Sustainability-in-Tech : Search Engine Sustainability Shock
With the integration large language models (LLMs) into search engines, some are predicting that the massive increase in computing power needed could mean huge carbon emissions.
What Are Large Language Models?
Large language Models (LLMs) are types of artificial intelligence (AI) models that are trained on vast amounts of text data to understand natural language. These models are typically based on deep learning architectures such as neural networks, and are capable of generating human-like language and carrying out a variety of natural language processing tasks. OpenAI’s ChatGPT and Google’s Bard chatbots are examples of LLMs.
Integrating LLMs Into Search Engines
Following the massive success of OpenAI’s ChatGPT (OpenAI has close working links with Microsoft), both Google, Microsoft, and now Chinese search company Baidu have all announced plans to upgrade their search engines by integrating generative AI tools which use LLMs to enable their search engines to understand and respond to complex questions. This is intended to give search engine users a better search engine experience and enable the search engines to compete with eachother in this new area.
For example:
– Microsoft has announced that it is to introduce a “new, AI-powered Bing search engine and Edge browser” (in preview Bing.com), using OpenAI’s LLM, to “deliver better search, more complete answers, a new chat experience and the ability to generate content.”
– Google has announced that it is testing and will soon be introducing its own conversational AI chatbot, powered by LaMDA, Google’s own AI, and that it will be integrated into the Google search engine.
Environmental Implications
In addition to worries about inaccuracies in the answers given by chatbots e.g., Bard’s recent costly wrong answer given in an advert for the chatbot, one major concern that many have overlooked is how much carbon emissions could be increased through the wider use of LLMs.
How And Why?
As highlighted in quotes from University of Surrey Professor Alan Woodword (in Wired), “There are already huge resources involved in indexing and searching internet content, but the incorporation of AI requires a different kind of firepower.” Professor Woodword is referring to his view that the wider use of LLMs could be a step change in online processing that could massively increase the power and cooling resources needed by large processing centres which could, of course, have a much bigger environmental impact i.e., more carbon generation. There may also be increased challenges in how data centres will deal with the extra heat produced.
How Much?
An idea of how big environmental problem this could be may come from a third-party study published on a Cornell University arXiv archive which states that “larger models translate to greater computing demands and, by extension, greater energy demands.” The research paper highlights how training GPT-3, autoregressive language model that ChatGPT is partly based on, consumed 1,287 MWh and that this led to emissions of more than 550 tons of carbon dioxide equivalent. To put the figure in perspective, this is the same amount of CO2 that would be produced by a single person taking 550 roundtrips between New York and San Francisco. Adding to this the fact that more LLMs are being introduced, and integrating chatbots into search engines such as Bing and Google which have tens of millions of users per day has some tech commentators, such as Martin Bouchard of Canadian data centre company QScale to estimate that this will mean “at least four or five times more computing per search.” In order to process this demand, more hardware and more data centres will be needed, which is an unwelcome prospect considering that data centres already account for one per cent (IEA) of the world’s greenhouse gas emissions. This may also make it very challenging for big tech companies to meet their green targets e.g., Microsoft aiming to be carbon negative by 2050.
AI Can Also Help Reduce The Impact Of Itself
That said, there are several ways that AI could be used to help offset the extra energy and carbon impacts that the increased use of Large Language Models (LLMs) produce. For example:
– Helping to develop more energy-efficient training methods. AI researchers can use machine learning algorithms to optimize the training process and reduce the number of computations required to train a model, which can significantly reduce the energy consumption.
– Cloud providers can use AI to optimise their data centres and reduce their energy consumption. For example, machine learning algorithms can be used to predict the demand for cloud resources and allocate them more efficiently, reducing the number of idle servers and minimising energy waste.
– Researchers are also exploring the use of green computing technologies to reduce the energy consumption of LLMs. AI algorithms can be used to optimise the scheduling of computing tasks and reduce the number of idle processors, which can significantly reduce the energy consumption.
– Sustainable computing practices can be adopted to ensure that LLMs are developed and used in an environmentally responsible way. This includes using renewable energy sources, reducing waste, and recycling materials whenever possible.
What Does This Mean For Your Organisation?
