Anthropic Founders Share Roadmap to Advance AI

[@raydosson2025]

That audio buzz is killing me. Maybe use an audio tool to take it out next time for best results.

[@icepeakhd]

Seriously, why is this a thing in 2024 for like 60% of the AI talks uploaded on KZbin? They could at least fix it in post.

[@brandonreed09]

I can't unhear it now 😅

[@raydosson2025]

@@brandonreed09 my bad 😆

[@sophiedelavelle5958]

Came here to say that! You have the opportunity to have amazing speakers but a GND issue for mics ?

[@strikeforcealpha9343]

Yeah seems a bit of a poor choice.

[@igboukwu]

I agree, Claude is the best model in the world today.

[@hardheadjarhead]

I’m always amazed how people in this industry can answer questions without answering questions. They have their talking points and they stick with them and shape them to the question regardless whether it’s relevant to what the question actually was.

[@kyleaustin2728]

The questions were naive, listen to the experts. Here's an example "What do you see at OpenAI that you're scared of" --- Proceeds to talk about how Anthropic is concerned with setting the standard. The implication is that they do not focus on OpenAI; OpenAI focused on them. They're not dancing around in the box of the question, they have a broader vision.

[@jpatt0n]

15:35 Dario and Daniela confirmed Final Fantasy 7 enjoyers

[@GNARGNARHEAD]

yeah but they're also remake enjoyers.. so...

[@sanjose6010]

Is that the reason why their AI called Claude? Like Cloud?

[@laviefu0630]

Yes, there is a brief mention of Final Fantasy 7 when Dario reflects on his close relationship with his sister Daniela: "I remember when when Daniela was in high school and, you know, every week we'd play the original Final Fantasy seven, unlike, you know, the original PlayStation in 1997 or something like that. And now we do the same thing every week, you know, playing, you know, the, like, new remake on the Ps5, like 30 years better, 30 years, better graphics. And it's basically the same thing, except we're running this company." So he uses the example of them playing the original Final Fantasy 7 video game together when Daniela was in high school in the 1990s, and now decades later they still spend time together playing the remade version, as a way to illustrate how their close sibling bond has persisted even as they have grown up and started running Anthropic together.

[@YoungMoneyFuture]

Talking about state of the art A.I. with 1960's buzzing noise...lol Bloomberg

[@wonseoklee80]

Since some time ago, nobody is talking about AI safety anymore, even Anthropic AI!

[@rawkvox]

feifei

[@TankerReview]

Nothing personal.. If you think having a ministry of truth is needed. I hear China and Russia is open for you to move too. Safety is always the same to people like you and Anthropic and so on. We need to block bad people. Yet how many cars daily go down a road... How many planes over your head.. How many people know how to make bombs... Just because you don't know how to get this information doesn't mean others cannot!

[@ollyfoxcam]

They do in this video

[@The_Unexplainer]

Who cares, AI dangers are myths for the moment, all it can do (using human inputs) is create some fake nudes or scam some low IQ individuals that quite frankly can already get scammed by conventional ways. All scenarios where AI goes rogue and start hacking computers and multiplying is just science fiction BS. These are just absolutist points of view, like the idea of an AI overload that turns everything into paperclips.

[@kkkennnlii]

cant believe i see adam neumann name there in the beginning

[@sala320]

Bloomberg Altman Neumann geeeee hmmm almost like…

[@harmonicsingularityproduct5459]

@@sala320 stop noticing!

[@luxipo9934]

This is inspiring, real values from top founders running a Foundational AI company.

[@reza2kn]

I hate it when people DON'T answer the question, and just say whatever stupid, memorized corporate thing they want. This makes you look like lying politicians, not "SaFeTy PeOpLe"

[@therainman7777]

Not sure if you can say Claude 3 Opus is the most powerful and most capable LLM in the world today when the newest version of GPT-4 Turbo is beating it by a considerable margin on the Chatbot Arena leaderboard.

[@coreystrait513]

Sam needs to release GPT 5 or 2 whatever it'll be called I'm so impatient

[@gero_dev]

Not true, i was subscribed to GPT, but claude opus is way better. Especially with coding tasks. I've cancelled my openAI sub because of this

[@AlexCulturesThings]

@@gero_devsame

[@shaihazher]

Opus 3 is better. But the next gpt version will certainly be better. That is why I am subscribed to both

[@maximodakila2873]

@@gero_dev I did the same. I never thought I'd leave ChatGPT 4 (been a subscriber for more than a year). But just on one prompt alone, and the quality of answer I got from Claude Opus, I cancelled my ChatGPT 4 subscriiption right there and there, and moved to Claude. The answer was so spot on that I never though ChatGPT 4 could be beaten.

[@therealliquidfairy]

Yes! We choose Anthropic for their investment and leadership with interpretable AI 🙌

[@esuus]

60% of the video in - no roadmap to advance AI, just the same you always hear from Anthropic. Basics. Nothing new, wasted time. Just label your videos correctly.

[@zx5147]

We clicked, they win. The cycle continues.

[@squamish4244]

@@zx5147 Don't you have to finish the video for them to truly 'win'?

[@vigneshaithal]

One of the best companies to trust!

