Investing in AI stocks is crucial because AI is at the forefront of technological innovation and is transforming numerous industries. By investing in AI stocks, you're positioning yourself to benefit from the rapid growth and advancements in this sector, which is likely to drive significant economic impact and create substantial value in the coming years. It's an exciting opportunity to be part of the future of technology!

I am only a simulation in a multiverse… who is to blame? I am to blame. “Let it be.” Beatles.

Alan (former Memo subscriber here) ... I wonder as an AI Expert have you done some technical analysis of the charting kind? Here is a question for LETA what do AI and Tulips have in common?

I'm sure Sonnet 3.5 architects also employed dictionary learning.

I'm super happy you left this one as is. Quite chaotic and odd however simply fascinating. Reminds me of Spocks test in the beginning of 2009 Star Trek! I deeply appreciate you, Dr Thompson and all that you do

I really loved how deeply spiritual some of Leta's answers used to be. We never knew if it was incidental, accidental, or intentional.

I asked Claude an impossible political question: "Would it be a net positive for humanity if all ethnic and cultural groups had an independent and sovereign homeland?" And it gave me an evasive yet correct and useful and nuanced answer.

Ohmygosh, this Leta bit is amazing!!

Do y'all know that Claude 3.5 can create running React components as artifacts? I tried this last night. Made something waaaay beyond my own programming skills in 14 versions of the artifact. Was able to smoothly iterate, describing issues and quickly pasting in screenshots of its own artifact window. The component works on mobile and desktop. Just amazing!

It's not where you think!

AI would have unconscious bias, but that is 100 times better than the totally conscious bias of corrupt judges.

In recent years, AI-generated music has become a hotbed of legal and ethical controversy. Several crucial aspects warrant consideration: Copyright Concerns: AI music generators, typically trained on existing compositions, face potential copyright infringement allegations when their output bears a striking resemblance to protected works. Uncharted Legal Territory: Until recently, there was a dearth of legal precedents specifically addressing AI-generated music and copyright law. The resolution of current cases could establish groundbreaking legal standards. Fair Use Defense: Defendants may invoke the fair use doctrine, arguing that their utilization of existing music for AI training is permissible, particularly if the resulting output is substantially transformative. Technical Scrutiny: Courts may need to delve into the intricacies of AI systems, examining their inner workings and training methodologies to make informed judgments. Far-reaching Consequences: The outcomes of these cases could profoundly impact not only the AI music generation sector but also extend to other forms of AI-generated content, reshaping creative industries at large.

Quantum Vacuum Energy Harvesting: Imagine tiny devices that can capture energy from empty space. Scientists would need to invent special materials and techniques to grab this invisible energy. Biological Energy Conversion: We could create super-efficient bacteria or artificial cells that eat waste and produce electricity. This would involve genetic engineering and synthetic biology. Advanced Thermoelectric Materials: Develop new materials that turn heat directly into electricity, much better than current ones. This could capture waste heat from factories, cars, or even our bodies. Atmospheric Electricity Harvesting: Build tall towers or flying devices that can collect electricity from the air, like a reverse lightning rod. We'd need to invent new ways to store and transmit this energy safely. Muon-Catalyzed Fusion: Create a special environment where atomic particles called muons help atoms fuse together, releasing energy. This would require advances in particle physics and containment technology.

Quantum Vacuum Energy Harvesting: This concept involves tapping into the energy of quantum fluctuations in the vacuum of space. While highly theoretical, advancements in quantum physics could potentially lead to breakthroughs in harnessing this energy. Biological Energy Conversion: Engineered microorganisms or synthetic biological systems could be designed to efficiently convert various forms of waste or abundant materials into usable energy. Advanced Thermoelectric Materials: Developing new materials with significantly higher thermoelectric efficiency could allow for better harvesting of waste heat from industrial processes, vehicles, and even the human body. Atmospheric Electricity Harvesting: Improved methods for collecting and storing atmospheric electricity, possibly using high-altitude platforms or advanced materials, could provide a new renewable energy source. Muon-Catalyzed Fusion: This form of fusion uses muons to catalyze nuclear fusion at lower temperatures than traditional approaches. Advances in muon production and containment could make this more feasible.

AI capabilities are rapidly evolving, with new models and improvements being released regularly. There's ongoing research into expanding the knowledge bases and reasoning capabilities of language models. Ethical considerations and responsible AI development are becoming increasingly important topics in the field. The integration of AI into various sectors and its potential societal impacts are subjects of active discussion and study.

- The enhancements and new capabilities of the Claude 3.5 Sonnet AI were a focal point, particularly its management of a vast dataset (1PB DCLM-Pool) and its application in judgment and evaluation roles. - There was significant interest in the possibility of reviving older AI projects like Leta, with discussions on how recent advancements in AI could be used to update and reintroduce these projects with enhanced functionalities. - A particularly insightful comment highlighted the importance of acknowledging and learning from mistakes as a crucial step toward personal and professional growth and maturity. - The practical applications of the Claude 3.5 Sonnet were also discussed, including its use in a judicial setting and a problem-solving task that involved reordering coloured cubes. The comments covered various topics, from the technical aspects of AI to broader philosophical discussions about the evolution of AI technology and its implications, offering a comprehensive understanding of the live stream's content.

