/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Vishvanand Subramanian

@vishvanands
27 posts
2025-03-14
trying hard to steelman this position from openai but unless it's possible to hide malware in the model weights, what exactly is the risk of downloading open source models from china?  apart from the obvious risk to openai trying to charge 2000 a month for a product deepseek will put out for free
2025-03-14 View on X
TechCrunch

OpenAI calls DeepSeek “state-controlled” and recommends that the US ban “PRC-produced equipment and models that violate user privacy and create security risks”

https://techcrunch.com/... Threads: Vishvanand Subramanian / @vishvanands : trying hard to steelman this position from openai but unless it's possible to hide malware in the model ...

trying hard to steelman this position from openai but unless it's possible to hide malware in the model weights, what exactly is the risk of downloading open source models from china?  apart from the obvious risk to openai trying to charge 2000 a month for a product deepseek will put out for free
2025-03-14 View on X
TechCrunch

Google releases its proposal for the White House's AI Action Plan, arguing that “fair use and text-and-data mining exceptions” are “critical” to AI development

Google, following on the heels of OpenAI, published a policy proposal in response to the Trump Administration's call for a national “AI Action Plan.”

2025-01-26
thinking about how openai arranged a 500 billion dollar investment for future models while a small lab spent 5 million as a side project to match the performance of their flagship model..
2025-01-26 View on X
MIT Technology Review

Rather than weakening China's AI capabilities, US sanctions appear to be driving startups like DeepSeek to innovate by prioritizing efficiency and collaboration

The AI community is abuzz over DeepSeek R1, a new open-source reasoning model.  —  The model was developed by the Chinese AI startup DeepSeek …

thinking about how openai arranged a 500 billion dollar investment for future models while a small lab spent 5 million as a side project to match the performance of their flagship model..
2025-01-26 View on X
VentureBeat

Yann LeCun says DeepSeek “profited from open research and open source” like Meta's Llama and is proof that open source models are surpassing proprietary ones

“Marc Andreessen, a co-inventor of the pioneering Mosaic web browser, co-founder of the Netscape browser company and current general partner at the famed Andreessen Horowitz (a16z)...

2025-01-25
thinking about how openai arranged a 500 billion dollar investment for future models while a small lab spent 5 million as a side project to match the performance of their flagship model..
2025-01-25 View on X
VentureBeat

Yann LeCun says DeepSeek “profited from open research and open source” like Meta's Llama and is proof that open source models are surpassing proprietary ones

If you hadn't heard, there's a new AI star in town: DeepSeek, the subsidiary of Hong Kong-based quantitative analysis …

2024-08-30
["Is Enthusiasm For Meta's Llama Slowing Down?"]  Wonder if @zuck's post is a response to this @theinformation headline.  If so, great to hear that open source models are killing it 🚀
2024-08-30 View on X
Reuters

Meta says its Llama models were downloaded almost 350M times, are used by AT&T and others, and usage via cloud providers more than doubled from May to July 2024

we just published a bunch of updates on the adoption we're seeing.  And yes, we have a lot more work to do on dev tools and resources which we're bringing online as quickly as we c...

2024-07-14
The slow exodus of people who regularly engage on threads is the only outcome if Meta chooses to prioritize big influencers over the actual community here.
2024-07-14 View on X
augment

Threads needs to foster community builders rather than get celebrity creators from platforms like Instagram, who tend not to do well on a microblogging platform

It's been just over a year since Meta launched Threads, a microblogging platform built upon Instagram's foundation. Threads: @laurenlopes.author , @saramontourlewis , @thetechsavvy...

2024-07-04
Threads hits 175M monthly active users on its 1 year birthday.  Meanwhile, it's trending on X as they wonder whether anyone uses Threads.
2024-07-04 View on X
The Verge

Mark Zuckerberg says Threads has 175M+ MAUs, up from 150M in April, ahead of its one-year anniversary on July 5; source: Meta may turn on ads in Threads in 2025

Threads hits 175M monthly active users on its 1 year birthday.  Meanwhile, it's trending on X as they wonder whether anyone uses Threads.
2024-07-04 View on X
Platformer

Q&A with Adam Mosseri on Threads' first year, not leaning into political content, driving engagement, focusing on real-time content, the fediverse, and more

and even Threads is warming to it Leo Grieco / Proactiveinvestors UK : Meta's Threads hits milestone but lags social media giants Sean Tilley / We Distribute : A Beginner's Guide t...

2024-06-29
Amazon has hired Adept AI's top executive and founding team in a move that mirrors Microsoft and Inflection a few months ago.  Now we'll find out if the EU will make good on its warnings to investigate companies that poach founding teams rather than acquire startups.
2024-06-29 View on X
GeekWire

Amazon hires the CEO and co-founders of Adept, which builds AI agents that automate enterprise workflows, to join its AGI team and will use some of Adept's tech

Amazon is amping up its AI efforts by hiring executives from Adept, a San Francisco-based startup building “agents” that automate enterprise workflows.

2024-06-26
OpenAI just wrote a long post delaying their Voice mode update to July.  Between this and Anthropic's blockbuster Sonnet 3.5 launch, OpenAI is struggling to maintain its crown.
2024-06-26 View on X
9to5Mac

OpenAI makes ChatGPT for macOS available to all users after first rolling it out to Plus subscribers; the app can be launched from anywhere using Option + Space

Over the last month, OpenAI has gradually been rolling out its first-ever ChatGPT app for Mac with Plus subscribers.

