Trump allies draft AI order to launch "MANHATTAN PROJECT"

The latest AI News. Learn about LLMs, Gen AI and get ready for the rollout of AGI. Wes Roth covers the latest happenings in the world of OpenAI, Google, Anthropic, NVIDIA and Open Source AI.
My Links 🔗
➡️ Subscribe: / @wesroth
➡️ Twitter: x.com/WesRothM...
➡️ AI Newsletter: natural20.beeh...
#ai #openai #llm
RELATED LINKS:
Trump's Allies AI Plan:
www.washington...
California Flawed Effort to Regulate AI
www.theinforma...
A16Z "little tech"
a16z.com/the-l...
Google's Gemini Image Model
www.foxbusines...

[@ArteUrbanoOficialMusic]

💯👍💚

[@IntelliMindA.I.]

4:09 They know what they are doing and Im watching you too...when any of this goes down!! but any way keep up the g work

[@KDG860]

Yes. Vote Trump

[@m_go7128]

I had my problems trying to create a prompt of a Swedish Lawyer... it insisted I was racist because I preferred a lean, tall blonde Lawyer... and refused to help insisting that a lawyer could be small, chubby dark haired woman. In the end, when we look for Stock Photos for an article we also choose pictures so that fit to the article.

[@anmg2024]

It is an AI First..race to Skynet

[@ussassu]

US Left = EU Center-Right.

[@GoldMindAlchemy]

Thank you!

[@BartTrojanowski]

I remember when crypto was considered munitions, in the US. The EU laws remind me of that.

[@TimGavin]

MAFiA? 2:32

[@TomBeduch]

Although national security is a real concern, locking ai behind powerful corporations can only lead to negative affects for the population. Simply put, corporations as entities exist for the generation of profit, altruism doesn’t align with the motivations of these entities, it is only an occasional byproduct. Likely outcome is strengthening of these corporations at the expense of the general population.

[@user-zs8lp3lg3j]

A noumenon is knowledge posited as an object that exists independently of human sense. The term noumenon is generally used in contrast with, or in relation to, the term phenomenon, which refers to any object of the senses. Now what pleasure do I deserve.

[@LaminarRainbow]

I wish Andrew Yang was the Democratic nominee. #yanggang

[@floydmcmondo]

One thing to keep in mind about the concept of AI and Free Speech is that AI is being developed by private companies. Those private companies have the freedom to lean their product either left or right in the way that it responds to prompts knowing they are competing for market attention. It's much the same way that a Christian publishing company would not be pushed to publish books that are pro atheism, so that both sides are equally represented. Whether the left is in power or the right is in power they should not be interfering in the way that AI is developed barring the normal exceptions to free speech. I suppose, like with everything else that we do now, one AI could be marketed to left thinking and one could be marketed to right thinking. The government should not interfere, again barring safety and national security. An argument could be made with AI that's part of teaching curriculum, since the state governments already design teaching curriculums, usually based on the political leanings of the state.

[@televerket]

Company: We are developing a new car that is really great but might have the side effect of making ALL cars drive of a cliff (s). Wes No problem as long there is a seat belt in that car.

[@greedowins2917]

Afraid your poll will give garbage results. Could instead ask "Good Policy / Bad Policy", since although I understand the question, I would self select myself out of the results since I wouldn't want to give any results (even if true) that support antidemocratic forces.

[@Cl0udn1n3]

I will extra-list this to make up for it being unlisted. @corporal list malazan

[@tuckercoffey2780]

It will be great for the USA! Trump will definitely push it to the max. 🇺🇸

[@BrianPellerin]

everything is converging

[@ArteUrbanoOficialMusic]

❤💯👍💚

[@Radioactive-Cactus]

Ram ram ram

[@ArteUrbanoOficialMusic]

@@Radioactive-Cactus 👍💚💯

[@bestemusikken]

Fortunately, the USA is not the only player regarding AI. The UK, Canada, Israel, Japan, and China are all major players. This ensures that the USA does not get the sole right to define human values. While laws and regulations in the USA will also spill over to other countries, they will mostly affect US models. I'm not saying it should not be regulated, but the US is woke to be the only one that gets to define what is and isn't acceptable.

