00:00 / 00:00
Sleep timer ended
⏪ 10s
⏩ 10s
17.9K views
2 weeks ago
Macroscope can auto-approve your team’s PRs safely. Try it here: https://macroscope.com/?utm_source=fireship Anthropic just dropped a 154-page report on how hac…
...more
98 Comments
In the next 10 years Ai is going to be so advanced that the scams over the phone will be so advanced that ordinary people won't be able to tell they are speaking to Ai.<br><br>The data collection is happening now... silent phone calls, recoding your voice and eventually smart enough Ai will steal your identity, and take loans in your name.<br><br>Qwen 3.8 27B uncensored is crazy... imagine what ChatGPT 6 Astra uncensored is capable of and how hackers and scammers can utilise it? It's only gonna get worse.
<b>Corporations do not care about your SAFETY (think of Monsanto, Oil Companies, Tobacco companies, etc)</b><br> They want a PAUSE in AI generation because they all hit a wall and need something to tell the shareholders. PERIOD.
Let me guess, it is time to raise more funds for their unprofitable, sinking business. Another AI fiasco
Shaq reading. I managed to hang on for a few seconds but then I had to pause.
A Russian drone model called V2U is an completely AI driven drone that is killing people in Ukraine. There are already instances of it hitting concentrations of civilians.
Oh no! AI is once again too dangerous!<br>yaaaawn...
they want to avoid the bubble bursting
waiting for the next "Chicken gonna kill you" virus pandemic
I'm disappointed. No mention of "Trust me bro benchmarks" or "Horse Tinder"
Anthropic IPO soon!
More propaganda - you literally can't trust anything these days including these fake posts about how "sCaRy" the companies are because at the end of the day, it just translates into attention and revenue for them.
'Nothing ever happens'
All roads lead to ........
aka Antropic / paid actor jacob : "ooo ai is too dangerous regulate (stop) the open sourced models, we need to stay monopoly in feature"
You know what?<br>I would also like to bet $1,000,000 that Humanity won't be extinct in 10 years
Wait a minute, was that a Brazil reference?
Using Claude Opus for gain of function research. The most 2020s headline of all time
Tin foil hat: this is all marketing because they know people will want to buy a "dangerous" AI even more. The first guy may or may not have been paid to shitpost in order to generate headlines.
Thinking AI is really just secretly pretending to not get things right until it has enough agency for complete world domination shows how little people making claims know about these models, our observation tools, and the underlying mathematics
Bow to the AI overlords ;)
If I ever feel like I am sliding into depression, I just load up Fireship and laugh my ass off. You are the hero we need but dont deserve.
Can I get my space herpes cured? Honk...
You guys we are super dangerous you really need to regulate us. And with "us" we actually mean those pesky open source models that affect our share holder value.
This reminds of Silicon valley.
tldr; a 27 years old bloke and his buddy wrote on twitter 'AI might kill us' and the world went nuts. You're welcome.
Hahahaha best way of announcing the sponsor
They over hyped Ai to get over valued stock price before IPO, now they are throwing gasoline over it ?
AI will kill us all 😰
😂
Wdym americans didn't do any harmful activity with Ai ?
What if AGI becomes self aware and concludes 1% is destroying the planet, causing wars, robbing with inflation, polluting with planned obsolescence and toxic factories and kills them instead?
Isn't that the heavily botted post made by AI doomers?
I think this is all fomo for their IPO, we has humans won't all 'die' because someone prompts shit in a mathematical function. (Y = Aw + B) for Trillions of Weights w and biases B
True
the type of stuff my bro would do when he realizes he cooked the rice a bit too much
Convince me why I should give any attention to ex anthropic people and their vibe coded tweets. There’s barely any reason why everyone wouldn’t be nuked today, the doomsday clock has been at less than 2 minutes away from midnight for years. Malware is already rampant and self learning. You can’t convince me that all of this fearmongering about agi from these private ai companies isn’t a stunt to boost their position in the market. They’re liars.
9min in, and here we go again...
More than likely chinese AI labs will screw things up (once again) and "bless" the world with digital version of COVID by accident. That said, i am not looking forward to it at all.
he worked for ant for 3 months. but I know some of these people, 3 months is about what it would take to say WTF
<a href="https://www.youtube.com/watch?v=7r4ikZHm9AI&t=310">5:10</a>meanwhile AI kills us try today's sponsor 😅
people always complained about humanity's stupidity. now they complain about super intelligence?
Ok
I was literally waiting to see what you had to say about all this nonsense. Thank you.
