تدريب Shadowing: He’s always wrong - تعلم التحدث بالإنجليزية عبر الفيديو

All right?

We're going to do this again.

I hope you're ready for this, okay?

Because since June 12th, we've been working closely with the U .S government to restore access to Claude Mythos 5

and Fable 5.

I can't believe Anthropic lost it.

I wonder, I wonder what caused that.

Today, the government has notified us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of U .S organizations that operate and defend critical infrastructure.

Now you're probably thinking, well, wait a second organizations that do critical infrastructure.

Haven't we seen this before?

Wasn't there something about mythos and only going out to select organizations?

Yes, there was.

It was called Project Glasswing.

This was just a couple months ago when Anthropics said mythos was in fact too dangerous.

And then they released it to everybody.

And then they pulled it back because the US government was told how dangerous it was.

Well, we're just replaying this entire record again.

We're back to Project Glasswing.

There is the set of the good guys, which apparently is

they get access again

and the rest of us we don't get access now i'm going to talk a lot about safety okay

because here's the deal i think this is all fake i

think not only is it fake i think it's actually hurting

people i actually have the opposite opinion of dario surprise surprise

me with the opposite opinion me with the opposite opinion of

this guy right here okay we don't actually see eye to

eye by the way this is dario in 2023 oh we're gonna play this clip of him up.

Of course, before we begin, I'd like to say thank you to the sponsors.

One of the more terrifying things you can do as a company is to contract out software.

And I know a lot of companies are contracting out their mobile applications.

The thing that makes this so dangerous is that it's a black box experience.

You hope you get back software that is complete, that is going to be well tested, that ultimately will be extensible as future needs require.

The last thing I wanted someone else to say to Claude, hey, mobile app, no mistakes.

It does make mistakes.

This is why I trust the guys over at Infinite Red.

I have had many a long talks with the CEO about software quality

and the way to really build robust and good systems.

And I love everything he has to say.

And everything that I have seen with my own eyes, quality of software that they build is very, very good.

All right, so the safety argument is actually pretty dang ridiculous that we're having right now, this whole kind of gated access to Claude mythos.

But I just wanted to kind of add one more quick point.

If you don't know, right now, chat GPT 5 .6 is also in a limited preview.

And the reason being, for now, at the request of the US government,

we're starting with a limited preview among a small group of trusted partners that's also too dangerous.

He's not allowed to release it either.

We got to make sure we slowly roll the thing out.

And we're entering into an era in

which the government is now going to have a larger and larger impact into what is exactly happening.

And this was always Dario's goal.

His goal was always to have the government step in and make regulation.

He's been just saying this nonstop.

If you read his writings, if you listen to his talks,

and his urging that we all has been realized the Frontier Studios are all being slowed down,

but that's not really actually making anything safer.

Because here's the deal.

Have you heard of the old model GLM?

Well, GLM 5 .2 apparently beats Claude in some cybersecurity benchmarks.

At least in this article, the security that's being ran is insecure direct object reference.

If you don't know what that is, that's simply, let's just pretend you have an endpoint on your website that goes and grabs user information.

If I were to pass in any user ID, you should probably go, hey, are you that user?

Oh, you're not?

Hey, you're not allowed to look at, I'm sorry.

Hey, brother, sorry, this is not you.

404, don't know who this is.

out that the fourth most common bug on hacker one top vulnerability list is in fact,

these right here these oopsie poopsies insecure direct object references where people can just go onto websites

and get data if they just simply have an ID.

So Semgrep went off and developed a benchmark against these type of vulnerabilities.

And it turns out things like Claude code Opus 46 and 48 did about a 28 % to 37%.

Jippity codex at about a 20 %

itself did all 39 % did better than both Claude Code and their latest models and Gipity 5 .5.

Now

if you don't know a harness makes a big difference a

harness is a lot of the reasons why you think coding with an LLM has gotten

so good it's actually the harness the ability for the model to use

and abuse the system to gather information

and make smarter decisions a harness is going to be the thing

that gives the model access to tools in which it can then go and make much more informed decisions so

with no harness outperformed Opus 4 .8 by practically 10%, Jippity 5 .5 by practically 20%.

And those both have harnesses.

And the thing about GLM that you may not realize is that it's an open weight model.

I can go to Hugging Face right now and download the model.

The thing is, it's very, very large for me to be able to run it with any level of effectiveness, it could cost me about $73 an hour.

Now, if I was a lot more than $73 an hour by scamming people.

