After announcing Gemini, a project by DeepMind aiming to surpass artificial intelligence models like ChatGPT yesterday, Google has now updated its privacy policy and is suggesting that it will only be using publicly available data to train its AI models. In other news, there is a growing number of tech layoffs due to the rise of AI which has concerned many researchers who believed tech roles would largely remain safe initially. This and more in todays AI roundup. Let us take a closer look.

A report by Gizmodo has revealed that Google has updated its privacy policy and now suggests that it will use any data that is publicly available (can be read by Google) to train its AI models.

Google uses the information to improve our services and to develop new products, features, and technologies that benefit our users and the public. For example, we use publicly available information to help train Googles AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities, mentions the new policy.

The most important part of this newly updated privacy policy is that earlier Google said that the data will be used for its language models, which have now been replaced with AI models. This raises serious concerns about the privacy of individuals who post things online, as now it is not just about who has access to the data but the users do not even control how the data can be used.

It was speculated that this was one of the reasons why both Reddit and Twitter made drastic policy changes to keep AI data harvesting at bay.

A new report by CNN claims that there is a growing number of layoffs occurring in the tech sector, a majority of which are linked to AI. Many employees have been fired and hiring has been frozen as companies figure out which roles can be taken over by AI.

Highlighting such an instance, the report mentioned IBM CEO Arvind Krishnas statements from an interview with Bloomberg where he mentioned that the company was going to stop its hiring to understand where AIs role can be more impactful.

Meesho and the Vision and AI Lab (VAL) of the Indian Institute of Science (IISc) will be collaborating together to conduct research into generative AI, a report by Business Standard said. The two have also signed a memorandum of understanding (MoU) of one year.

As per the MoU, Meesho will let its data scientists work with researchers from IISc to focus on multimodal representation learning and generative AI capabilities. Meesho believes this collaboration will result in the expansion of the e-retail sector by harnessing the abilities of AI.

The universities in the UK have prepared guiding principles around generative AI. The central focus is to provide education and awareness on such technologies as institutes struggle to adapt teaching and assessment methods to adjust to the rise of AI, a report by The Guardian stated.

Unlike the previous sentiments that AI should be banned from educational institutions, the new guidelines emphasize the need to learn and adapt to this growing technology to tap into its potential while also making the students informed about the risks of plagiarism, biases, and inaccuracies of AI.

An Instagram page by the username ai.magine_ has shared a series of photos showcasing characters from the popular sitcom Friends being reimagined in Indian ethnic dresses. The post showcases Chandler and Monica tying the knot at an Indian wedding. It also shows Joey attending the wedding wearing an Indian sherwani.

More here:
5 things about AI you may have missed today: Google to use public info to train AI, tech layoffs, more - HT Tech

(Ole.CNX/Shutterstock)

Palo Alto-based Inflection AI, an OpenAI competitor, announced it has raised $1.3 billion in a funding round led by Microsoft, Reid Hoffman, Bill Gates, Eric Schmidt, and new investor Nvidia for a total of $1.525 billion raised.

This latest round places the companys valuation around $4 billion, according to a Reuters report.

Inflection AI claims to be building the largest AI cluster in the world along with partners CoreWeave and Nvidia. When completed, the system will be comprised of 22,000 Nvidia H100 Tensor Core GPUs, a release stated.

The deployment of 22,000 NVIDIA H100 GPUs in one cluster is truly unprecedented and will support training and deployment of a new generation of large-scale AI models. Combined, the cluster develops a staggering 22 exaFLOPS in the 16-bit precision mode, and even more if lower precision is utilized, the company said.

The current system contains over 3,500 H100s and recently completed the reference training task of open source benchmark MLPerf in 11 minutes. Inflection AI collaborated with Nvidia and CoreWeave to run the MLPerf tests and fine-tune and optimize the cluster.

The company has developed a large language model, Inflection-1, that enables interaction with Pi, a personal AI chatbot. The company says the new funds will support its continued work of building and supporting Pi. Inflection AI describes Pi as a new class of AI designed to be a kind and supportive companion offering text and voice conversations, friendly advice, and concise information in a natural, flowing style.

