Model

Semron Seeks to Substitute ‘Memcapacitors’ for Chip Transistors

Gettyimages 1425576452
The electric field approach minimizes the movement of electrons at the chip level, reducing energy usage — and heat. TOPS/W is a bit of a vague metric, but the takeaway is that memcapacitors can lead to dramatic energy consumption reductions while training AI models. Now, it’s early days for Semron, which Kirschen says is in the “pre-product” stage and has “negligible” revenue to show for it. EnCharge, like Semron, is designing computer chips that use capacitors rather than transistors, but using a different substrate architecture. Semron will be a key element in solving this problem by providing a revolutionary new chip that is inherently specialized on computing AI models.

AI Startup Founded by Ola’s Creator Reaches Unicorn Status with $50M in Funding

Gettyimages 625022782 2
Krutrim, an AI startup founded by Ola founder Bhavish Aggarwal, said it has raised a funding round that values it at $1 billion. It’s also the first Indian AI startup to become a unicorn, it said. Matrix Partners India — which has also backed Aggarwal’s other two startups, ride-hailing firm Ola and EV startup Ola Electric — led the $50 million “first round” in Krutrim, it said Friday. The startup plans to launch a conversational AI assistant that understands and speaks multiple Indian languages, the startup said. The startup plans to make a beta version of its eponymous chatbot available to consumers next month and follow it with rolling out APIs to developers and enterprises.

“OpenAI Reduces Prices and Repairs ‘Uncooperative’ GPT-4 Model”

Gettyimages 1647521480 E1694685235777 2
OpenAI is always making slight adjustments to its models and pricing, and today brings just such an occasion. It’s also a popular API, being lower cost and faster than GPT-4 on a lot of tasks. Hence the steady ratcheting down of prices — though it’s also a natural result of streamlining the models and improving their infrastructure. GPT-3.5 Turbo also gets a new model version, 0125 (i.e. And the company also released a new version of its free moderation API — which identifies potentially harmful text.

“Global Debut: OnePlus 12 Series Unveiled by OnePlus”

Oneplus 12
OnePlus launched the OnePlus 12 series globally today, bringing an R-series device to North America and Europe for the first time. The company also launched the $99 OnePlus Buds 3 wireless buds apart from the OnePlus 12 (starting at $799) and the OnePlus 12R (starting at $499). Last year, OnePlus launched just the $699 OnePlus 11 in the U.S., while the OnePlus 11R was limited to India and China. The OnePlus 12 boasts a 6.82-inch 1440p display with the company claiming it reaches up to 4,500 nits of brightness. OnePlus is opting for a triple camera setup in the OnePlus 12 with a 50-megapixel Sony LYT-808 sensor, which has a f/1.6 aperture.

Spain Challenges Legality of Meta’s Use of Content Moderating Subcontractors

Gettyimages 1245156419
The Barcelona court accepted that the mental problems suffered by the worker are not a common illness but a work accident, per the newspaper. Meta’s subcontractor had processed his absence from work as common ailment and sought to deny responsibility for any psychological harms suffered from reviewing violent content uploaded to Facebook and Instagram. Meta also noted it provides technical solutions to subcontractors which are intended to enable content reviewers to limit their exposure to graphic material they are being asked to moderate as much as possible. In the article the newspaper quotes a worker describing the support provided by their employer, and Meta’s subcontractor, as “very insufficient”. Legal rulings that impose requirements on third party content reviewers to take care of workers’ mental health could put limits on the model, however.

“Revolutionizing Graphic Design: How Recraft’s Unique Approach Addresses the AI Debate”

Recraft
The latest is Recraft, an AI graphic design generator aimed at professionals, which has raised a $12 million Series A round led by Khosla Ventures in Silicon Valley, together with former GitHub CEO, Nat Friedman. Admittedly there are now myriad Generative AI design tools out there, such as Jasper, Adobe Sensei, Let’s Enhance, and many others. And then you generate images using this style of reference. But to date, many of the generative AI design solutions have been targeting consumers, rather than professionals that require high degrees of control. Recraft delivers on professional workflows such as vector images, style controls and end-to-end content production, all powered by in-house built foundation models.”

” “Unlock the Full Potential of the Samsung Galaxy S24 with Google Gemini-Powered AI Enhancements”

Google Bard Gemini V2
Gemini, Google’s cutting-edge generative AI model family, is coming to Samsung’s new flagship smartphone — the Galaxy S24. Google and Samsung made the announcement to coincide with the Galaxy S24’s unveiling this afternoon. Beyond Gemini, the Galaxy S24 will benefit from Google’s Imagen 2 text-to-image model, which will underpin photo editing features in the Galaxy S24 Gallery app. Neither company said exactly how Ultra might be used on the Galaxy S24, however — or any other Samsung devices for that matter. Mum’s the word on whether Galaxy S24 users can expect the same Assistant and Bard upgrades; hopefully we’ll learn more during today’s Galaxy S24 press conference.

Anthropologists discover deceptive capabilities of trained AI models

Gettyimages 1548038240 1
A recent study co-authored by researchers at Anthropic, the well-funded AI startup, investigated whether models can be trained to deceive, like injecting exploits into otherwise secure computer code. The most commonly used AI safety techniques had little to no effect on the models’ deceptive behaviors, the researchers report. Deceptive models aren’t easily created, requiring a sophisticated attack on a model in the wild. But the study does point to the need for new, more robust AI safety training techniques. “Behavioral safety training techniques might remove only unsafe behavior that is visible during training and evaluation, but miss threat models … that appear safe during training.

Greetings from Vegas: My Experience at CES

A Postcard From Las Vegas By Haje
This week Devin, Kirsten and I have been running around in Las Vegas for CES to figure out what’s going on in the world of tech. Siri in particular could do with getting its act together, before it’s completely left in the dust by its competitors. Biggest news from CESSamsung’s Ballie, the spherical home robot first seen at CES 2020, has rolled back into the spotlight at CES 2024 with some trend-forward AI enhancements. Kyle reports how AI tools can easily manipulate ID images, creating deepfakes convincing enough to pass KYC tests. Plus, Alexa’s got a new AI model making it more opinionated and emotional.

Alternative: Struggling to Choose Between Airbnb or a Traditional Hotel? Discover Overmoon’s Hybrid Vacation Rental Concept

Overmoon Home 04 1
Historically, vacation rental companies have managed homes for homeowners. Overmoon is a three-year-old vacation rental startup with a different model that essentially cuts out the middle man. But Fraiman doesn’t believe that so many proptech companies are failing because of high interest rates. “I think the inability to raise capital is the real reason and interest rates are a contributing factor,” he said. Plus, higher interest rates means fewer buyers out there, which means less competition.