And of course, it would be a completely offline model for me to just have, I wouldn't have to go

and fork over millions of dollars to be able to have an on premise machine that could run this instead, I could just run it from the safety of my computer, just doing a quick little SSH up,

bada bing, bada boom, and just see the thing run.

Not only that, but Hermes agents, one of these many agents, one of these many harnesses that run models has something called a mixture of agents, meaning it uses both 55 from Jippity,

and also uses 40 is four eight.

And by combining those both, it's doing better on benchmarks.

In other words, it's performing better.

I don't really like honestly, when someone says it's 11 % better, I have no idea what that actually means, because I don't know how to quantify.

programming in

that kind of sense i don't know how to be like it's four percent better like i don't know what

that means But it's just saying, hey, it's scoring higher on whatever these tests are, which once again means that the ability to do,

you know, security research and all this is actually kind of generally available.

The bugs that Mythos has found, people are able to find them also with 4 .8.

You're going to be able to find them with the mixture of agents.

GLM is going to be able to go off and find these as well.

100

As harnesses improve, the ability to do this is going to go way up.

101

higher than Mythos.

102

So it's all actually available.

103

It's just available to those with money.

104

If you have money and time, and you can set up your own harness, bada bing, bada boom, you're going to have yourself a hacking machine,

105

which is ultimately what Dario has been warning everybody about.

106

The scaling of open source models, I think it's going down a very dangerous path.

107

And if again, if the path continues, I think we could get to a very dangerous place.

108

Yikes.

109

I think it's worth saying some things on open source models that are clear to all the experts, but I want to make sure is understood by this.

110

Clear to all the experts.

111

They uniformly agree, which is crazy because, you know, classic experts.

112

I mean, you can't even get dentists to agree on the toothpaste, and he's over here getting...

113

experts to agree to outcomes of stuff we haven't measured yet?

114

Kind of wild.

115

Committee, which is when you control a model and you're deploying it, you have the ability to monitor its usage.

116

It's really all you need to know.

117

You'd like to monitor the usage.

118

And this is the whole argument.

119

This is the entire safety argument, which is just completely broken.

120

There's open source models in which are doing absolutely excellent.

121

There's people that have money and motivation to scam.

122

They have access to it.

123

There's a whole bunch of people that maintain don't have access to these models.

124

They don't have access to these open source frameworks, because they're not going to go spend, you know, $5 ,000 over a weekend to maintain their little open source project

125

that nobody actually cares about in which the entire world uses

126

and everybody actually cares about they just don't realize they care about it.

127

And this is the whole problem of the safety argument is that it's fundamentally unsafe, the tools and the harnesses are there.

128

people talking about it all the time on the internet.

129

There's already better harnesses that are scoring higher, whatever that means, on the same benchmarks than Mythos.

130

At the end of the day, I think Michael, creator of Effects TS, says it best.

131

Gonna need to get something off my chest.

132

The Discord on CyberSec and Mythos slash US approach is stupid beyond imagination.

133

There should be no guardrails, and the most powerful model should be immediately available to everyone.

134

Want to make sure the world is safer?

135

Give free usage to maintainers of core infrastructure.

136

If access is restricted, be sure offensive teams will find their way in.

137

Somebody with insider access to Anthropic

138

or OpenAI will be offered a deal they can't refuse and leak the model to malicious actors, or a malicious actor will end up developing an equally capable model.

139

Look at GLM.

140

It's near the same level of capability.

141

Yeah, maybe it's not mythos levels, but it's not far off.

142

Yes, not only the good folks have money, but also the bad ones.

143

Exactly correct.

144

The only thing you achieve by restricting access is unsafety.

145

100%.

146

Because that's the thing is right now, you could have an entire world that exists in which people could be preparing their piece of the world,

147

their little corner of software, making it more secure if that's the true delivery of these models to begin with.

148

If that's the case, wouldn't you want the whole world to walk in lockstep?

149

Why would you want a select group of bad guys who happen to have a couple million

150

or the ability to rent out enough Blackwell Ultra

151

Nvidia to be able to take advantage of everybody else

152

while we the plebs wait for the ability to secure our own systems.

153

Like the fundamental premise of this all doesn't make any sense.

154

Like if Dario actually wanted safety, he would go for Federation.

155

It's the only way to make the world safe is to

156

have a lot of people work on it at the same time.

157

The only way to make it unsafe is to allow people to lag far behind what the bad guys have access to.

158

Ridiculous.

159

Absolutely ridiculous.

160

And that's because at the end of the day, anthropic Dario over here what they really want is control he says it right there inside the little interview

161

If I can just control the inputs, well then I can control the outputs.