Pi, which stands for personal intelligence, is marketed as an AI focused on prioritizing the interests of people both in its functionality and monetization. Imagine your personal AI companion with the single mission of making you happier, healthier, and more productive, wrote Mustafa Suleyman, CEO and co-founder of Inflection AI in a blog post.

Instead of the big tech companies prioritizing advertisers and content creators, Inflection AI is trying a different approach, Suleyman says.

We dont have all the answers, but we are setting out to develop a personal intelligence that really does work for you, thats in your corner, always on your team. Our mission is to firmly align your AI with you, and your interests, above all else. It means designing an AI that helps you articulate your intentions, organize your life and be there for you when you need it, he wrote.

Suleyman, a co-founder of Googles DeepMind, created Inflection AI in 2022 along with LinkedIn co-founder Reid Hoffman and DeepMind alum Karn Simonyan, Inflection AIs chief scientist.

The founders have made Inflection AI a public benefit corporation (PBC), which they say gives them a legal obligation to run the company in a way that balances the financial interests of stockholders, the best interests of people materially affected by our activities, and the promotion of our specific public benefit purpose, which is to develop products and technologies that harness the power of AI to improve human well-being and productivity, whilst respecting individual freedoms, working for the common good and ensuring our products widely benefit current and future generations.

A powerful benefit of the AI revolution is the ability to use natural, conversational language to interact with supercomputers to simplify aspects of our everyday lives, said Jensen Huang, founder and CEO of Nvidia in a release. The world-class team at Inflection AI is helping to lead this groundbreaking work, deploying NVIDIA AI technology to develop, train and deploy massive generative AI models that enable amazing personal digital assistants.

We are very excited to partner with Inflection AI, a pioneering AI company with an outstanding team, to bring the power of supercomputing to cutting edge consumer products, said Michael Intrator, CEO of CoreWeave.

Related

View post:
OpenAI Rival Inflection AI Raises $1.3B to Enhance Its Pi Chatbot - EnterpriseAI

Recent progress in AI has been startling. Barely a weeks gone by without a new algorithm, application, or implication making headlines. But OpenAI, the source of much of the hype, only recently completed their flagship algorithm, GPT-4, and according to OpenAI CEO Sam Altman, its successor, GPT-5, hasnt begun training yet.

Its possible the tempo will slow down in coming months, but dont bet on it. A new AI model as capable as GPT-4, or more so, may drop sooner than later.

This week, in an interview withWill Knight, Google DeepMind CEO Demis Hassabis said their next big model, Gemini, is currently in development, a process that will take a number of months. Hassabis said Gemini will be a mashup drawing on AIs greatest hits, most notably DeepMinds AlphaGo, which employed reinforcement learning to topple a champion at Go in 2016, years before experts expected the feat.

At a high level you can think of Gemini as combining some of the strengths of AlphaGo-type systems with the amazing language capabilities of the large models, Hassabis told Wired. We also have some new innovations that are going to be pretty interesting. All told, the new algorithm should be better at planning and problem-solving, he said.

Many recent gains in AI have been thanks to ever-bigger algorithms consuming more and more data. As engineers increased the number of internal connectionsor parametersand began to train them on internet-scale data sets, model quality and capability increased like clockwork. As long as a team had the cash to buy chips and access to data, progress was nearly automatic because the structure of the algorithms, called transformers, didnt have to change much.

Then in April, Altman said the age of big AI models was over. Training costs and computing power had skyrocketed, while gains from scaling had leveled off. Well make them better in other ways, he said, but didnt elaborate on what those other ways would be.

GPT-4, and now Gemini, offer clues.

Last month, at Googles I/O developer conference, CEO Sundar Pichai announced that work on Gemini was underway. He said the company was building it from the ground up to be multimodalthat is, trained on and able to fuse multiple types of data, like images and textand designed for API integrations (think plugins). Now add in reinforcement learning and perhaps, as Knight speculates, other DeepMind specialties in robotics and neuroscience, and the next step in AI is beginning to look a bit like a high-tech quilt.