So much has been reported about the amazing capabilities of LLMs and the new generation of chatbots led by the arrival of ChatGPT, and of how search engines could be seriously upgraded by incorporating them, that the possible environmental impacts appear to have been overlooked and under-reported until now. Data centres are already struggling to cope with demand and the need to reduce energy consumption and carbon emissions, and incorporating chatbots (which already have large energy requirements) into search engines which process hundreds of millions of searches per day looks likely to have a huge negative environmental impact i.e., higher energy requirements, greater carbon emissions, and the need for even more data centres. Now may be the time for tech and computer giants to get together and focus on finding new and innovative ways to minimise the environmental impact of these new technologies e.g., perhaps using more environmentally friendly AI-based solutions. Also sourcing more green and sustainable energy and being transparent and ethical in the use of data could help, but in the short term, it looks as though the rise of these new super-powerful chatbots is likely to create more environmental challenges than solutions.
Tech Tip – Useful YouTube Shortcuts
If you’d like a fast and easy way to navigate within YouTube, here are some top keyboard shortcuts.
– While watching a YouTube video, use the following letters on their own on the keyboard (no need for shift) for the following functions: F (full screen), K (play/pause), J (rewind 10 secs), and L (fast-forward 10 secs).
– Shift+P plays the previous video.
– Shift+N plays the next video.
Featured Article : Do Hackers Get Pensions & Sick Pay?
Kaspersky research has shown that hackers are now being recruited with dark web job ads offering huge salaries and benefits!
What?!
According to Kaspersky’s analysis of more than 200,000 employment ads posted on the dark web between January 2020 and June 2022, ‘employers’ have been seeking applications from the best hackers by offering favourable terms of employment including remote work (obviously), full-time employment, flexitime, paid time off, paid sick leave, and even the promise of working with ‘a friendly team’ !
Massive Salaries
If the research by Kaspersky’s Digital Footprint Intelligence team is accurate, it seems that online crime does pay. For example, the median levels of pay offered to IT professionals in the ads varied between $1,300 and $4,000 per month.
Other ways that Kaspersky’s team observed ‘employers’ seeking to tempt applications from hackers in dark web job ads included offering:
– Prospects of promotion and incentive plans/bonuses (or fines, depending).
– Levels of compensation (paid in cryptocurrency for privacy) depending on how much effort hackers invest, their contribution, and how successful the ‘business’ is on the whole.
Reverse Engineering
The highest median salary of $4,000 could be found in ads for reverse engineers. Although not all reverse engineers are hackers, reverse engineering refers to taking apart a software or hardware system and analysing its components to understand how it works. In some cases, reverse engineering can be used for malicious purposes, such as hacking into a system or creating malware. In these situations, the reverse engineer uses the information gained from the reverse engineering process to exploit vulnerabilities or gain unauthorised access to a system.
Other Suspect Job Ads – ‘Developers’ and Malware Coders
Kaspersky’s research noted that although the innocent-sounding “Developers” (accounting for 61 per cent of the job ads) were the most in-demand specialists on the dark web, within this speciality, 60 per cent of the developer ads sought people who create internet products like phishing pages. Another popular skill sought in the ads was malware coding where job descriptions were found to include the development of Trojans, ransomware, stealers, backdoors, botnets, and other malware types, along with the creation and modification of attack tools. Other specialists sought included:
– Attackers. These conduct attacks on networks, web applications and mobile devices, and accounted for 16 per cent of the total ads, making them the second most popular jobs among cybercriminal employers. Kaspersky noted that: “Most of the attackers’ jobs on the dark web were associated with actions that would compromise corporate infrastructure. The goals of these actions are ransomware infection, data theft, or stealing cash directly from accounts.” Groups hiring attackers focused on selling access to compromised systems to other cybercriminals or hacking web and mobile applications.
– Designers who can create a phishing page or letter that would be hard to distinguish from a real one. These were found to be the third most in-demand professionals with 10 per cent of the observed ads.
Boost In Demand Following The Pandemic
The Kaspersky research also noted that the fact that the greatest number of ads were posted in March 2020, was most likely related to the outbreak of the COVID-19 pandemic and the ensuing changes in the structure of the job market.
What Is The Dark Web Job Market Like?
The dark web job market is characterised by a mix of both legal and illegal job offers. Some employers offer semi-legal or potentially legal jobs, such as creating IT learning courses, while others offer more dubious or illegal employment arrangements. These may include selling illegal drugs, engaging in fraudulent schemes, or hacking activities, or working with hacker groups.
Some people are attracted to dark web jobs because of the potential for easy money and high financial gain. However, not all salaries offered on the dark web are significantly higher than those earned legally and depend on experience, talent, and effort. Other individuals may turn to the dark web job market due to dissatisfaction with their current employment, changes in the job market, or a lack of certain candidate requirements.
Dark web jobs may also appeal to freelancers and remote workers because of the flexibility and freedom they offer, such as the ability to take time off and choose a schedule. Unlike the legitimate job market, however, getting dubious jobs on the dark web comes with the real risk of being arrested and prosecuted, and there is no guarantee of being paid.
