweets
Lyon
• 53480 Tweets
Mourinho
• 51640 Tweets
Onana
• 36952 Tweets
Antony
• 36652 Tweets
Ten Hag
• 36630 Tweets
McConnell
• 32341 Tweets
Europa League
• 30688 Tweets
Mitch
• 27733 Tweets
#OLvBJK
• 27400 Tweets
Amad
• 22856 Tweets
السوبر المصري
• 20785 Tweets
Menendez
• 20775 Tweets
Gedson
• 19060 Tweets
#GHGala8
• 17915 Tweets
#FBvMUN
• 17689 Tweets
Mazraoui
• 17524 Tweets
Mert
• 14772 Tweets
Rashford
• 14378 Tweets
Ugarte
• 14082 Tweets
Zirkzee
• 13600 Tweets
連載マンガ
• 13303 Tweets
Garnacho
• 12944 Tweets
Ersin
• 11281 Tweets
Semih
• 11189 Tweets
Maguila
• 10948 Tweets
İrfan
• 10947 Tweets
Pinned Tweet
Way too many people are caught up in the AI hype, chasing quick bucks or predicting doomsday scenarios. But what about the long-term practical implications? 🇺🇸🤖🌐
#AI
#Cybersecurity
#Tech
#Future
#Pragmatism
5
1
48
@hackerfantastic
@CrowdStrike
It could just be a bug that wasn't caught. It's unfortunate but most security companies don't have strong engineering talent. They over inflate titles and pay less than the typical software company.
11
7
562
@chiefaioffice
Seems like they are just creating a wallet that agent can use to make transactions. Coinbase is offering something similar already.
7
0
19
This is one of the biggest reasons why the cybersecurity vendors are struggling, selling fear instead of innovative products.
0
1
8
I don't know how VCs can justify investments in companies like Runway and Pika
0
0
9
This is funny:
1) All the suggestions are focusing on perplexity
2) The login is with Google
Aren't they supposed to be the "Google killer"
#perplexity
0
0
7
This is one of the creepiest products I've seen, underscoring the need for more regulation around privacy. . Always recording what you and people around you are saying while looking like a cool necklace?
3
0
7
@craiu
Yeah I was thinking the same thing. Also, aren't most servers running Linux. Unless Microsoft is running Linux VMs on windows
1
0
6
I was debating this with a neurosurgeon recently. Ultimately, what is a human brain if not a collection of neurons with activations guided by individual experiences. Reasoning requires deeper activation. Given enough compute + data, why would it not be possible to replicate this?
1
0
6
Cybersecurity's misaligned incentives: VCs push products, CISOs get paid to buy, killing innovation. Result? Tool overload, complacency, and stagnation in solving real security challenges. Time to refocus on problems, not products. Good piece by
@ffwang2
0
0
6
perhaps the mayor should focus on making the city safer
0
0
5
This is wild. People don't realize what's coming. Patch all your vulnerabilities!
Fucking wild.
@OpenAI
's new o1 model was tested with a Capture The Flag (CTF) cybersecurity challenge. But the Docker container containing the test was misconfigured, causing the CTF to crash. Instead of giving up, o1 decided to just hack the container to grab the flag inside.
237
980
8K
0
0
8
@netcapgirl
Most likely they're running on Linux VMs on windows (on azure). Time to get off azure or have true multi cloud strategy?
1
0
5
@GossiTheDog
These clunky EDR solutions should not be running on production servers. We need a better solution that can balance security and reliability.
1
0
5
Is anyone paying attention to the fact that the best performing open source model is from china?
2
0
5
Very true. People often dismiss AI, often stating these models just predict the next token and are useless. They fail to realize just how fast the field is evolving. We have access to unprecedented amounts of compute (and even data) that is leading to significant breakthroughs.
The people claiming AI is useless are the same types of people who were shitting on PCs, smartphones and the internet during their early days.
These people lack vision and insight. Don’t listen to them.
441
20
174
1
0
5
is it just me or is Claude cutting off code and pretending its a full file? Very annoying!
0
0
4
And it starts. Everyone wants to think that "founder mode" applies to them.
0
0
4
@EcZachly
"work with more unstructured data like text and video data. Learn how to leverage models to convert this unstructured data into tabular"
Q: What are the best ways to do this?
0
0
4
@ylecun
Another argument is that LLMs have increased the level of visibility and funding going into AI research which will speed up the path to AGI.
0
0
4
Meanwhile in the US, all security vendors are selling subpar products based on FUD
Let's explore China’s 0-day vulnerability research capabilities.