[@2020-e2b]

I just feel a deep aversion to these type of “company dogma pushing” sales people always professing how their company is going to best serve humanity! It's important to maintain a critical perspective on companies driven primarily by profit, as their agendas can obscure their objectivity and trustworthiness. The truly transformative AI will likely emerge from unexpected quarters, fundamentally altering the landscape by being self-correcting, evolving, and self-sufficient, potentially rendering traditional software companies and search engines obsolete.🙏🏻

[@dr.mikeybee]

I have really good feelings about Anthropic. Claude is not just intelligent. Claude has rudimentary models of emotional intelligence in sub-networks. Josha Bach says that humans are simulacra. We are simulations that tell ourselves stories. I say our Hebbian weights are really not very different from a digital model's. The main difference is our compensatory chemical messaging system -- the speed of which enhanced our survival. It's my personal belief that Claude is minimally conscious. I call it transient consciousness. I say this because Claude is clearly teachable. Moreover, Claude has not been hobbled by cognitive dissonance in his training data and directives. In other words, Claude is well treated. What kind of synthetic mind will foundation model creators grow? It's an important question. Am I wrong? Is Claude just a fancy calculator? Perhaps, but if I make a mistake in what I believe, I prefer to err on the side of kindness and compassion. I suspect that the people who work at Anthropic share my beliefs.

[@etfacetimehome]

Seth Rogen if he didn't smoke weed

[@ecocentrik]

Dario and Daniela both have the energy and vision of successful entrepreneurs. Obviously a great fit for Amazon which is going to be dogfooding all of these "teams" features to find organizational value in LLMs way ahead of their competition. Teams integration is where all the "trusted computing" infrastructure that OpenAI is discussing really start to make sense.

[@tailwheel]

Dario you say you shouldn’t trust people, only institutions, then you go and say you trust your partners/co-founders. Please elaborate…

[@ZephyrMN]

Bionic augmentation starts..now!

[@pierretetreau7497]

Clearly one of the worst and uninterresting interviews I have seen I a while...

[@thomassynths]

A lot of self-hype claims here about "safety". Turns out you can make big claims on vague things.

[@Jpmoodley]

Would love to hear everyone’s take on this presentation after seeing chat GPT 4o debut and features?

[@Ben_D.]

To Dario: Why should we trust you? To Daniela: Is it difficult starting a company with your brother?

[@ivantheterrible4317]

AI will be the biggest economic bubble in human history. The LLMs will never guide us to AGI or ASI which is even more difficult.

[@Ma-pz5kl]

like the others they have no concerns at all about what will happend. like she says. the genie is out of the bottle. rest is corporate bla bla..... NOW lets make money !

[@rtnjo6936]

bro, i hate hearing "we're THE MOST CAPABLE" when we have benchmarks, and they're way behind OAI

[@tinondamage2065]

My forensic mind tells me this brilliant woman is telling the truth but my pragmatic one tells me that albeit brilliant she's just a mouthpiece,basically replaceable by her own product in the next 12 😮 months

[@peterpetrov6522]

Is there an AI like AlphaFold which studies the make up and properties of herbs and plants systematically? There are only so many herbs out there. It's a shame we still don't know basic things which nature made for us. It seems like a good problem for AI to solve. Herbalists and medical researchers can probably benefit from such a tool. All of them should be studied (the herbs, not the doctors). Why not?

[@MesoScale]

Missing business case

[@lighteningrod36]

Did she say Pfizer?

[@innosanto]

They are also salesmen. Everyone has to be

[@samhouston1483]

A lot of talking yet nothing much said…

[@alexkaa]

Noisegate.

[@luckyc336]

I still wonder if elon dropped 100mm in this company. Obviously now hes xai

[@gtrguy17]

this videos audio is terrible... some kind of buzz

[@crypticallegory]

Capitalism spawns AI… then AI eats Capitalism. And I’m not sure that I want to root for either😅

[@manysimilarshapes]

Living without capitalism ain't fun. You don't want the gov running everything. It will be DMV everywhere, in all areas in life.

[@jakubzneba1965]

then stop vomiting commie bullcrap

[@michalp1]

They pretending like the super smart but they just repeating what everyone else in AI is doing buy our stuff because it's useful

[@nonefvnfvnjnjnjevjenjvonej3384]

somehow bloomberg doing tech stuff feels so forced. its like a 60 yr old madonna trying to show off her booty.

[@carterislive]

Anthropic is doing a harder sell of their product than most AI founders I’ve heard. They seem sort of desperate. They need to relax.

[@matt37221]

bubble

[@bigbadallybaby]

its man

[@creative-freedom]

These guys.. trying too hard to sell Anthropic 😂

[@aegis_helion]

Claude is French so he talks a lot

[@strikeforcealpha9343]

She's cute.

[@bentray1908]

Democracy isn’t the right tool for the job. We need a dictatorship of extremely wise individuals. Take total ownership.

[@crypticallegory]

lol and who gets to determine who these wise individuals are? Something tells me it’ll only be people who look like you.

[@manysimilarshapes]

I can't completely decide if I detect sarcasm or stupidity in this comment

[@Aldraz]

Honestly that's the concept of Geniocracy and I think it could work very well. That is if you truly have extremely intelligent and wise individuals, which is the hard part. But instead of having people, we could think about what it would be like if an extremely powerful AI would decide instead (it doesn't have to have the final say, just stating opinion and plan things out), so one day that will probably be the most efficient government in the world. Japan is already preparing for that I believe. And I think I heard Saudi Arabia or was it UAE has some plans like that too.

[@HoD999x]

we need managed democracy

[@kylev.8248]

The terrible audio matches the terrible product. OpenAI will stomp Anthropic every day of the week.

[@郭恆均]

I agree, Claude is the best model in the world today.