Ask QX, an Indian AI for the world

Ask QX, an Indian AI for the world


QX Lab AI launched the world’s first node-based, hybrid Gen AI platform ‘Ask QX’ in 12 Indian and over 100 global languages on February 2, 2024, in Dubai. Businessline spoke to the co-founders of QX Lab AI just before the launch to get an insight into an Indian foray in generative artificial intelligence (Gen AI) with some unique characteristics. Participating in the discussion were Tilakraj Parmar (CEO), Arjun Prasad (Chief Strategy Officer), and Tathagat Prakash (Chief Scientist). Excerpts:

While there has been a great deal of noise around Gen AI in the last 12 months following OpenAI’s launch of ChatGPT, you have come out of the blue. Can you tell us about your journey and the product.

Tilakraj Parmar: We have been working on the product for nearly eight years now. We are one of the first to build a product on a hybrid, node-based architecture. Our product is about 30 per cent GPT-based and the rest is a unique blend.

We are launching with 100+ of the planned 300 global languages and 12 Indian languages. This involved a significant effort to train and tune the programme just as the architecture and model-building.

Arjun Prasad: We are 3 Indian founders, bootstrapped and working for eight years on tech and then launching only when we were ready for the customer. Our strategy is to go for the B2C market and empower everyone with the power of AI.

Tilakraj: We have trained around 372 billion parameters which is roughly 6 trillion tokens — one token is roughly 750 words. Currently we are in the beta phase of web version that was launched in January.

We have almost 8 million users now, which is a big number for us because we were not expecting this kind of a response. We have had somewhere around 60 million prompts — questions that people ask on the platform.

So the goal is to not just focus on the English language. This is where we differ from the rest of the crowd. We want to get into the regional language market where we want people to think and write and create in their own language.

Do you allow only text-based interaction or will we have voice input as well? Also, new developments from OpenAI allows for the creation of one’s own GPT with specific inputs like spreadsheets and other structured documents. Does QX allow that?

Tilakraj: On the web, we don’t have voice input yet, but on the app that we’ve just launched, it is present. In our voice input feature you can ask questions, but the answer will come in text as of now. Voice response will be coming soon.

Arjun: The multimodal product is in alpha phase. It’s in our roadmap and we will release it at some point in the year. We are one of the very few companies that actually believe in delivering the tech-first and then talking about it. It’s in alpha phase right now and we want it to be fine-tuned to make sure it’s a beautiful product.

We will soon have the text-to-image, text-to-voice, voice-to-text options. We are also working on ensuring that the usability — our UI, UX — is user-friendly.

Given that a big promise — and I am sure, a big challenge — is the Indian languages’ capabilities of the system, can you tell us how you went about building it?

Tilakraj: The team’s Indian origin had a huge impact on the platform’s language features, particularly in terms of cultural understanding and linguistic diversity. We decided not to work with translations but with original material in the given language. While it was a challenge to find so much digitised material in Indian languages — other than the top 6 or 7 languages. It wasn’t just literature that we needed. We need current, on-going material from all areas as well as in languages.

Tathagat Prakash: To address this challenge, we looked towards synthetic data. Over a period of time this synthetic data will be replaced by actual data. But the synthetic data would have done its job — of not making the user feel like they are dealing with a translation.

Challenges include diverse structures, scripts, phonetics, scarcity of quality datasets, capturing contextual nuances, integrating cultural context, resource constraints for less common languages, orthographic disambiguation, speech recognition challenges, semantic ambiguities and the need for efficient algorithms in various technical environments.

Some recent developments suggest that guardrails around AI are not very strong. Have you done anything about this concern in QX?

Tilakraj: We use contextual inferencing and we have also put “ethical dilemma” in it, where if somebody puts in a query about some famous movie stars or great personalities from India or anywhere one else in the world, the program will stop. “Is it necessary to give that answer?” These are few things that we’ve worked very hard on. Before launching the product, we wanted to be very sure about contextual inferencing and the ethical dilemmas it brings in so that you know the product doesn’t affect the end user adversely.

Published on February 5, 2024





Source link

The movement towards lightweight materials

The movement towards lightweight materials


If a car is made 100 kg lighter, its fuel consumption decreases by about half a litre for every 100 km, according to the Fraunhofer magazine. Multiply this by the number of cars that ply on the roads, you will see that a little weight reduction makes an enormous difference. The magazine goes on to point out that if the weight of an Airbus 320 is reduced by 100 kg, it can save 10,000 litres of fuel per year.

Making things lighter is the newest tool available to mankind for fighting climate change.

Though lightweight materials are useful in all their applications, their utility is most evident in mobility. For decades, researchers have been toying with magnesium alloys for designing light-weighting automobiles, but there have been problems. According to a report, the average magnesium content in US-made cars has remained unchanged at about 5 kg for two decades. So, there is a need to move over to something else.