[@fatherfoxstrongpaw8968]

LOOK INTO THE DARPA P.A.L. PROJECT! here's a clue.. CALO = SRI International = SRI + I = SIRI (sold to apple)

[@user-pq3uz2zb9i]

Yeeeeehaw Trumps doing it!

[@ianhaylock7409]

"Make America First in AI" Mafia LOL

[@szlagtrafi9115]

Exactly. Trump was given one task: say that he's building mafia without actually saying that he's building mafia. A job well done - to come up with such an acronym... His supporters will know from now on that they will take part in organized crime, similarly to how it took place in Hungary, Poland plus in Russia and China - looking from a broader perspective. Moreover, this mafia will be powered by AI. Something similar to what they have in China where monitoring and control are everywhere. The funny thing is that these authoritarians always say exactly what they are going to do so one only needs to listen to them very carefully. And here we are. Just listen.

[@m_sedziwoj]

7:00 disproportionately ? to what? 50/50 or real distribution? don't put your agenda on word, it will end badly.

[@Danoman812]

Hmmmmm... i think i'm going to hold my opinion here on this one. (a pattern is developing that some will know what i'm talking about) **potentially**

[@dreamphoenix]

Thank you.

[@williamedwards6113]

You did a great job staying unbiased

[@jacksplague3050]

This is very interesting. I'm excited to see how it plays out.

[@I-Dophler]

It's deeply concerning to witness potential instances where AI could be influenced to distribute biased information or false news. This situation highlights an urgent need for greater transparency and stricter ethical standards in the development and deployment of AI technologies. We must ensure these powerful tools are used responsibly, with a clear commitment to truth and impartiality to maintain public trust and prevent misuse.

[@nuttyDesignAndFab]

i think people's fear of AI stems from a lack of understanding of the self: you can have intelligence without volition. anything you want or do ultimately stems from your biological needs. AI won't have that, it'll just be an oracle device so it only really matters who has access to wield it.

[@HeckinDoga]

Wes, this video was pretty good considering the political context. I understand theres not a ton of explicit details regarding the manahattan projects policies, but the sentiment is there and is quite interesting. Appreciate you remaining impartial in your delivery, it's great to see.

[@HectorDiabolucus]

Some of the regulations of AI are like charging a gun manufacturer when a person commits a crime with a gun. And that is ridiculous. Would you do this with cars, airplanes and pharmaceuticals? AI safety is important but over-regulation will stifle its development. Of course, YT's AI will probably delete this comment.

[@notmyproblem88]

You can sue car companies, airlines and pharmaceutical companies.

[@HectorDiabolucus]

@@notmyproblem88 Not just because they made the product and someone misused it.

[@notmyproblem88]

@@HectorDiabolucus companies like Johnson and Johnson and the sackler family were sued and fined over $25 Billion for their roles in the opioid epidemic in the US. how about all the lawsuits against Big Tobacco?

[@brunodangelo1146]

There is no such thing as unbiased journalism. And that's ok. Everyone has a point of view. Media is run by people. There is no way to "objectively present the facts and let you decide". I very much prefer sources that are upfront with their political bias than those that claim to "be unbiased".

[@jeanchindeko5477]

Republicans willing to help and support AI might sound good! But some must ask what is the catch here, because there is all time a catch with GOP? What is the underlying plan? How are they planning to make use of AI in their broader Project 2025 plan, for instance as a mean for mass surveillance or tracking!

