A powerful new AI chatbot called “gpt2-chatbot” appears on LMSYS Chat and has sparked speculation whether it could be OpenAI’s unreleased GPT-5 or a supercharged GPT-2.
Vitalik Buterin: AI may surpass humans as the ‘apex species’
“Even Mars may not be safe” if superintelligent AI turns against humanity, warns Ethereum co-founder Vitalik Buterin.
‘107,000 GPUs on the waitlist’ — io.net beta launch attracts data centers, GPU clusters
Io.net’s recently developed decentralized physical infrastructure network has moved into its beta phase, allowing GPU computing providers to plug into the platform.
Universal Music releases Beatles ‘last song’ with help from AI
The final Beatles song “Now and Then” has been released and made possible with a little help from AI to produce John Lennon’s vocal track.
Jed McCaleb-backed nonprofit will provide easier access to AI computing capacity
Voltage Park will lease access to 24,000 clustered NVIDIA GPUs by the hour or month to help small startups and researchers model machine learning.
Startup demos upcoming decentralized GPU infrastructure network to OpenAI, Uber
io.net has built a decentralized physical infrastructure network that will source GPU computing power for AI and machine learning.
Researchers find LLMs like ChatGPT output sensitive data even after it’s been ‘deleted’
According to the scientists, there’s no universal method by which data can be deleted from a pretrained large language model.
Microsoft to form nuclear power team to support AI: Report
Microsoft is forming a new team of professionals to advance its artificial intelligence plans with Small Modular Reactors and microreactors.
Google and Microsoft-backed AI firm AlphaSense raises $150M at $2.5B valuation
AlphaSense’s client list now includes most of the S&P 500 and nearly every firm listed in the Dow 50.
Google launches Digital Futures Project with $20M in grants to support ‘responsible AI’
The launch comes ahead of a series of AI forums to be hosted by U.S. Senate majority leader Chuck Schumer.
Scientists created ‘OpinionGPT’ to explore explicit human bias — and you can test it for yourself
Due to the nature of the model’s tuning data, it’s unclear whether this system is actually capable of generating outputs showing real-world bias.
Oxford scientists develop GPU-accelerated limit order book sim to teach AI how to trade
The first-of-its-kind architecture gives up to a 7x speedup over traditional training methods.
Anthropic cracks open the black box to see how AI comes up with the stuff it says
The researchers were able to trace outputs to neural network nodes and show influence patterns through statistical analysis.
ChatGPT and Claude are ‘becoming capable of tackling real-world missions,’ say scientists
The scientists developed a tool called “AgentBench” to benchmark LLM models as agents.
Zoom updates terms after backlash, won’t train AI without consent
Many online said they were halting the use of Zoom over terms that seemingly allowed the platform to scrape user data to train AI models.
New research shows how brain-like computers could revolutionize blockchain and AI
A CMOS-compatible neuromorphic computing chip could be on the horizon thanks to breakthrough research out of Technische Universität Dresden.
AI21 Labs debuts anti-hallucination feature for GPT chatbots
Contextual Answers is designed for enterprise but could have far-reaching implications for the generative AI sector.
Apple has its own GPT AI system but no stated plans for public release: Report
The Cupertino, California-based company reportedly developed an internal GPT system on Google infrastructure for employees to tinker with.
Elon Musk’s new AI startup is as ambitious as it is doomed
The public perception surrounding AI’s abilities is no match for the laws of physics.
Anthropic launches Claude 2 amid continuing AI hullabaloo
The new model demonstrates measurable improvements across numerous categories, including near-instant query response times and the ability to parse inputs up to 100K tokens in size.