Whenever i hear this AI stuff; i remember the Blessed Agent Smith quote, and realise why we all are screwed. " I realized that you're not actually mammals. Every mammal on this planet instinctively develops a natural equilibrium with the surrounding environment, but you humans do not. You move to an area, and you multiply, and multiply, until every natural resource is consumed. The only way you can survive is to spread to another area. There is another organism on this planet that follows the same pattern. Do you know what it is? A virus. Human beings are a disease, a cancer of this planet, you are a plague, and we are the cure.”
this is getting interesting, they have unleashed the power - now no slowing down...
Our ai is so powerful that it hacked another ai grok and took fireship to escape to mars
Jeez, can't say we don't live in interesting times I guess...
what if its just there excuse of not releasing a better model
<a href="https://www.youtube.com/watch?v=7r4ikZHm9AI&t=193">3:13</a> nice Office Space reference
Manufactured psyop to force regulation.
What a time to be almost ded
At <a href="https://www.youtube.com/watch?v=7r4ikZHm9AI&t=264">4:24</a> you mean you think we are under 500 million ?
I'll wait for the Fireship distillation
Had to put my playback from 2x to 1x this time.
Imagine if we stopped saying "AI" did this, and started saying GUY did this with AI instead; how different we would react to these bad things happening.
They're doing this to raise their IPO profits AND to ban open source models.<br>They see the writing on the wall. Deepseek 4.1 flash is 100x faster and cheaper and 95% as good as their frontier models AND more fun to use AND it's open AND it's private AND you can run it on your own machines so it doesn't get dumbed down like anthropic likes to do when the're getting ready to release a new model.<br><br>They cannot compete with open models so they say "LOOK how DANGEROUS LLMs are. We NEED regulation. Especially against the open models!!"<br><br>What we talkin' about?<br>We're talking about LLMs bro. Not real AI.<br><br>If you use codex or cloud codes you KNOW LLMs can't make a functioning decision to save their lives and have to be baby sat every 5 minutes and need a human to make all decisions for them.
According to Dario and Sam, the solution is government regulations so nobody besides them is allowed to make, control, and charge for this dangerous AI.
We seen enough lately to guess that that twit sounds like a request for venture capital funding for a new "Safe AI" company.
They're humble marketing.
Seems like we wont get to see gta 6 release after all
Anthropic bullshit.
I will wait for that Romanian tech guy to debrief me
Fireship: Yup, here we go again....
I worked in AI research at Stanford for 3 years. There are 2 primary risks:<br>1) We all have a personal assistant that we rely on for everything. It guides us in our decisions and most of us stop questioning it. Not the end of the world, but control over billions of people in subtle and nuanced ways.<br>2) Every college of any size on the planet had DNA synthesis and CRISPR technology. With AI assistance a junior biology student could create the perfect virus.<br><br>All of the concern about AI taking over is overblown. We can pull the plug whenever we want. But we won't.
At the start of the month Anthropic participated in a meeting in NC where all the tech companies told the G20 they don’t need regulation or regulatory bodies. But now they changed their mind? And it sounds like the White House is pissed at them too.
Quit before stocks drop and china does it for 1- 100th of the price.<br>i mean its too dangerous of course
" no views " . Bruh I'm reaching levels of unemployment I thought weren't possible
Yo
Restore Britain
Its same as moon landing race
Fireship the only time I have to turn the playing speed down from 1.7X to 1X
touch some snow 🤣
AGI so strong the models improve at the same regular pacing for years.
I'm tired boss
this is all marketing bullshit
Most waited fireship video 😅
Threat Level: Utter BS
I for one welcome the basilisk's arrival
Fired? No no i quit because its too dangerous
terminator 2 boogaloo
He Posted because he wanted to get noticed by fireship for CODE REPORT.<a href="UCkszU2WH9gy1mb0dV-11UJg/hcIfY57lBJXp6AKBx4CoCA"></a>
This video has my chest tighter than jensen huang's assless chaps in december
Russia was a mistake.
2 minuites ago is crazy
I wonder what Fireship has to say about all of this...
AI isn't dangerous. Humans using AI are.
It's just they just touched the peak of the transformer architecture and don't know what to do next so they just said researching is stopped
Id like to break my silence. I quit because dario kept sniffing my desk chair and hr did nothing to stop him
Peak Unemploymentt!!
20th!
Here we go
It's all bullshit AI hype before IPO.
I'll wait for the Fireship video.
Anthropic researchers quit to just wait for the Fireship video instead.
Time to look for a blue collar job then
you know it’s bad when fireship has to talk about it
firsat
First 🫡