How Are Criminals Able To Simply Advertise For Hackers Online Like This?
Criminals are confident enough to post job ads on the dark web because it has a level of security and privacy through its use of encryption and anonymising technologies. The most popular tool for accessing the dark web is the Tor (The Onion Router) network, which routes internet traffic through a series of servers around the world, making it difficult to trace the origin of the traffic. This makes it more difficult for policeand even for hackers and other malicious actors themselves, to monitor or interfere with the traffic on the dark web.
However, the dark web is not completely secure and anonymous. While the technology provides a degree of privacy and security, law enforcement agencies have the resources and expertise to penetrate and monitor dark web activity. For example, they have been able to dismantle some dark web marketplaces and arrest individuals who were found to have been engaging in illegal activities on the dark web.
Also, the dark web has its own set of security risks, such as the possibility of being scammed or hacked, or being infected with malware. It’s also important to remember that many illegal activities, such as buying or selling illegal drugs or stolen goods, or posting job ads for illegal purposes, are still against the law on the dark web and can result in severe legal consequences if caught.
Do People Actually Post Their CV On The Dark Web?
Yes. According to Kaspersky, the research analysed 867 ads that contained specified keywords, 638 of the ads were vacancy postings and 229 were CVs. The statistics suggest, therefore, that jobseekers respond to ads by prospective employers more frequently than they post CVs but, nonetheless, some post CVs on dark web forums that target diverse areas of expertise and job descriptions, such as moderating Telegram channels to compromising corporate infrastructure.
Monitoring The Dark Web Is A Form Of Defence
Following the shocking research results, as noted by Polina Bochkareva, Security Services Analyst at Kaspersky: “IT headhunting is one of the numerous topics which is constantly discussed on the Darknet. Nowadays, tracking cybercriminal’s interest and continuous analysis of their activities is vital for companies that want to proactively respond to cyberattacks and keep their information security at the highest level. The more you know about your adversary – the better you are prepared”.
Protection
Some ways that businesses can protect themselves from common cyber-attacks include:
– To protect against phishing attacks, businesses can implement anti-phishing software, provide regular security awareness training to employees, and encourage them to be cautious when receiving emails from unknown sources.
– Businesses should regularly backup their data, keep their software and operating systems up to date, and train employees to identify suspicious emails and attachments to avoid falling victim to ransomware and other types of malware.
– To protect against Distributed Denial of Service (DDoS) attacks, businesses can implement DDoS protection services, monitor network activity for signs of an attack, and have a plan in place for responding to a DDoS attack.
– SQL injection attacks involve injecting malicious code into a website’s database in order to steal data or compromise the website. To protect against these attacks, businesses can use parameterised queries to protect against SQL injection, keep their database software up to date, and regularly monitor their websites for signs of an attack.
– Seeking the help of their IT Support Company or other security professionals.
– Using the multiple sources of Threat Intelligence information to stay aware of actual TTPs used by threat actors.
What Does This Mean For Your Business?
The Kaspersky research highlights how the dark web has long been a haven for cyber criminals and that there appears to be a flourishing job market there where ‘employers’ now blatantly use the same enticements that are used in legitimate jobs to attract the most specialised criminals. It looks unlikely that law enforcement or regulation will be able to make much of an impact in shutting down the worst aspects of the dark web in the near future, and even if they do, criminals are likely to move to other secure platforms and channels. It seems, therefore, that the best thing businesses can do is to take the measures mentioned above and remain focused on making their own cyber security as robust as possible and to keep on top of new security measures they need to take going forward.
Tech Insight : Stopping SafeSearch From Being ‘Too Safe’
Following anecdotal reports that some users can’t turn off the strict SafeSearch settings when searching using Bing, we look at what SafeSearch is, and how to control it.
What Is SafeSearch?
SafeSearch is a setting in Microsoft’s Bing search engine that’s designed to filter out inappropriate web content.
How?
It works on the search terms used and can be set to ‘Off,’ ‘Moderate,’ or ‘Strict.’ To choose the setting:
Using Edge
– Go to bing.com and click on the three dots top right.
– Select ‘Settings,’ and ‘More.’
– Choose your SafeSearch preference: ‘Strict,’ ‘Moderate,’ or ‘Off.’
– Select Save.
Using Google Chrome
– Go to bing.com and click on the three lines (top right).
– Select ‘SafeSearch.’
– Choose your SafeSearch preference: ‘Strict,’ ‘Moderate,’ or ‘Off.’
