Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
ruby.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
If you are interested in the Ruby programming language, come join us! Tell us about yourself when signing up. If you just want to join Mastodon, another server will be a better place for you.

Administered by:

Server stats:

1.1K
active users

ruby.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#Limitations

0 posts0 participants0 posts today
Public
N-gated Hacker News

🧠🤖 Behold, the groundbreaking revelation that large language models aren't the omnipotent deities of #reasoning we all assumed. A study from the prestigious #Santa #Fe #Institute, nestled between their endless newsletters and fellowships, confirms what your toaster could have told you: #AI still struggles with basic #logic. 🚀🔍
santafe.edu/news-center/news/s #limitations #technology #news #HackerNews #ngated

santafe.eduStudy: Large language models still lack general reasoning skills | Santa Fe InstituteUsing a variety of analogy puzzles, SFI researchers have shown that the reasoning abilities of OpenAI’s GPT-4 model fall short when faced with small changes. Recognizing the limitations of these tools is critical to knowing when and to what extent they can be trusted.
Public
Andro Abhi

Official Silt Strider Trailer #Morrowind #Roleplay #Server

youtube.com/watch?v=nqh9B7A-y_…

Silt Strider is a #TES3MP server focused on realistic, grounded and lore-friendly roleplay, not just with #players, but with its #NPCs as well. With access to our #AI #Interaction #System, you can speak and roleplay with both players and NPCs seamlessly.

The community originally came from our previous project, #Tales from #Nirn, a #project creating a server on #Skyrim Together, which ran for the span of about 2 years, before closing down due to #limitations on #scripting and server capabilities.

With the help of Wabbajack to create an auto-installing mod-list, as well as the robustness and versatility of TES3MP, Silt Strider is a great place to make friends and roleplay in a world that truly honors its setting.

Music: Nerevar Rising by Fredrik Jonasson

YouTubeOfficial Silt Strider Trailer (Morrowind Roleplay Server)By Silt Strider (Tales from Nirn)
Public
Akire Bubar

I think in order to feel #DisabilityPride, a person needs to have a workable answer to this question: What does it mean to believe in oneself _and_ accept one's own limitations?

How do YOU relate to the things you're bad at but have to do anyway? Or the things you feel judged for not being able to do at all?

See the (free!) #Patreon post for this:
patreon.com/posts/self-worth-v

A short journal comic in black ink and colored pencil on white paper. At the top, a narration box outlined in black and shades of green and yellow reads, "Recognizing I can't do something feels like admitting FAILURE without even TRYING." Below this is a drawing of a woman from the torso up. She has light brown hair and is wearing glasses and holding her chin thoughtfully in one hand. A colorful speech bubble above and to the left of her reads, "What does it mean to Believe in Oneself and accept one's own limitations? What does that Look like? How does it work? How do I learn this? Can I learn this?" The comic is signed "Akire Bubar" by the artist in the bottom left corner.
#comics#JournalComic#GraphicMedicine
Public
卡拉今天看了什麼

Goldman Sachs: AI Is Overhyped, Wildly Expensive, and Unreliable

Link
📌Summary:
Goldman Sachs, one of the largest investment banks in the world, published a research paper expressing doubts about the economic viability of generative AI. The paper notes that there is "little to show for" the huge spending on generative AI infrastructure and questions whether this spending will ever pay off in terms of AI benefits and returns. The paper suggests that AI optimism is driving large growth in stocks like Nvidia and other S&P 500 companies, but argues that the stock price gains are based on the assumption that generative AI will lead to higher productivity and efficiency. The paper also suggests that companies are acting like generative AI is going to change the world, but that the technology is currently deeply unreliable and may not change much of anything at all.

🎯Key Points:
- Goldman Sachs questions the economic viability of generative AI.
- The bank notes that there is "little to show for" the huge spending on generative AI infrastructure and wonders whether it will ever pay off.
- The paper is based on a series of interviews with Goldman Sachs economists and researchers, MIT professor Daron Acemoglu, and infrastructure experts.
- Goldman Sachs argues that AI optimism is driving large growth in stocks like Nvidia and other S&P 500 companies.
- The bank suggests that the stock price gains are based on the assumption that generative AI will lead to higher productivity and efficiency.
- The paper argues that generative AI is currently deeply unreliable and may not change much of anything at all.
- The bank expresses skepticism about both the cost of generative AI and its ultimate transformative potential.

🔖Keywords:
#GenerativeAI #EconomicViability #GoldmanSachs #Doubts #Unreliable #Hype #Expensive #Investment #S&P500 #Nvidia #Productivity #Automation #Efficiency #Bubbles #LongTermReturnForecastingFramework #Skepticism #Cost #TransformativePotential #Limitations #Architecture #ComplexProblems #Data #Training #Reddit #RedditData #Doubling #Limitations #Expense #VirtualReality #Metaverse #Blockchain #VentureCapital #AIIndustry #Revenue #ProcessingPower #InfrastructureCosts #DataBreaches #AT&T #Hackers #CallandTextMetadata #Customers #AT&TCustomers #Massive #Significant #DataBreaches
404 Media · Goldman Sachs: AI Is Overhyped, Wildly Expensive, and UnreliableOne of the world's largest investment banks wonders if generative AI will be worth the huge investment and hype: "will this large spend ever pay off?"
#sp500
Public
Cyber-Defence Campus

Dr. Ljiljana Dolamic's recent lunch seminar at the Cyber-Defence Campus was a captivating overview of Large Language Models (#LLMs), shedding light on their #breakthroughs, #dangers, and #limitations. Here's a glimpse into the key takeaways:

💡 Dr. Dolamic began by tracing the evolution of language modeling tasks, emphasizing the significant shift brought about by Neural Language Modeling. Despite the transformative impact of models like the Transformer architecture in tasks such as machine translation, LLM developments continues to capture the interest of NLP practitioners.

📑 One highlight of the discussion was the spotlight on #GPT-4 and its integration with user-friendly interfaces like Chat-GPT. Promoted as a solution for a myriad of tasks and bolstered by features like fine-tuning and Retrieval Augmented Generation (#RAG), these models have garnered attention for their potential to streamline processes. However, Dr. Dolamic also aptly addressed the inherent risks accompanying LLMs, including the susceptibility to hallucinations and biases.

➡ As we look back on the seminar, it serves as a timely reminder of the complexities inherent in the realm of language modeling. By acknowledging both the promises and pitfalls of LLMs, participants left equipped with a deeper understanding of this evolving landscape.

Public
Annabelle

You know, while it is on my mind. #ElonMusk and what he has left for a #Twitter team. Is doing one #shitty job at trying to sell #TwitterBlue to people.

1) With the cutting of #Dms and making it #verified only for some users.

2) The #limitations for #Tweeetdeck and how #RateLimits are fucking that over hard. Plus the fact they #removed it from #Legacy instantly after #announcing that Tweetdeck was going to blue exclusively. That is terrible #business and pushes people away. Like me.

ExploreLive feeds
About