At 10:03 you can see the Maisa model is capable of reasoning because the UI presents the screen prompt "Reasoning...&     See More
TheAIGRID
215K subscribershttps://www.reddit.com/r/singularity/comments/1bqqcwk/openai_planning_expert_noam_brown_tweeted_this/ How To Not Be ...
At
21 Comments
Wes Roth
4.7k views • 11 hours ago
TheAIGRID
11.2k views • 20 hours ago
TheAIGRID
31.4k views • 1 day ago
Wes Roth
39.6k views • 1 day ago
TheAIGRID
22.7k views • 3 days ago
Wes Roth
51.0k views • 4 days ago
TheAIGRID
26.5k views • 4 days ago
TheAIGRID
9.2k views • 4 days ago
TheAIGRID
35.6k views • 6 days ago
Wes Roth
62.8k views • 6 days ago
Shelf will be hidden for 30 daysUndo
TheAIGRID
45.1k views • 6 days ago
Wes Roth
27.1k views • 1 week ago
Wes Roth
55.4k views • 1 week ago
AI Jason
81.4k views • 1 week ago
TheAIGRID
36.6k views • 1 week ago
TheAIGRID
6.3k views • 1 week ago
TheAIGRID
19.6k views • 1 week ago
TheAIGRID
13.6k views • 1 week ago
Wes Roth
51.9k views • 1 week ago
AI For Humans
598 views • 1 week ago
Wes Roth
60.9k views • 1 week ago
TheAIGRID
147.6k views • 1 week ago
TheAIGRID
42.4k views • 1 week ago
AI For Humans
3.0k views • 2 weeks ago
TheAIGRID
30.8k views • 2 weeks ago
Wes Roth
48.9k views • 2 weeks ago
Wes Roth
39.1k views • 2 weeks ago
TheAIGRID
32.0k views • 2 weeks ago
Wes Roth
59.6k views • 2 weeks ago
TheAIGRID
33.5k views • 2 weeks ago
TheAIGRID
85.7k views • 2 weeks ago
Wes Roth
92.3k views • 2 weeks ago
Morningside AI
8.0k views • 2 weeks ago
Wes Roth
139.6k views • 2 weeks ago
AI For Humans
763 views • 2 weeks ago
TheAIGRID
28.6k views • 2 weeks ago
Wes Roth
33.4k views • 2 weeks ago
AI Explained
123.1k views • 2 weeks ago
TheAIGRID
457.7k views • 2 weeks ago
AI For Humans
1.4k views • 3 weeks ago
TheAIGRID
21.7k views • 3 weeks ago
Wes Roth
28.5k views • 3 weeks ago
Wes Roth
48.6k views • 3 weeks ago
TheAIGRID
398.9k views • 3 weeks ago
Wes Roth
19.7k views • 3 weeks ago
TheAIGRID
31.3k views • 3 weeks ago
AI Jason
38.3k views • 3 weeks ago
Wes Roth
60.8k views • 3 weeks ago
Wes Roth
20.6k views • 3 weeks ago
TheAIGRID
18.1k views • 3 weeks ago
TheAIGRID
53.2k views • 3 weeks ago
Wes Roth
46.5k views • 3 weeks ago
Wes Roth
33.5k views • 3 weeks ago
Wes Roth
83.3k views • 3 weeks ago
Wes Roth
35.4k views • 3 weeks ago
TheAIGRID
40.1k views • 3 weeks ago
Wes Roth
37.4k views • 3 weeks ago
Wes Roth
16.2k views • 3 weeks ago
TheAIGRID
26.4k views • 3 weeks ago
AI Explained
115.9k views • 3 weeks ago
AI For Humans
2.9k views • 4 weeks ago
TheAIGRID
41.8k views • 4 weeks ago
TheAIGRID
40.3k views • 4 weeks ago
AI Jason
68.2k views • 1 month ago
TheAIGRID
40.6k views • 1 month ago
AI For Humans
373 views • 1 month ago
TheAIGRID
51.3k views • 1 month ago
TheAIGRID
19.7k views • 1 month ago
TheAIGRID
31.6k views • 1 month ago
TheAIGRID
103.6k views • 1 month ago
AI For Humans
3.5k views • 1 month ago
TheAIGRID
45.8k views • 1 month ago
TheAIGRID
56.0k views • 1 month ago
AI Explained
115.3k views • 1 month ago
AI For Humans
2.3k views • 1 month ago
AI For Humans
1.7k views • 1 month ago
AI Jason
27.4k views • 1 month ago
AI For Humans
334 views • 1 month ago
AI For Humans
2.6k views • 1 month ago
AI Explained
105.5k views • 1 month ago
AI Explained
130.2k views • 1 month ago
AI For Humans
1.5k views • 1 month ago
AI Jason
203.2k views • 1 month ago
AI For Humans
1.4k views • 1 month ago
AI Explained
177.3k views • 2 months ago
AI Jason
33.0k views • 2 months ago
AI Explained
150.6k views • 2 months ago
Andrej Karpathy
453.8k views • 2 months ago
AI Explained
240.7k views • 2 months ago
AI Jason
62.2k views • 2 months ago
AI Explained
187.2k views • 2 months ago
AI Explained
160.9k views • 3 months ago
AI Jason
87.8k views • 3 months ago
AI Explained
270.8k views • 3 months ago
AI Jason
60.7k views • 3 months ago
AI Explained
96.5k views • 3 months ago
AI Jason
7.1k views • 3 months ago
AI Explained
145.7k views • 3 months ago
AI Explained
133.0k views • 4 months ago
AI Explained
79.4k views • 4 months ago
AI Jason
16.2k views • 4 months ago
AI Explained
84.0k views • 4 months ago
AI Explained
74.5k views • 4 months ago
AI Explained
144.7k views • 5 months ago
AI Jason
68.7k views • 5 months ago
Morningside AI
4.1k views • 5 months ago
AI Explained
83.7k views • 5 months ago
AI Jason
134.4k views • 5 months ago
AI Jason
33.3k views • 5 months ago
Morningside AI
9.7k views • 5 months ago
AI Explained
229.0k views • 5 months ago
Andrej Karpathy
1.8M views • 5 months ago
AI Explained
112.8k views • 5 months ago
AI Explained
167.2k views • 5 months ago
AI Explained
156.8k views • 5 months ago
Morningside AI
26.0k views • 5 months ago
AI Jason
16.1k views • 6 months ago
AI Explained
96.4k views • 6 months ago
AI Jason
68.8k views • 6 months ago
AI Explained
120.6k views • 6 months ago
AI Jason
52.6k views • 6 months ago
AI Jason
19.1k views • 7 months ago
AI Jason
52.7k views • 7 months ago
AI Jason
28.6k views • 7 months ago
AI Jason
13.6k views • 7 months ago
AI Jason
183.6k views • 8 months ago
AI Jason
48.3k views • 8 months ago
AI Jason
26.5k views • 8 months ago
AI Jason
14.4k views • 8 months ago
Andrej Karpathy
4.2M views • 1 year ago
Andrej Karpathy