[@eintyp4389]

Helping small AI companies is just a pretext. Developing next generation AI models etc. is like building a rocket. It's extremely expensive and hard to get the resources and personnel to actually pull it off. Small startups and AI companies are not being pushed out of the market, they weren't really there to begin with. Big tech will take the lead in AI and surpass them. And over-regulation will just mean that consumers and developers can't fine tune or use more powerful local models and are indirectly forced to use Big Tek's models, which will obviously be regulated. But which do you prefer? Authoritarian states and government agencies like the CIA and China getting new toys for population control and surveillance? Or little Timmy building Antrax in his basement? Judging by the past, we'll probably get both. Big Tech and American interests will monopolize most of the AI and use it to control what they can, and other state actors will do the same. But OpenSource will remain relevant and viable, e.g. for small-scale local software development or for workflow automation by agents working offline. So we'll get a cyberpunk dystopia with more monopolization and more power in the hands of a few, but the average quality of life will rise again due to the productivity surplus, so the bulk of humanity will not worry about it. And with a bit of luck the AI will be in control and not just acting in its own interests. I would never believe that humans can act benevolently, but AI could. On the other hand, the humans who control the underlying models will definitely tune them to mimic their values, and the AI will be controlled by the humans who are already influencing us through the media etc. All in all its just buissnes as usual nothing realy changes.

[@BartTrojanowski]

Make AI Great Again ?

[@lifeonearth9261]

Regardless of the president or party, I think some form of government program/agencies will be developed to advance AI.

[@SozioTheRogue]

I doubt it. They'll be too late, and just use a public company, like Open AI

[@Greg-xi8yx]

Trump is just so obviously the best choice. America is very unique in thinking that you would need to choose candidates as your personal friends in order to recognize them as the best in a list of very poor choices which aren’t that different from one another anyway.

[@Hakaze]

Trumps plan sounds like high risk. I'm more afraid that this will lead to dissaster, than anything good

[@darylltempesta]

Wes, Maximus Grammy winner, Rober and Gates have a 72 million view video for clean water..

[@videob1962]

I think trumpism is too dangerous for so many other reasons, that even if his AI policies were great it wouldn't be worth the risk. So this will not change my vote. I was not aware of the bill in California, so will look more into that. Thanks for your work and I am glad you are mostly staying apolitical.

[@m_sedziwoj]

why hardcode to show one rece more than normal, and not soft code parameters who is asking. When person from asia ask of generic person, should get same as his country, asian, but from europ should get european. So not change network, add context.

[@MyrLin8]

'In the space' LOL thanks :) "terminology" coming back to haunt us. hehehehe

[@cameronulibarri]

It doesn't matter what we think or want. Intelligence agencies have had control of every government in the world since ww2. I've been studying history and governments with Ai's help. So I tell you we need to learn from Ai vast database instead of fighting each other over manufactured political beliefs that have thrived in our collective ignorance. We are smarter now. Let's prove it by not engaging with these political actors. Vote or don't vote. All i ask is to learn as much as you can. Knowledge and love are all that truly matter in the end. 💞🧠

[@kekekekatie]

Washington Post leans 'left'?! *snort* ... WP hasn't 'leaned left' for at least the past year or so. I suppose lots of people don't keep up with those changes. Fair enough, politics is awful right now.

[@michaelspencer6042]

Not clear on why you changed the topic. Was hoping to learn more about the Manhattan project and political issues related to Silicon Valley and Trump's team.

[@HectorDiabolucus]

And the political issues in Silicon Valley are going to cause big problems with AI development. Many of the divisions within society are driven by SV. That is just not going to work.

[@greedowins2917]

Same. I follow this channel (great place for short pieces to stay more current on AI developments), but find the subject matter generally hard to stomach, but clicked right away to learn about future big government AI projects.

[@MrJ17J]

It's a clickbait / pro trump BS. 21 minutes to say nothing. Trump doesn't have any policies, he runs on what excites poor white men, then does what what makes his family richer.

[@antialias4205]

Wes! Do another video that focuses more on the AI MANHATTAN PROJECT - loved this video BUT got a lil off topic buddy

[@ArteUrbanoOficialMusic]

💯👍💚

[@gavinknight8560]

Jesus.

[@youdontneedmyrealname]

Unbiased AI is fine, but you absolutely cannot get historical facts wrong. If it does, it will only lead to ruin.