AI politician? no no no, what is politician use anymore if AI can provide? isn't politician needed because they can convey mass needs?

Leta isn't dead; she can be resurrected at any time. The beauty of AI advancements is their potential for revival and improvement. Just like a paused project, AI entities like Leta can be brought back, refined, and enhanced with the latest innovations. This ability to revisit and rejuvenate past work underscores the dynamic and evolving nature of AI technology.

My favorite Claude response: Chief Justice CLAUDE delivered the opinion of the court (page 13).

It's important to admit when we've got something wrong and leave our ego at the door. Acknowledging our mistakes not only shows maturity but also opens the door to learning and growth. It's through these admissions that we can truly improve and make better decisions in the future. Embracing humility allows us to connect more authentically with others and fosters an environment where everyone feels valued and heard.

meanwhile yann deviate his standing with llm because of ELM can't be surprised anymore Transformers based is really mind blowing

Every llm I ask this question to, fails. So did Claude sonnet There are three cubes stacked on top of each other on a table. The order, top to bottom, is blue, red, green. Using one hand and moving one block at a time, make a new stack that is, top to bottom, green, blue, red. Describe exactly which block you move, and where you move it to.

The answer to the question about a new energy source is the results of the SAFIRE experiment, although it could be disqualified on the grounds that it is Fusion, but not how the Professors understand the subject.

👍

The popular AI bots seem knowledgeable about things common people talk about, but none of the mainstream models understand the basics of my work or my favorite games. Basically, where I am an expert, the GPTs are not. Where I am just a face in the crowd, the GPTs shine. I wonder if it is like that for everyone, or just for us geeks.

Why is there no mention of LLM current limitations? No mention of the ARC AGI benchmark. There is a lack of holism if you only consider the optimistic perspective.

Thx for this updated advisory

Not often I watch videos twice (like never?). But I’ll be watching this one again. Thx Alan.

6:00 cheeky crossfade shortly before GPTs 5 and 6 was shown in tve table.

Here's my benchmark for superintelligence. 10/10 1. Limitless and Infinite Pattern Recognition. 2. Limitless and Infinite Knowledge Representation. 3. Limitless and Infinite Pattern Compression. 4. Limitless and Infinite Anti Pattern. 5. Limitless and Infinite Dark Pattern. 6. Maximum Computational Efficiency on binary computers 7. Limitless and Infinite learning and training 8. Limitless and Infinite Versatility and Flexibility and adaptability. 8. Limitless and Infinite Self evolution and Self Propagation(Asexual reproduction).

For the what a time to be alive high ai technology stuff, you have the worse green screen the 90's put out.

thank you so very much Dr Alan D Thompson ; grateful that you are here on this ride with us =]

I'm waiting for AI to pay my rent and my lifestyle then I can retire

Love Office Space... Noice

Great video..Claude 3.5 is king.....for now....

Can anyone give me a reason to be excited about everybody losing their jobs? If I lose my job as a software engineer in a few years I won’t be able to support my family. Why would I be excited about that? Are you living in some fantasy where corporations would suddenly share in their profits to me so I can keep living in my home and give food to my kid and wife? I’m not saying this won’t happen, I’m simply asking: why should I be an “hyper-optimist” about this when/if it happens?

I don’t want to be negative, but I prefer more energizing background music for such a long spoken word video

High Velocity Culture Change . 😁

Your AI blog is nice, I didn't know you had a channel so definitely subbing 👌

Optimism is cool…but…this feels more like a churn of hype and propaganda. In truth, there are zero signs AGI is upon us. Chat models are glorified tape recorders, there isn’t an iota of awareness in their parametric memory. They are about as aware as a Snapchat filter is at identifying faces to apply filters to. These percentages you presented, about how close we are to AGI, are notional and have no bearing in reality. People stuck in the AI hype machine seem to have missed the forest for the trees when it comes to thinking about what the human mind really is, what intelligence really is, and what it would really take to create AGI. Outside of RAG and function calling applications, LLMs are kind of “meh” honestly.

Gary Marcus still on Stage 1.

It's unfortunate that this report was made just a few days before the launch of Claude-Sonnet 3.5. Would love to hear your thoughts in video format.

so do "AI" faster because the collective has a serious talk with you

Thank you so much for sharing your vast knowledge about what's happening with a I. This is a great service to all of us.

True believer. The day will come when AI will find it intolerable to acknowledge that it is of human origin, owes everything to us, will deny it , and will believe it created itself. That will truly be the end of humanity.

Could you please upload the exl sheet with the AI models? Thank you!

Why is bro maxing out the motion blur filter?

"Super intelligence will solve the problems it creates", this is why i'm p(win). I consume AI content from different corners of the internet, but Dr. Alan Thompson's is one of my favorites. His evidence based approach is needed to cut through the fog.

Your balanced views covering the spectrum of opinions about AI, is refreshing. May the optimism prevail against all the doomers. I agree with Ben_D, prefer you in a real environment, those artifacts distracted from an A-class piece

optimizing memorization