CybAI news - - Page 2955

25 May

Gamma

stp2y0 CommentsNews

Briefings highlight generational AI scaleups, startups, and projects. I was fortunate to have interviewed co-founder & CEO Grant Lee. Read on to learn more about the origins of Gamma, how it is defining a new UX with AI, and why it is a generational company. You can also read this article as a Gamma doc here: link. Gamma is the canvas for modern work documents. Users can create visually engaging, interactive documents easily by themselves or with the help of an AI assistant.Why Gamma is a generational companyEfficient momentum: Gamma is the most popular product in its category while having…

25 May

Loft Orbital and SkyServe partner on AI-powered Earth observation application

stp2y0 CommentsChat-GPT

WASHINGTON — Loft Orbital is partnering with SkyServe to leverage Earth observation and edge computing capabilities on a newly launched satellite. The two companies announced March 19 that SkyServe will use Loft Orbital’s YAM-6 spacecraft, launched on the Transporter-10 rideshare mission March 4, to demonstrate artificial intelligence capabilities by analyzing optical and hyperspectral imagery the satellite collects. For what the companies call Mission Denali, SkyServe, an Indian startup, will install its SkyServe STORM platform on the spacecraft. Customers can then deploy AI models on the platform to perform analysis of imagery the satellite collects in real time. “We are proud…

25 May

Health And Biotech Startups Now Get The Majority Of US Series A Funding

stp2y0 CommentsRAG models

This year is shaping up as the first we’ve seen in which biotech and healthcare startups receive a majority of U.S. Series A commitments. So far in 2024, biotech and health companies have pulled in around $5.6 billion across 110 Series A rounds, per Crunchbase data. That accounts for 53% of all funding at the Series A stage, which is a closely watched barometer for the startup ecosystem. The biotech sector’s comparatively strong showing comes as overall Series A dealmaking looks on track to come in a bit above last year’s totals. However, funding remains down from 2022 and much,…

25 May

Onehouse Breaks Data Catalog Lock-In with More Openness

stp2y0 CommentsViral News

(Majcot/Shutterstock) Onehouse, the Apache Hudi-backer that bills itself as the most open data platform in the world, further opened up its platform today with the launch of a data catalog synchronization feature that streamlines user access to data residing in major cloud platforms. The feature complements the company’s investment in developing XTable, an open-source offering that delivers read-write interoperability among Hudi, Delta, and Apache Iceberg table formats. The advent of open table formats like Hudi, Delta, and Iceberg revolutionized data openness by enabling multiple query engines access the same piece of data without fear of data corruption. As the key…

25 May

Revolutionary weight-loss drugs like Wegovy come with a catch

stp2y0 CommentsNews

This article is an installment of Future Explored, a weekly guide to world-changing technology. You can get stories like this one straight to your inbox every week by subscribing here.Anti-obesity drugs cause people to lose more than just fat.More than 73% of American adults are overweight, according to the CDC. This puts them at increased risk of death and many serious health issues, but losing weight and keeping it off through diet changes and exercise — the standard approach — is notoriously difficult.That made the FDA’s 2021 approval of Novo Nordisk’s semaglutide (Wegovy) as an obesity treatment seem like something of…

25 May

Research Papers in Oct 2023: A Potential Successor to RLHF for Efficient LLM Alignment and the Resurgence of CNNs

stp2y0 CommentsAI

From Vision Transformers to innovative large language model finetuning techniques, the AI community has been very active with lots of interesting research this past month.Here's a snapshot of the highlights I am covering in this article:In the paper ConvNets Match Vision Transformers at Scale, Smith et al. invest significant computational resources to conduct a thorough comparison between Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), challenging the prevailing notion that ViTs outperform CNNs in image classification tasks. The Mistral 7B paper introduces a compact yet powerful language model that, despite its relatively modest size of 7 billion tokens, outperforms its larger…

25 May

If Scarlett Johansson can’t bring the AI firms to heel, what hope for the rest of us? | John Naughton

stp2y0 CommentsChat-GPT

On Monday 13 May, OpenAI livestreamed an event to launch a fancy new product – a large language model (LLM) dubbed GPT-4o – that the company’s chief technology officer, Mira Murati, claimed to be more user-friendly and faster than boring ol’ ChatGPT. It was also more versatile, and multimodal, which is tech-speak for being able to interact in voice, text and vision. Key features of the new model, we were told, were that you could interrupt it in mid-sentence, that it had very low latency (delay in responding) and that it was sensitive to the user’s emotions.Viewers were then treated…

25 May

ArXiv Pre-Print “Evaluating AI Systems under Uncertain Ground Truth: a Case Study in Dermatology” • David Stutz

stp2y0 CommentsGenAI

ArXiv Pre-Print “Evaluating AI Systems under Uncertain Ground Truth: a Case Study in Dermatology” In supervised machine learning, we usually assume access to ground truth label for evaluation. In many applications, however, these ground truth labels are derived from expert opinions. Disagreement among these experts is typically ignored using simple majority voting or averaging. Unfortunately, this can have severe consequences by over-estimating performance or mis-guiding model selection. In our work presented in this article, we tackle this problem by introducing a statistical framework for aggregating expert opinions. Abstract For safety, AI systems in health undergo thorough evaluations before deployment, validating…

25 May

Data Machina #246

stp2y0 CommentsViral News

New Trends in Vision-Language Models (VLMs.) The evolution of VLMs in recent months has been pretty impressive. Today VLMs exhibit some amazing capabilities. See the two links below on what VLMs can do and how they work:But still VLMs are facing some challenges for example in terms of: multimodal training datasets, resolution, long-form modality, vision-language integration, and concept understanding. Somewhat along those lines, I see 5 trends happening in VLMs: 1) VLMs run on local environment 2) Emerging VLM videoagents 3) Unified structure learning for VLMs 4) Personalisation of VLMs and 5) Fixing the VLM resolution curse. Let’s see…VLMs on…

25 May

The AI Revolution Will Not Be Monopolized: Behind the scenes

stp2y0 CommentsNews

FabNER Claude 2 accuracy on # of examples 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 20 examples F-Score Speed (words/s) GPT-3.5 1 78.6 < 100 GPT-4 1 83.5 < 100 spaCy 91.6 4,000 Flair 93.1 1,000 SOTA 2023 2 94.6 1,000 SOTA 2003 3 88.8 > 20,000 1. Ashok and Lipton (2023), 2. Wang et al. (2021), 3. Florian et al. (2003) SOTA on few- shot prompting RoBERTa-base * * EXPERIMENTS * * * EXPERIMENTS * * * EXPERIMENTS * * * CoNLL 2003: Named Entity Recognition Source link lol