OpenAI just wrote a long post delaying their Voice mode update to July.  Between this and Anthropic's blockbuster Sonnet 3.5 launch, OpenAI is struggling to maintain its crown.
2024-06-26 View on X
Bloomberg

OpenAI delays the release of its advanced Voice Mode for ChatGPT by a month to address safety issues and plans to roll out the feature to Plus users in the fall

no word from Scarlett Johansson yet Livemint : ChatGPT Voice Mode release pushed to July: Here's why Jay Bonggolto / Android Police : OpenAI pushes back ChatGPT's new ‘Voice Mode’ ...

2024-06-23
I once interviewed at OpenAI's first acquisition: Rockset.  Great retrieval infra tech and the team had some engineering legends from Google and Meta, so I learned a lot from just the interview.  Not sure if the company really took off, but OpenAI is probably going to integrate their tech for enterprise RAG while getting some great infra engineering talent too.
2024-06-23 View on X
Bloomberg

OpenAI buys Rockset, an enterprise search and analytics startup that raised $105M, integrating both the tech and staff of an acquisition for the first time

OpenAI acquires Rockset To Power the AI-Driven Future OpenAI : OpenAI acquires Rockset Markus Kasanmascheff / WinBuzzer : OpenAI Acquires Rockset to Boost Data Analytics Reuters : ...

2024-04-19
Given Llama-3 400B is on par with the current best model Claude-3 Opus and it's still training, we can soon expect to see the dream of open source realized: The best model in the world is now free and open source.
2024-04-19 View on X
The Verge

Meta details Llama 3: 8B- and 70B-parameter models, a focus on reducing false refusals, and an upcoming model trained on 15T+ tokens that has 400B+ parameters

What To Know About ‘Llama 3’ Model Marcus Gopolang Moloko / Memeburn : Meta AI with built in Llama 3 is on WhatsApp in South Africa Hamsat Abdurasheed / News.ng : Meta releases Lla...

Given Llama-3 400B is on par with the current best model Claude-3 Opus and it's still training, we can soon expect to see the dream of open source realized: The best model in the world is now free and open source.
2024-04-19 View on X
The Verge

Meta claims both Llama 3 models beat similarly sized models like Gemini, Mistral, and Claude 3 on certain benchmarks; humans marked Llama 3 higher than GPT-3.5

2023-12-26
NYT claims private information can be leaked during fine tuning of GPT 3.5.  Initially, I was alarmed that researchers were able to bypass basic safeguards in LLMs against revealing private information in the training data.  But perhaps the model was just guessing firstname.lastname corporate emails of popular journalists and happened to get a few right..
2023-12-26 View on X
New York Times

Researchers extract business and personal email addresses for 30+ NYT staff by fine tuning GPT-3.5 Turbo, bypassing the safeguards for privacy-related queries

Jeremy White / New York Times :

2023-12-25
NYT claims private information can be leaked during fine tuning of GPT 3.5.  Initially, I was alarmed that researchers were able to bypass basic safeguards in LLMs against revealing private information in the training data.  But perhaps the model was just guessing firstname.lastname corporate emails of popular journalists and happened to get a few right..
2023-12-25 View on X
New York Times

Researchers extract business and personal email addresses for 30+ NYT staff by fine tuning GPT-3.5 Turbo, bypassing the safeguards for privacy-related queries

This is a visualization of a large email database from the Enron Corporation, which is often used to train artificial intelligence systems, like ChatGPT. Threads: @carnage4life and...

2023-11-19
There was definitely some tension between Satya Nadella and Sam Altman at the OpenAI event.  Looks like it boiled over and Sam walks away with 0 OpenAI stock despite spearheading the most influential product launch in a decade.
2023-11-19 View on X
The Verge

Source: OpenAI's board had agreed to resign and let Sam Altman and Greg Brockman return, but has since waffled; Altman would want significant governance changes

The OpenAI board is in discussions with Sam Altman to return to CEO, according to multiple people familiar with the matter.

2023-11-18
There was definitely some tension between Satya Nadella and Sam Altman at the OpenAI event.  Looks like it boiled over and Sam walks away with 0 OpenAI stock despite spearheading the most influential product launch in a decade.
2023-11-18 View on X
Axios

Source: Microsoft learned that OpenAI was ousting CEO Sam Altman just a minute before the news was published

Microsoft, which has invested billions in OpenAI, learned that OpenAI was ousting CEO Sam Altman just a minute before the news was shared with the world, according to a person fami...

There was definitely some tension between Satya Nadella and Sam Altman at the OpenAI event.  Looks like it boiled over and Sam walks away with 0 OpenAI stock despite spearheading the most influential product launch in a decade.
2023-11-18 View on X
The Official Microsoft Blog

Satya Nadella says Microsoft has a “long-term agreement with OpenAI with full access to everything” it needs and remains “committed” to their partnership

its biggest outside source of capital and computing power—was stable following the sudden firing of CEO Sam Altman. https://www.theinformation.com/ ... @erinkwoo @amir Dan Ives / @...