Researchers are therefore turning to composites, mainly carbon composites. Lightweight composite materials (LWCM) have been used in racing cars for a while — the task now is to mainstream it. Composites are an artificial combination of two different materials. The base material is called the matrix and the one bound to it is called filler. By toying with different combinations of matrices and fillers, researchers can come up with something good. But this is an arduous task. After developing each material, the scientists must test it for its ability to take different forms of stress, durability under various temperatures and other qualities such as water absorption.

This is where modern science is stepping in. Machine learning (ML) is helping fast-track material development. In a paper, researchers Nesrine Amor et al, of the Technical University of Liberec, Czech Republic, point to the growing use of computational intelligence in finding lightweight materials, especially composites. Testing the composites is also much faster and less expensive now. In the past, researchers would perform simulations to predict the material properties using ‘Finite Element Methods’— a time-consuming process. But in recent years, ‘computational intelligence’ (such as machine learning) has become the backbone of engineering. “ML methods work as excellent tools for LWCM due to their short development cycle, accurate prediction ability and strong data analysis. Nowadays, ML methods are widely used for the prediction of thermal and mechanical properties, dielectric constant and new materials design,” they say.

Scientists are finding success in bringing in lightweight materials into vehicles. Recently, the German Fraunhofer Institute announced that its applied polymer research scientists are working on a drive shaft made of fiber-reinforced plastic. Drive shafts transfer power from the gearbox to the wheels and must withstand high loads. Therefore, they have always been made of alloys of steel, aluminium or titanium. If the Fraunhofer scientists succeed, a drive shaft made of carbon fiber-reinforced plastic may become a reality. These drive shafts will be 60 per cent lighter than their metal counterparts but also have better mechanical properties.

Beyond mobility

The need for light-weighting materials goes beyond vehicles. For example, building materials could be made lighter and energy efficient. One research speaks of the immense potential that ‘lightweight concrete’ holds. The scientists investigated the merits of ‘lightweight expanded clay aggregate’ (LECA). They studied different levels of LECA mixed with conventional aggregates for making concrete and tested the lighter concrete for mechanical strength and water absorption. They found “10 per cent replacement” to be the “optimal percentage”. A cubic metre of this concrete weighs 1,821 kg, compared with 2,548 kg of conventional concrete.

But one could move away from concrete itself. Research is on for the use of alternative materials like bamboo, compressed earth blocks, fly ash bricks, and glass fiber–reinforced concrete.

“Lightweight construction is not just about making slight reductions in mass,” notes the Fraunhofer magazine. “This construction philosophy focuses on selecting materials, designing products… in a way that conserves materials and makes products lighter, while also improving functionality and safety at the same time,” it says.

In sum, lightweight materials are a new philosophy sweeping through the research world.





Source link

A new device for cooling data centres 

A new device for cooling data centres 


India is fast becoming a data-centre hub. The Indian market, estimated at 700 MW, is expected to grow to 1.5 GW in the next 3-4 years. Data centres are big energy guzzlers. The International Energy Agency (IEA) estimates that in 2022, data centres around the world consumed 240-340 terawatt hour (TWh) of electricity, roughly 1-1.3 per cent of global final electricity demand.

Why do they need consume this much? Because energy is needed to cool the servers at the data centres. There are three ways of cooling the servers — room level, rack level and chip level. Of the three, chip-level is said to be the most energy efficient.

Chip-level cooling is done by phase-change materials (like ice) or by ‘heat pipes’. Most heat pipes feature a porous wick which moves the working fluid to the evaporator from the condenser. But the wicks are costly.

Then there are these heat pipes called thermosyphons. They simply siphon out the heat — a liquid moves through it, picking up heat from the hot regions and dumping the heat into the atmosphere. Thermosyphons are gravity-assisted wickless heat pipes that utilise gravitational force to transport the working fluid from the condenser section to the evaporator section.

A group of scientists led by Prof Pallab Sinha Mahapatra and Prof Arvind Pattamatta from the Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai, have developed an ‘Integrated Flat Thermosyphon Heat Sink (IFTHS)’ with enhanced performance. The IFTHS has a condenser with integrated hollow fins, the inner surface of which is used for condensation and the outer surface is used for air-side convective cooling.

In their paper published in Applied Thermal Engineering, the researchers have said that a super-hydrophobic (water-repelling) condenser and a mini-channel used in the device gave superior cooling performance. “The lifespan of electronic components can be increased by 3.73–17.75 times and the room air cooling load can be reduced by 8.04–25.12 kJ/kg when the IFTHS configurations are used for cooling electronic chips in data centre servers instead of the baseline configuration. The novel IFTHS configurations developed in the present study can contribute to energy-efficient chip-level thermal management in energy-intensive data centers,” the paper noted.