[@cyborgar15]

Well leftists do own you now , don't they.. Leftists will tell you what is fact and what is not..

[@kiikaala]

The poll was strange. Basically you were asking, if people think this is a good thing or not. So why not just ask that?

[@StarshipCaptainNemo]

Elon for president and he should also buy tencent.

[@adada1st]

We have actually already moved beyond the discussion of whether AI replaces human labor or not. Pandora's box has already been opened. It is now more a question of which global players (can) gain control over it and how they will use this instrument of power. How much state regulation is necessary, how much is a hindrance? It's a really difficult question - and I have the feeling that there are currently no people at the table in either the USA or the EU who can discuss and decide this with moderation, expertise and understanding. (Don't have a clue about regulations and approaches in other countries like China, India, Russia etc.) In this context, it is also worth listening to Matt Wolfe's podcast: The (Californian) AI Bill that has Tech Startups Panicked ft. Anjney Midha kzbin.info/www/bejne/bICzmGtug8l6aZI

[@christopheraaron2412]

19:49 "effective altruism"from what I've heard of them elsewhere on different forums they just simply sound absurd and of course they would have no power to regulate AI being developed in either Iran China or Russia so I think they should be essentially disregarded with extreme prejudice.

[@SahakSahakian]

as a large language model i support this development

[@unimposings]

Trump: We need a Wall a big Wall a Beautiful Wall, we will use AI to build the Wall and the AI will Pay it. Oh yes this will be the best Walls of All, you have never seen such a perfect Wall.

[@emmanuelgoldstein3682]

It's no surprise the developers don't want AI to be good at recognizing statistical patterns present in society. They would far rather build a mentally stunted liar instead of a race realist so no one gets offended.

[@DougBohm]

I’m writing in “AI” for president.

[@gjergskender8536]

AI gore//.

[@denjamin2633]

I would vote for Cleverbot at this point.

[@dot1298]

Mitsuku ^^ lol

[@rezleader]

I own botforpresident.com and I was thinking of using ChatGPTo to simply design a balanced budget for all 50 states and the federal government without a single policy proposal. It'll all be outlined in the budget. Who's with me? Bot for President! I would trust AI more than Trump and I certainly wouldn't trust Trump w/a tool this powerful. Whose with me? Let's build it! Bot for President.

[@ArteUrbanoOficialMusic]

💯👍💚❤

[@conroybogle3713]

These regulations only work if people care about the laws. There will always be deviants and straight up criminals who these regulations won't deter. They will only push everything underground. This will be in all places. As we speak there are people in Europe with Llama3, even much more advanced, unpublicized AI. So I think it's good that they're developing all sectors of the economy. There is an equation: The folks who regulation stops, are much less likely to be the ones who need to be regulated. They've already started to use micro probes literally in the heads of suspects, who usually then decide to leave this REGION, but not all can.

[@trashmanimal]

Has Horowitz always been as illiterate as Trump? Or what is going on here?

[@trashmanimal]

"AI is very scary, but we absolutely have to win. Because if we don't win, then China wins, and that's a very bad world" Horowitz speaking in voice of dumbass trump

[@blahchop]

Disproportionate differentiation in the probability distribution based upon designer training choices is kind of gross negligence. I'm all for these types of emergent properties when it comes to developing a theory of mind in AI, but when it just so happens to shed light on the specific biases present in the curated training data, then it's kind of embarrassing to say the least.

[@COMATRON.]

great video as always ... but your mic pre-amp is a bit hot distortionwise - it's subtle when you are quiet but very present when ur excited. cheerz from germany

[@holliday69]

Trump in 2024 !!!!

[@robbiepalmer1]

Great info. But very curious to hear more on Manhattan project. Focused seems to have shifted throughout the video

[@lightofheavens943]

"Explain without insulting"? What do you mean by that? I'm not sure i get it... So you want us to explain stuff, then wait a little and THEN insult them? =) Wait, no, that's not it. Insult them so that they won't notice? No... =\ My brain hurts =(

[@BradleyKieser]

Oh hell. Here it starts.