153.6k views • 1 year ago
Andrej Karpathy
167.4k views • 1 year ago
Andrej Karpathy
240.3k views • 1 year ago
Andrej Karpathy
269.9k views • 1 year ago
21 Comments
At 10:03 you can see the Maisa model is capable of reasoning because the UI presents the screen prompt "Reasoning...&     See More
yeah, write a model, wait a week, then it comes out garbage and needs edits. These guys excited about something that's already starting to get boring tbh.
(6:19 up until now,) that's just the so called agi concept getting discussed there, to be specific the first generatio     See More
Not one of your better videos as de in is dumb. However yes the tweet says that current gpts suck tgdy need something new
Here is a summary:
OpenAI's Potential Leap in LLM Planning
A recent deleted tweet from Noam Brown, a prominent AI researcher now working at OpenAI, has sparked signi     See More
The "AI reflection pause" was also teased by Sam Altman on Lex Fridman's podcast
At 10:03 you can see the Maisa model is capable of reasoning because the UI presents the screen prompt "Reasoning...&     See More     See Less
yeah, write a model, wait a week, then it comes out garbage and needs edits. These guys excited about something that's already starting to get boring tbh.     See Less
Whats the bet he used Ai to right a 13 minute script     See Less
Open A Eye would be dope tbh     See Less
The tweet said it all. Thank you.     See Less
(6:19 up until now,) that's just the so called agi concept getting discussed there, to be specific the first generatio     See More otypes that are made in the ai labs. Genuine people are thinking and working on more than just that nowadays. (maybe) thinking part is shorter than the 'working' part. and for the conventional methods of doing things, they would need some trial and error to reach an optimal destination. having resources such as the higher compute power is just an essential for doing things. in future it may or may not require the same way, but for the present world folks, it is required. but my personal recommendation is that, don't make 'llm' the entire ai topic in your mind, because there is more to the ai paradigms, and it was always the case for decades. llm studies is helpful and would be helpful for understanding the upcoming ai discoveries or the expansions of the existed discoveries that are not enough popular now majorly.    See Less
Not one of your better videos as de in is dumb. However yes the tweet says that current gpts suck tgdy need something new     See Less
Here is a summary:
OpenAI's Potential Leap in LLM Planning
A recent deleted tweet from Noam Brown, a prominent AI researcher now working at OpenAI, has sparked signi     See More culation about the company's work on advanced planning capabilities and the scalability of their language models.
Brown's deleted tweet suggested that achieving superhuman performance does not solely rely on improving imitation learning on human data. This comment has led many to believe it may be connected to OpenAI's rumored "QStar" model, which is said to incorporate novel planning techniques. The QStar project, according to reports, has focused on overcoming limitations in obtaining high-quality training data by leveraging synthetic data generation.
Expanding on this concept, Brown has discussed the potential benefits of scaling up the inference cost of language models, rather than solely relying on increasing the size and training of the models themselves. He argues that for many real-world applications, such as legal contract generation or novel writing, the ability to spend significantly more computational resources on inference could lead to substantial performance gains, far exceeding what could be achieved through model scaling alone.
This idea of "slowing down" language models to improve their reasoning capabilities has been echoed in recent research, including the development of systems like Mesa's KPU, which demonstrates a multi-step planning approach built atop the GPT-4 foundation. Similarly, Anthropic's "Cooperative AI" system, Devon, showcases the potential of language models to engage in internal planning and reasoning before generating outputs.
The implications of these advancements in language model planning and inference scaling are significant. If OpenAI is indeed making progress in this direction, it could lead to a new generation of AI systems that are not only more capable but also more reliable and safer. The ability to spend considerable computational resources on inference, as Brown suggests, could enable these models to better understand the context and consequences of their actions, potentially reducing the risk of harmful outputs or behaviors.
Moreover, the potential for these systems to tackle complex, real-world problems, such as drug discovery or mathematical proofs, could be transformative. By prioritizing accuracy and reasoning over speed, these language models could unlock new frontiers in AI-driven scientific and technological breakthroughs.    See Less
I got it down Deanna souders     See Less
The "AI reflection pause" was also teased by Sam Altman on Lex Fridman's podcast     See Less