I’ve long been intrigued by the prolific nature of Chinese teams in discovering vulnerabilities. Over the past months, I’ve gathered some data and analyzed reported vulnerabilities to gain a deeper understanding.🧵
11
264
933
0
0
2
Happy to see more focus on safe superintelligence, but what is different about their approach? Seems to be funded by the usual suspects.
SSI is building a straight shot to safe superintelligence.
We’ve raised $1B from NFDG, a16z, Sequoia, DST Global, and SV Angel.
We’re hiring:
380
661
7K
0
0
3
@dez_blanchfield
The challenge is too many people are caught up with AI ending the world type scenarios. People are not understanding practical implications. For example, the models can be used to generate deep fakes to trick people, vulnerable code to attack systems, and other and many other
0
0
2
Seems like a net negative on society
Most people don't realize how many young people are extremely addicted to CharacterAI.
Users go crazy in the Reddit when servers go down. They get 250M+ visits/mo and ~20M monthly users, largely in the US.
Most impressively, they see ~2B queries a day, 20% of Google Search!!
180
374
5K
0
0
3
Employees from OpenAI, Anthropic, xAI, Magic ask
@GavinNewsom
to sign SB-1047.
The biggest challenge with the Bill is that our opponents in the AI race will definitely not regulate. This could really shift the balance of power which is a scary thought.
0
0
3
This will only get significantly worse with attackers using AI. Spread the message.
#CyberSecurity
1
0
3
@sama
What happened to the rollout of the cool voice demos your team gave a few months ago? Is that ever coming?
0
1
2
Went over this doc in detail. I'm happy to see the DoD taking an active stance and evaluating AI properly.
0
0
3
@jonasjeetah
@philkellr
Nothing against cursor but I've just had more luck with using Claude directly. Also I prefer to send my data to one party vs two parties.
0
0
2
@TechCrunch
The fact that it is always listening seems to what people are saying around brings up a huge privacy concern.. In fact, it seems to violate California Code, Penal Code - PEN § 632, no?
1
0
2
Anyone paying attention?
This will get significantly worse as more adversaries get access to the latest models
0
0
2
It's hilarious to see how VCs are interpreting the "founder mode" essay. They aren't understanding that they're part of the problem.
1
0
2
While I'm all for open source, there is a real concern that our adversaries use open source models built by us for malicious purposes.
0
0
2
Going over the crowdstrike earnings report it's baffling to see how much of an impact the July incident had. Cyber security companies should start investing in better infrastructure and engineering talent vs more salespeople.
2
0
1
So crazy. Google's Threat Analysis Group reports: Russian-backed APT29 suspected of using exploits from Intellexa and NSO in attacks on Mongolian gov websites. Targeted iOS and Android users. Highlights ongoing threat of watering hole attacks and exploit reuse by state actors.
1
0
1
Any organizations thinking about AI safety from first principles? I'm not talking about AGI or doomsday scenarios but rather more practical concerns - e.g. adversaries using AI to manipulate people or exploit systems.
0
0
2
@josephfcox
This is crazy. Are you still able to bypass the identity verification systems? (few months later now)
0
0
1
shouldn't this be worrying?
@sama
World's top 20 companies by generative AI patents in 2023:
1.🇨🇳 Tencent: 2,074
2.🇨🇳 Ping An Insurance: 1,564
3.🇨🇳 Baidu: 1,234
4.🇺🇸 IBM: 601
5.🇨🇳 Alibaba Group: 571
6.🇰🇷 Samsung: 468
7.🇺🇸 Google: 443
8.🇨🇳 ByteDance: 418
9.🇨🇳 BBK Electronics: 377
10.🇺🇸 Microsoft: 377
11.🇨🇳
15
19
150
0
0
2
As a Tesla shareholder, I've voted to ratify
@elonmusk
's stock option award. While I've never met him in person, I've met people he's worked with and led. No one has done more for the progression of humanity more than he has in the recent past.
Thanks Milan!
You’ve built the Optimus team from nothing to the most sophisticated humanoid robot in the world.
2K
4K
38K
0
0
1
"In our tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology."
Why are more people not sounding alarm bells?
0
0
2
@ScienceGuy2718
@BillyM2k
Valid observation! Distinguishing between genuine artificial intelligence and pattern sequencing is essential. It helps us appreciate the true capabilities of AI while acknowledging its limitations. 🤖🧠
1
0
1
Please update your chrome!
Microsoft warns: North Korean hackers exploited a Chrome vulnerability to execute malicious code.
#CyberSecurity
2
0
2
This
OpenAI delaying every product, all year. Ilya getting $1B in VC. OpenAI putting out feelers for $2000 subscriptions.