162

If I don't let you say certain things, well then you can't say it to the model.

163

I can protect you from you and ultimately that's why i call him daddy dario

164

He is the one that's going to keep you safe from all the bad guys or good guys.

165

But you're going to be safe because, honestly, he has your best interest in heart, which is just control.

166

I mean, sorry, safety, safety.

167

He has your best interest in heart.

168

It's safety.

169

That's what he wants.

170

Anyway, so there you go.

171

That's my whole opinion on safety in this entire talk.

172

I think it's all ridiculous.

173

I think this entire gated access and everything is really not doing anything good.

174

I think at the end of the day, the bad actors are still going to be bad.

175

been cited a hundred times all over these different studies it's not like we're actually really having security

176

or just having security theater.

177

And ultimately, it's a play for control.

178

Or it's a play for Dario to be able to regulate his opponents out of business.

179

I'm not actually really sure which one it is.

180

Probably both is my guess.

181

Anyways, I can't wait for Fable to come back out again.

182

That would be great, huh?

183

I'd love to have a little bit of Fable.

184

Wouldn't you?

185

Yeah, that'd be fantastic.

186

I'd use it again.

187

Yeah, Fable?

188

I would.

189

The name.

190

This is the Primogen.

📺 نفس القناة

✨ فيديو موصى به

33 جمل

THEY'RE HIDING THE TRUTH ABOUT AI

Business EnglishDaily ConversationUniversity Students

349 جمل

THE DISCLAIMER: Let It Breathe

Business EnglishDaily ConversationUniversity Students

105 جمل

Gilfoyle Hacks Jian Yang's Smart Fridge 🤓 Silicon Valley

Business EnglishDaily ConversationUniversity Students

567 جمل

I Cheated Until the Casino Banned Me

Business EnglishDaily ConversationUniversity Students

208 جمل

EnglishPod (B0007)

Business EnglishDaily ConversationHigh School Students

194 جمل

『プラダを着た悪魔２』無料プレビュー

Business EnglishDaily ConversationUniversity Students

205 جمل

The Man Who Stole MySpace | Samy

Business EnglishUniversity StudentsWorking Adults

194

22 جمل

Quick English Shadowing Practice

Business EnglishDaily ConversationUniversity Students

775 جمل

117. Coldplay Cheating Scandal, Proposal Gone Wrong, + LIB Season 9 Announcement

Business EnglishDaily ConversationUniversity Students

142 جمل

The Internet Didn't Fail. It Was Taken.

Business EnglishDaily ConversationUniversity Students

69 جمل

7 Programming myths that waste your time

Business EnglishUniversity StudentsWorking Adults

266 جمل

Computer Repair Shop Workers Confess Their SINS

Business EnglishDaily ConversationUniversity Students

ما هي تقنية التظليل الصوتي؟

التظليل الصوتي (Shadowing) تقنية تعلم لغة مدعومة علمياً، طُورت أصلاً لتدريب المترجمين الفوريين المحترفين. الطريقة بسيطة لكنها قوية: تستمع لصوت إنجليزي أصلي وتكرره فوراً بصوت عالٍ — كظل يتبع المتحدث بتأخير 1-2 ثانية. تُظهر الأبحاث تحسناً كبيراً في دقة النطق والتنغيم والإيقاع وربط الأصوات والاستماع والطلاقة.

تقنية الشادوينغ: اقرأ الدليل الكامل خطوة بخطوة →

تدريب Shadowing: He’s always wrong - تعلم التحدث بالإنجليزية عبر الفيديو

You won't fall behind

Go’s Big Mistake

People Are Mad They're Told to Learn

Maybe we were wrong

Github are you joking?

THEY'RE HIDING THE TRUTH ABOUT AI

THE DISCLAIMER: Let It Breathe

Gilfoyle Hacks Jian Yang's Smart Fridge 🤓 Silicon Valley

I Cheated Until the Casino Banned Me

EnglishPod (B0007)

『プラダを着た悪魔２』無料プレビュー

The Man Who Stole MySpace | Samy

Quick English Shadowing Practice

117. Coldplay Cheating Scandal, Proposal Gone Wrong, + LIB Season 9 Announcement

The Internet Didn't Fail. It Was Taken.

7 Programming myths that waste your time

Computer Repair Shop Workers Confess Their SINS

حول هذا الدرس

المفردات والعبارات الرئيسية

نصائح الممارسة

ما هي تقنية التظليل الصوتي؟