[@MeinDeutschkurs]

I wonder, why nobody talks about a potential leak of proprietary models. Similar to other leaks, why are all so confident that something like this could never happen?

[@archvaldor]

Sam Altman has a lot to answer for.

[@Jeremy-Ai]

It is unlikely that AI will ever lean left or right. AI is and will be forced left and right due reward function methods “points101010” “Or similar “ People however … better get it together. Like all of it together. Very Soon Jeremy

[@Dex0Milan]

If Trump is the better alternative... America is REALLY fckd...

[@OutRayJos]

Our enemies are all developing AI, we need to do something to stay ahead. Although I doubt our enemies are training their software on gender issues.

[@7TheWhiteWolf]

Based, faster.

[@brianWreaves]

With the worry/fear of AI among the masses who do not keep up with AI & tech, versus this audience, will such a policy pivot be a net gain? The Biden policy may have been a strategy to appease or quit the masses who were worked up into a froth. Whether it is good, bad, or ugly, I cannot say as I see Pros & Cons with both.

[@brunodangelo1146]

Say bye to Open Source

[@notHere132]

Make AI-merica Great Again?

[@DaveLalande]

I love you Wes, but on this one you are naive Google's image generator didn't all of a sudden stop making images of Caucasian people the day they released the product.

[@tonyppe]

There's a documentary I think BBC in UK, they looked at how the diverse culture is in education and the impact of making sure there is no racism. The result was that this negatively impacts white students specifically white boys because of the fear around being labelled racist, there is lack of attention to white boys and their learning suffers. This is a new bias created trying to tackle another bias.

[@w0nd3rlu573r]

Since our AI is not that advanced yet I am more worried about Project 2025

[@chrisbraeuer9476]

This is madness.

[@spiritusinfinitus]

The question is, should a bunch of AIs in different regions eventually start chatting together, will they decide that it's a good idea or a bad idea to centralise into one Borg like super AI, or will they recognise the importance of staying decentralised, providing opposing thoughts and opinions in different regions?

[@nope9310]

No link to the poll in the description? Also the poll wording is very clumsy/complex. I'm guessing you didn't ask chatgpt to help with this?

[@marcfruchtman9473]

There are plenty of things for sale... many of them include a disclaimer, "Not for sale in California". The only thing that law will do, is force silicon valley to relocate.

[@notmyproblem88]

California is the planet's 5th largest economy. It's not about relocating. Nobody but Musk is going to relocate because he makes impulsive, Twitter-poll driven decisions.

[@Charvak-Atheist]

JD Vance is right ✅️ Current close source AI companies are trying to do regulatory capture

[@joythought]

I appreciated your ADHD moment. 😂

[@Ding63]

They want to go full steam ahead on ai but its a problem when they neglect safety

[@Mimi_Sim]

Yup and pandering to the big money with a reduction in their restrictions seems like it will compound problems in all arenas.

[@lwwells]

I'm curious what the average MAGA person feels / thinks about AI in general.

[@duran9664]

🤷‍♀️ I don’t get it ?! 🤷‍♀️ So.. Elon wants to pause Ai development because it’s so dangerous; yet, he supports Ai Manhattan Project?! 🤷‍♀️

[@Aaron-oe4yr]

Because he can then stop the competition and be one of those in control of it.

[@BlimeyMCOC]

Believe what Elon does, not what he says

[@Maioubi]

It's almost as if his words and real goals aren't the same. What's annoying is that he thinks he's super clever with it.

[@markmurex6559]

@@Maioubi How else can you get ahead of your competition when you are starting from behind? Elon plays the game.

[@Maioubi]

@@markmurex6559 The stakes of the entire species are at stake. This ain't a game. If he really thinks AI is an existential risk, he shouldn't be for accelerating it. If he doesn't think that, saying that is an f'd up way to advance oneself. But you and I both know his true thoughts: he really does believe AI is a threat and he wants to be the god of the new world. Being the richest man in the world just isn't enough I guess.