– Select Save.
What’s The Problem?
Some users are reporting a range of issues related to SafeSearch. For example:
– Even on a ‘Moderate’ setting, Bing filtering out search terms that are not inappropriate, e.g. one that contain some ‘x’ technical terms.
– SafeSearch automatically choosing a ‘Strict’ setting for some users.
– SafeSearch overriding a Moderate setting with a Strict setting with the explanation that a country or region requires it.
– Users not being able to turn SafeSearch to ‘Off.’
– SafeSearch settings being set and locked by the account, device, or network administrator.
About Strict
If a ‘Strict’ setting is initially locked on, e.g. it’s enabled by parents or administrators to help protect children and prevent them from accidentally accessing adult content, the setting is then out of the user’s control. It cannot be turned off by the user without the password or other authentication credentials that were set up by the person who enabled the feature.
Bing’s SafeSearch can also override a Moderate setting with a Strict setting if the SafeSearch feature is locked on the Strict mode, e.g. by a parent or administrator.
Country Or Region Issues?
Although some countries or regions may have different laws or cultural norms that impact what is considered appropriate content, plus some countries may require that certain types of content be filtered by default, regardless of the SafeSearch setting chosen by the user, this doesn’t explain the country/regional blocking/change of setting that some users have recently experienced. SafeSearch should always be under the control of the user or the person who set up the parental controls or other restrictions.
Ways To Fix SafeSearch If It Can’t Be Turned Off
Some ways to fix SafeSearch if it can’t be turned off include:
– Some reports online indicate that issues may related to a bug that can be fixed by an update. Therefore, make sure you have the most recent Windows updates, e.g. Win + I to open Settings, go to Windows Update and click the Check for updates button.
– Try using a different browser. Some online reports indicate that different browsers (e.g. Google Chrome) don’t have the issue.
– Check and consider changing your Windows settings. For example, from the Start menu go to Settings, click on Privacy & Security. In Windows permissions, select Search permissions. For SafeSearch, select Off – Don’t filter adult content from my web results.
– Set up a new user account with new SafeSearch settings on your Windows computer.
– Try disabling third-party antivirus on the computer which may be interfering with search results and triggering SafeSearch issues. Use caution here for obvious reasons!
– Use the Group Policy Editor to change your system settings and turn off Windows SafeSearch.
– Use the Group Policy Editor to turn SafeSerch off. To do this: press Win + R and in the Run dialog type gpedit.msc and click OK. In the Group Policy Editor window, select Computer Configuration > Administrative Templates > Windows Components > Search. Select Set the SafeSearch settings for Search (right hand pane). Set to Not Configured, restart the computer, and check if the issue has been fixed.
What Does This Mean For Your Business?
Features like SafeSearch are important for protecting younger users from being exposed to inappropriate content and for managing search habits for other users. However, there are circumstances where, for different reasons (as mentioned above) SafeSearch settings could be counterproductive. Keeping Windows updated and trying the suggestions mentioned above may resolve any issues you may have experienced.
Tech News : Google Bard’s $100 Billion Bungle
$100 billion/£82.7bn is reported to have been wiped off the value shares in Google’s parent company Alphabet following a rather public error made by Bard, Google’s new answer to ChatGPT.
Bard
Originally called ‘Apprentice Bard,’ Bard is Google’s conversational AI service which uses Language Model for Dialogue Applications (LaMDA), Google’s family of conversational neural language models. Bard is the Google equivalent of OpenAI’s ChatGPT chatbot.
What Went Wrong?
The much-anticipated Bard has already been rolled out to “trusted users” and scheduled for public release in “the coming weeks.” However, an advert for Bard which featured a promotional tweet contained a demonstration question which Bard had answered wrongly. As part of the answer to the question “What new discoveries from the James Webb Space Telescope (JWST) can I tell my nine-year-old about?”, Bard said “JWST took the very first pictures of a planet outside of our own solar system.”
This was quickly identified as being an incorrect answer because the first pictures of these ‘exoplanets’ were actually taken by the Very Large Telescope in 2004.
$100 billion/£82.7bn Wiped Off
The knock in confidence caused by the mistake led to shares in Google’s parent company Alphabet sliding by up to 9 per cent in regular trading. This equated to $100 billion/£82.7bn being knocked off the company’s market value by the end of last Wednesday.
Highlights The Importance Of Testing
A Google spokesperson commented on the somewhat painful and embarrassing mistake for the company by saying that it “highlights the importance of a rigorous testing process” and pointed to the fact that Bard is currently in the hands of its “Trusted Tester program”. The spokesperson added that it would combine external feedback with its own testing to make sure Bard’s responses were up to scratch in future.