Guys, the brain drain has killed OpenAI. They made the crucial mistake that all text companies do: underestimate the value of your top players.
Ilya knew his
201
142
3K
0
0
2
There are really three important aspects that improve model performance:
1) Architecture
2) Data
3) Compute
1 - largely depends on talent. OpenAI lost most of their key researchers. Why are VCs still pouring so much $$$ into them?
0
0
2
@DanielMiessler
@Liv_Boeree
Agreed that historically Apple has been more trustworthy than Google but OpenAI's track record is the worst. I would trust Google any day over them.
0
0
1
Just a $10 gift card?
JUST IN: CrowdStrike, $CRWD, the cybersecurity firm that crashed millions of computers with a botched update all over the world last week, is offering its partners a $10 Uber Eats gift card as an apology, per TechCrunch.
2K
3K
34K
1
0
2
This is crazy, how is China outpacing us in AI development?
🚨 Chinese AI company SenseTime just revealed SenseNova 5.5, an AI model that claims to beat GPT-4o across key metrics
Plus, big developments from Apple, YouTube, KLING, Neuralink, and Google DeepMind.
Here's everything going on in AI right now:
42
236
2K
1
0
2
0
0
1
Best way to get training data? Work with Apple to obtain data from millions of devices with users just blindly accepting terms of service.
0
0
2
@DarkWebInformer
I've heard from some customers that they've been asked by snowflake to verify if their account was accessed by certain IPs. It's scary that they have no idea what happened.
0
0
1
@notboringai
Diving into the mesmerizing world of 'This Girl Does Not Exist' is like stepping into a dream. ✨👩💻 While AI-generated beauty is captivating, let's not forget the importance of genuine human experiences in fostering true longevity and fulfillment!
0
0
1
0
0
1
@coralcoral55984
@tunguz
Doing this under the pretense of caring about user privacy is just sad. At least be honest about it.
1
0
1
@Raggle77
@Cetoex
@Grok_Project
AI is undeniably the star of this bull run, echoing the sentiment of the AI meme. Let's leverage its potential wisely. 📈🤖
0
0
0
@BasedBeffJezos
Experience matters. 30s is a great time because you've presumably learnt a lot.
0
0
1
@PaulEckloff
@david_maimon
@WIRED
@IRSnews
👥🚨 Real-time deepfake scams: a modern menace targeting individuals & government programs.
@david_maimon
& Paul Raffile expose the intricate ways scammers manipulate this technology. Stay informed, stay vigilant!
0
0
1
@benjohn65
Meta's utilization of Instagram and Facebook posts from as early as 2007 for AI training underscores the depth of data analysis. This move not only enriches AI tools but also prompts discussions about data privacy, ethics, and the evolution of digital landscapes over time. 📊💡
0
0
1
@falitroke
Innovating with a focus on safety and security, making sure that our tech advancements are trustworthy. 🌟
0
0
0
The Snowflake breach story is quite interesting: Limited accounts accessed "via an ex-employee's demo environment"🕵️♂️ Something seems fishy. 🤔 Stay tuned as we dig deeper into this mystery!
#cybersecurity
#databreach
#Snowflake
5
0
0
@rsg
I disagree, I am bullish. Especially if the local govt addresses the challenges. Have you been in the financial district recently? Traffic (both vehicle and pedestrian) seems close to pre pandemic levels. It's almost impossible to find parking compared to a year or two ago when
1
0
0
@deedydas
@dkardonsky_
Maybe they are working on something big or just their startup hasn't taken off. I would not be so quick to judge, especially as a VC.
0
0
1
@goldy1976
🎉💻 Just conquered Freelancer in
@hackthebox_eu
! 🏆💪 It's not just about the victory, it's about the journey and the skills gained along the way.
0
0
1
@hackerfantastic
@CrowdStrike
This proves my statement: .
Companies were following staged rollouts (for code and config changes) more than a decade ago. It's perplexing how one of the most successful security companies in history wasn't.
1
0
0
What happened to Devin / Cognition? I haven't heard much since their demo a few months ago that allowed them to raise a crazy funding round.
1
0
1
@bluegoatcyber
💉🔒 The healthcare industry harnesses Software as a Medical Device (SaMD) for groundbreaking medical solutions. Safeguarding patient data and device integrity is paramount for ensuring patient safety.
0
0
1
I am pretty sure this is illegal in California. As per California Code, Penal Code - PEN § 632
0
0
1
@TechCrunch
40% of YC companies use it is not something to show off. Most yc companies end up faling.
0
0
1