But Gemini wont be the first multimodal algorithm. Nor will it be the first to use reinforcement learning or support plugins. OpenAI has integrated all of these into GPT-4 with impressive effect.

If Gemini goes that far, and no further, it may match GPT-4. Whats interesting is whos working on the algorithm. Earlier this year, DeepMind joined forces with Google Brain. The latter invented the first transformers in 2017; the former designed AlphaGo and its successors. Mixing DeepMinds reinforcement learning expertise into large language models may yield new abilities.

In addition, Gemini may set a high-water mark in AI without a leap in size.

GPT-4 is believed to be around a trillion parameters, and according to recent rumors, it might be a mixture-of-experts model made up of eight smaller models, each a fine-tuned specialist roughly the size of GPT-3. Neither the size nor architecture has been confirmed by OpenAI, who, for the first time, did not release specs on its latest model.

Similarly, DeepMind has shown interest in making smaller models that punch above their weight class (Chinchilla), and Google has experimented with mixture-of-experts (GLaM).

Gemini may be a bit bigger or smaller than GPT-4, but likely not by much.

Still, we may never learn exactly what makes Gemini tick, as increasingly competitive companies keep the details of their models under wraps. To that end, testing advanced models for ability and controllability as theyre built will become more important, work that Hassabis suggested is also critical for safety. He also said Google might open models like Gemini to outside researchers for evaluation.

I would love to see academia have early access to these frontier models, he said.

Whether Gemini matches or exceeds GPT-4 remains to be seen. As architectures become more complicated, gains may be less automatic. Still, it seems a fusion of data and approachestext with images and other inputs, large language models with reinforcement learning models, the patching together of smaller models into a larger wholemay be what Altman had in mind when he said wed make AI better in ways other than raw size.

Hassabis was vague on an exact timeline. If he meant training wouldnt be complete for a number of months, it could be a while before Gemini launches. A trained model is no longer the end point. OpenAI spent months rigorously testing and fine-tuning GPT-4 in the raw before its ultimate release. Google may be even more cautious.

But Google DeepMind is under pressure to deliver a product that sets the bar in AI, so it wouldnt be surprising to see Gemini later this year or early next. If thats the case, and if Gemini lives up to its billingboth big question marksGoogle could, at least for the moment, reclaim the spotlight from OpenAI.

Image Credit: Hossein Nasr / Unsplash

See more here:
Here's Why Google DeepMind's Gemini Algorithm Could Be Next-Level AI - Singularity Hub

Ever since ChatGPT went viral last fall, companies have touted many ways artificial intelligence can make our lives easier. Theyve promised superhuman virtual assistants, tutors, lawyers and doctors.

What about a superhuman chemical engineer?

London-based startup Orbital Materials would like to create just that. The startup is working to apply generative AI the method behind tools like ChatGPT expressly for accelerating the development of clean energy technologies. Essentially, the idea is to make computer models powerful and sharp enough to identify the best formulas for products like sustainable jet fuel or batteries free of rare-earth minerals.

Jonathan Godwin, an Orbital Materials co-founder, imagines a system thats as accessible and effective as the software engineers use today to model designs for things like airplane wings and household furniture.

"That, historically, has just been too difficult for molecular science, he said.

ChatGPT works because its adept at predicting text - heres the next word or sentence that makes sense. For the same idea to work in chemistry, an AI system would need to predict how a new molecule would behave, not just in a lab but in the real world.

Several researchers and companies have deployed AI to hunt for newer, greener materials. Symyx Technologies, a materials discovery company formed in 1990s, wound down after a sale. More recent companies have gained traction making petrochemical alternatives and programming cells.

Still, for many materials needed to decarbonise the planet, the technology isnt there yet.

It can take decades for a new advanced material to move from discovery to the market. That timeline is way too slow for the businesses and nations looking to rapidly cut emissions as they race to meet net zero targets.

"That needs to happen in the next 10 years, or sooner, said Aaike van Vugt, co-founder of material science startup VSParticle.