The Race Is On
The release and the popularity of ChatGPT has not just been followed by the impending release of Bard (following testing), but there is also news that Microsoft’s revamped Bing search engine will be using large language model technology, and that Google looks likely to incorporate Bard’s technology in its search engine. All these developments are evidence that the AI chatbot race is well and truly on. Also, it looks likely that the addition of powerful chatbots to search engines will change the format of search engine results. For example, by incorporating AI chatbots like ChatGPT and Bard, instead of just delivering pages of web links in answer to search engine queries, users may be given one definitive answer, thereby making searches quicker with easier to digest answers.
What Does This Mean For Your Business?
OpenAI’s ChatGPT may have gained a substantial head start in the market with this type of advanced chatbot, nevertheless it was inevitable that Google would very soon be hot on its heels with its own version. One of the factors that affects the quality of answers from chatbots, however, is training. Whereas, in addition to its initial training, ChatGPT has been trained by millions of people since its release as a free version in November, Bard still appears to have some way to go. The wrong answer being used as part of a high profile advertising campaign is reminiscent of other high profile tech launch blunders like Tesla’s Cybertruck demo event where its bullet and sledgehammers-proof windows were easily smashed. The damage for Google with Bard, however, translated into a share slide and, no doubt, a little damage to market-confidence in Bard’s ability compared to ChatGPT. Bard is still undergoing testing, so it remains to be seen how it performs when it’s fully released for all of us to try.
Tech News : Quantum Computing Breakthrough
Researchers from the University of Sussex and Universal Quantum have announced that they have made a major breakthrough in quantum computing that could help solve the most complex real-world problems.
What Are Quantum Computers?
Quantum computers can carry out complex calculations at high speed. Whereas traditional computers store data in binary ‘bits’ (ones and zeros) and work by creating and storing long strings of these ‘bits,’ quantum computing’s ‘qubits’ (quantum bits) can do both at once. This is because a qubit can hold a zero, a one, or any proportion of both zero and one at the same time, and an array of qubits can use something called ‘superposition’ to represent all 2^64 possible values at the same time. This means that quantum computers can store more data in fewer bits (i.e. much more information can be stored in fewer qubits), meaning information can be processed much more quickly than with a traditional computer. The power of a quantum computer is stated in its quantum volume number/how many qubits, for example, a 14-qubit system.
Quantum computers can, therefore, be used to dramatically speed up tasks that have traditionally taken a long time. The results can be astounding, e.g. where crunching numbers that would take a classical computer a week, could take a quantum computer less than a second.
The Challenges
Although quantum computers operate on the 100-qubit scale, experts say that millions of qubits are needed to solve important problems that are currently out of reach of today’s most powerful supercomputers.
Even though quantum computers are developing, they are becoming constrained by the size of their microchip, which limits the number of quantum bits such a chip can accommodate.
The Breakthrough
The recent breakthrough, featured in a research paper published in ‘Nature Communications,’ is that a new powerful technique (called ‘UQ Connect’) allows multiple chips to slot together like a jigsaw puzzle to make a more powerful quantum computer. This ‘UQ Connect’ technique uses electric field links to enable qubits to move from one quantum computing microchip module to another with unprecedented speed and precision.
World Record Numbers
Using the UQ technique, The University of Sussex and Universal Quantum team have reported being successful in transporting the qubits with a 99.999993 per cent success rate and a connection rate of 2424/s. Both these numbers are believed to be world records and orders of magnitude better than previous solutions.
Boundless Applications
Professor Sasha Roseneil, Vice-Chancellor of the University of Sussex, highlighted the value of the breakthrough, saying: “It’s fantastic to see that the inspired work of the University of Sussex and Universal Quantum physicists has resulted in this phenomenal breakthrough, taking us a significant step closer to a quantum computer that will be of real societal use. These computers are set to have boundless applications from improving the development of medicines, creating new materials, to maybe even unlocking solutions to the climate crisis.”
What Does This Mean For Your Business?
The development of quantum computers has had many engineering, computer science and mathematical challenges to overcome just get it to this point. The recent discovery of how to effectively link multiple chips together in a way that is fast and accurate is the next big step forward in boosting their power and overcoming single chip constraints. This could mean that we are entering the realm of scaled-up, practical quantum computers that are capable of solving some of the biggest challenges faced by many different industries, and some of the biggest challenges facing us all, such as how to successfully treat some of our most serious diseases and tackle the climate crisis.