The conventional air-cooled heat sinks used for cooling CPUs in rack servers can be replaced by the IFTHS without any additional modifications, it says.

Prof Marco Marengo, Professor of Thermal Sciences, from the Department of Civil Engineering and Architecture, University of Pavia, Pavia, Italy, has reviewed this work. In his comments, published in IIT-Madras Tech Talk, Prof Marengo noted that by using both inner and outer surfaces for condensation and air-side convective cooling, the IFTHS outperforms (the conventional) flat thermosyphon.

“This technology has the potential to enhance the cooling effectiveness of these systems in energy-intensive sectors and for data centres,” he says.





Source link

Stay younger, longer

Stay younger, longer


How nice it would be if you could just pop a pill and stay young! That might seem like the stuff of fiction, but if you sniff the scientific air, you can smell something good.

Senescent cells — cells that have lost power to divide and multiply — contribute to age-related decline. Scientists are looking at some small-molecule drugs that eliminate the senescent cells, but it appears that you have to keep taking them all the time. But now, scientists have discovered a better drug.

Our bodies have what are known as T-Cells, a type of white blood cells, which are the principal players in our immune system — they attack the ‘invaders’ and (most often) kill them. This wonder drug, developed by a group of scientists of the Cold Spring Harbor Laboratory, New York, USA, gets the T-cells attack the senescent cells.

“If we give it to aged mice, they rejuvenate. If we give it to young mice, they age slower. No other therapy right now can do this,” said Corina Amor Vegas, Assistant Professor at the Cold Spring Harbor Laboratory and one of the authors of the paper.

Well, the drug does not appear to be human-ready as yet. Wait until it becomes ready to be administered to humans. Don’t age.





Source link

Green coatings

Green coatings


A new technique of synthesising thin hard surface coatings by high velocity air fuel spraying, has the potential of emerging as an environment-friendly and safer alternative to hard chrome plating used for car parts, tools and kitchen utensils.

Chrome plating is generally used as it is hard and wear resistant. However, the presence of chromates, fluorites and hexavalent chromium makes it carcinogenic. This has lead the search for a safer, environment-friendly alternative with an equivalent or superior wear resistance and crack-free coating. Deposition of thin coating with industrially acceptable surface roughness is economical as it requires less powder and elimination of several grinding processes.

While with conventional thermal spray techniques, thickness build up is high and several grinding and polishing operations are needed to acquire the required thickness and roughness. A new technique called high velocity air fuel (HVAF), involving low temperatures and high particle velocities can deposit coatings using finer sized powders (5-15 µm).

Scientists from ARCI, Hyderabad, have carried out the synthesis of these thin hard coatings made of a composite alloy of tungsten, cobalt and chromium using high velocity air fuel spraying method. With this, thin coatings were deposited with torches of different capacities and by employing different nozzle sizes, says a press release.

The coating can be deposited on as-machined condition to achieve smooth surface and around 50 µm coating thickness. This significantly reduces the post coating finishing operations which reduces the processing and raw material cost significantly with better wear resistance than HCP, the release says.





Source link

Cool magnets

Cool magnets


Researchers have found a new alloy that can act as an effective magnetic refrigerant for minimising greenhouse gas emissions meet the global demand for higher energy efficiency to tackle global warming.

Magnetic refrigeration offers an energy-efficient and environment-friendly cooling technology as an alternative to the vapour-cycle refrigeration technology in use today. Hence efforts are on to fabricate magnetic refrigerators for household, industrial and technological applications.

Magnetic cooling effect (MCE) is the reversible temperature change of a magnetic material when it is subjected to an external applied magnetic field. Magnetisation and de-magnetisation leads to the magnet either giving out or absorbing heat.

Current research is focused on developing new magnetic materials for use as refrigerants. The material must be capable of operating for millions of cycles without any fatigue and failure, must have high thermal conductivity and should respond to an external magnetic field of about 2T (Tesla). Since most of the materials developed so far show giant magneto caloric effect (GMCE) with fields reaching as high as 5T, there is an urgent need to look for materials in which GMCE is achieved in lower fields.

Researchers from the SN Bose National Centre for Basic Sciences, Kolkata, experimented with an alloy called all-transition metal-based Heusler alloys (which are magnetic inter-metallics with face-centred cubic crystal structure) in their search for material exhibiting giant reversible MCE. “The search for the right kind of magnetic material has yielded positive results at the SN Bose Centre Lab. The synergistic combinations of giant MCE and MR by proper tailoring of Cu-doped Heusler alloys may lead to a diverse range of solid state-based technological applications,” says a press release from the Department of Science and Technology, Government of India.





Source link

YouTube
Instagram
WhatsApp