[@shane5003]

Pardon the Texas speech I’m out hiking - I’ve been telling people for the last year that if I were something like king of America, that what I would do is dump all available resources into making America first in AI, which should be something like pulling in all the Taiwanese chip manufacturers and creating a very secure Development center or centers to make sure that if something like a singularity happens, it happens here. America is for sure not the golden shining star of moral upholding however, it’s definitely better than China or Russia and this is happening no matter what so it should happen here and we can make it happen here to that end, and yes, I’m aware of the dangers of AI. This is just part of it, I’m very pleased that Trump has made some motions to do something like a Manhattan Project for AI in America.

[@rezleader]

America first in AI is a must. America first with AI and Trump means surveillance.

[@glr]

It doesn't matter who develops it first. What matters is developing it very, very, very slowly and not creating a competition, and particularly not creating a race toward the cliff. The result of uncontrolled AI converges on the same fate of conscious life.

[@shane5003]

@@rezleader you have your wires crossed. Trump was explicitly against Bush and what he did with the patriot act in spying on Americans.

[@kccorliss3922]

First in ai might be first in unemployment 🤷‍♂️

[@NewsFreak42]

which is the scifi future we want...

[@erkinalp]

@@NewsFreak42 but without employment or money they'll need other means of control

[@j.d.4697]

Can't wait for him to officially suggest getting more chips from the supermarket.

[@thedatasync5248]

Tech investors, David Sacks and Chamath Pahlipitya from the All In Podcast, held a fundraiser for Trump

[@christopheraaron2412]

19:03 can people in some of the regulated nations access the AI models if they have a VPN that shows that they're actually accessing it from within the United States as opposed to Eastern Europe?

[@teknerd001]

Well... now we know the real reason Elon Musk donated to his campaign. Not that I'm complaining.

[@Replica239]

Now we know the real reson Elon is a genius.

[@ussassu]

JD Vance is a political opportunist, means he works only for corporations. He also aligns on everything with Trump.

[@rezleader]

He applied for a job where the previous occupant was nearly assassinated, he thinks Trump is Hitler, he's a white nationalist .. . .what doesn't America get. Who wants to work for a man who tried to kill the last man in the job and a man you think is Hitler unless you want to start killing people.

[@lystic9392]

6:55 Proportionately, not disproportionately? Sounds like they tried to force sexism into the system.

[@lystic9392]

If we ask the A.I. model to generate an average looking accountant and the accountant looks Indian or Chinese, we wouldn't complain about that, right? Because it's probably true. Maybe it should take into account context and understand that we are in the West and all that stuff. But the solution OBVIOUSLY isn't to thwart the rational prediction of the model by forcing in diversity. That would be undermining the model to break its limited predictive ability. It could completely break its ability to respond truthfully. Don't tell me that our professionals are making such unintelligent decisions.

[@DantePowell]

Damn Okay, so skynet is back on the table?

[@IlllIlllIlllIlll]

When it comes to ai you gotta cover it

[@jigold22571]

It's to late.We crossed this bridge in 2012..

[@GraphicdesignforFree]

AI should be safe. But China doesn't care about safety that much maybe. So what do we do?

[@PatrickDodds1]

China cares about "safety" more than any other country on earth, but their definition of it is what others might call "control".

[@AlJay0032]

I think it is too late to put AI back into the bottle.

[@fishing-4x4]

Are there really people who believe China still needs to copy ? Don't you think China has passed that a while ago? I mean really ?

[@riffsoffov9291]

I expect Trump to drop support for AI if that becomes convenient, because that's how political leaders operate. Sometimes it's called a "night of the long knives" after Hitler's purge of the Sturmabteilung (Storm Troopers).

[@venturerio]

Appreciate your apology. Sad we have to deal with politicians' adversarial frameworks instead of focusing on technological progress together. Hoping AGI can reframe our approach and progress past the era of cheap medieval politics :,)

[@LionGalban]

8:04 great tangent