AI researchers think they can help. Before launching Orbital Materials, Godwin spent three years researching advanced material discovery at DeepMind, Googles AI lab. That lab released AlphaFold, a model to predict protein structures that could speed up the search for new drugs and vaccines. That, coupled with the rapid takeoff of tools like ChatGPT, convinced him that AI would soon be capable of conquering the material world.

"What I thought would take 10 years was happening in a matter of 18 months, he said. "Things are getting better and better and better.

Godwin compares his method with Orbital Materials to AI image generators like Dall-E and Stable Diffusion. Those models are created using billions of online images so that when users type in a text prompt, a photorealistic creation appears. Orbital Materials plans to train models with loads of data on the molecular structure of materials. Type in some desired property and material say, an alloy that can withstand very high heat and the model spits out a proposed molecular formula.

In theory, this approach is effective because it can both imagine new molecules and measure how they will work, said Rafael Gomez-Bombarelli, an assistant professor at MIT, who advised Orbital Materials. (He said he is not an investor.)

Right now, many tech investors are prowling for companies that can turn a profit by improving greener material production. Thats particularly the case in Europe, where regulators are forcing manufacturers to lower carbon emissions or face stiff fines. The markets for advanced materials in sectors like renewable energy, transportation and agriculture are set to grow by tens of billions of dollars in the coming years.

Some researchers, like those at the University of Toronto, have set up "self-driving labs that pair AI systems with robots to search for new materials at unparalleled speeds. Dutch startup VSParticle makes machinery used to develop components for gas sensors and green hydrogen.

Think of it like a DNA sequencer in a genomics lab, said co-founder van Vugt, who believes his equipment can help shorten the 20-year time horizon of advanced materials to one year, and, eventually, "a couple of months. His company is currently raising investment capital.

Orbital Materials, which raised US$4.8mil (RM22.40mil) in previously undisclosed initial funding, is planning to start with turning its AI gaze toward carbon capture. The startup is working on an algorithmic model that designs molecular sieves, or tiny pellets installed within a device that can sift CO2 and other noxious chemicals from other emissions, more efficiently than current methods. (Godwin said the startup, which has several AI researchers, plans to publish peer-reviewed results on this tech soon.) Carbon capture has failed to work at scale to date, though thanks to a slew of government incentives, particularly in the US, interest in deploying the technology is rapidly ramping up.

Eventually, Godwin said Orbital Materials would like to move into areas like fuel and batteries. He imagines mirroring the business model of synthetic biology and drug discovery companies: develop the brainpower, then license out the software or novel materials to manufacturers. "Its going to take us a little bit of time to get to market," said Godwin. "But once youre there, it happens very quickly.

But getting the AI right is only half the battle. Actually making advanced materials in areas like battery and fuel production requires working with huge incumbent enterprises and messy supply chains. This can be even costlier than developing new drugs, argued MITs Gomez-Bombarelli.

"The economics and de-risking make it just way harder, he said.

Heather Redman, a managing partner with Flying Fish Partners, which backed Orbital Materials, said most tech investors chasing the shiny penny of generative AI have failed to look at its applications outside of chatbots. She acknowledged the risks of startups working in the energy sector, but believes the US$1 trillion (RM4.67 trillion) potential of markets like batteries and carbon capture are worth the investing risk.

"We love big hills as long as theres a big gigantic market and opportunity at the top, she said.

Gomez-Bombarelli is aware how big these hills can be. He helped start a similar company to Orbital Materials in 2015, called Calculario, which used AI and quantum chemistry to speed up the discovery process for a range of new materials. It didnt get enough traction and had to focus on the OLED industry.

"Maybe we didnt make our case, he said. "Or maybe the market wasnt ready.

Whether it is now is an open question. But there are encouraging signs. Computing certainly has improved. Newcomers might also have an easier time selling AI because would-be customers could more easily grasp the potential. Gomez-Bombarelli said the pitch is relatively simple: "Look at ChatGPT. We can do the same thing for chemistry. Bloomberg

Here is the original post:
DeepMind alum wants to use AI to speed the development of green materials - The Star Online