Study Finds AI Models Unprepared to Tackle Financial Problems, Far from Replacing Professionals

A new study by researchers from Clemson University and Washington State University finds that ChatGPT isn't completely reliable to give financial advice yet. The researchers analyzed 10,000 responses from AI after asking it different questions about finance and the results revealed that AI still has a long way to go until it can take over finance jobs. ChatGPT and other AI models are good enough to specify broader concepts of finance and summarize the issues and their solutions but it still cannot work on it thoroughly and in detail.

Many people like Goldman Sachh predicted that 15% to 35% finance jobs will be soon assisted by AI and it may take over the jobs of asset and wealth managers. But this study highlights that there is no need to panic because AI still doesn't have adequate knowledge of finance that is needed to handle the large systems. Even though AI can pass finance certification exams, it still isn't enough and we need to find what these AI models can actually do.

The researchers of the study assembled 1083 MCQs from different finance licensing exams and tested AI models like ChatGPT, Bard and LLaMA. These exams are required to be passed by human financial professionals in order for them to get a license. The researchers evaluated the responses of AI models based on how well they could explain the reasoning behind choosing a certain answer and how much sophisticated language they used while giving explanations.

ChatGPT-4 showed the best results, with 18% to 28% more accurate answers than other AI models. When researchers gave ChatGPT-3.5 some examples of responses, it nearly matched the responses of ChatGPT-4 showing that it can also give accurate answers which can resemble human professionals. There was a 73.4% accuracy on answers about customer accounts, trading and prohibited activities by both these models, but both of them showed less accuracy(56.6%) on answers about investment objectives and client financial profiles. There were also inaccurate answers by both models on specified problems related to tax status and insurance coverage.

The team of researchers also asked ChatGPT to evaluate potential merger deals and are now exploring how well this AI model can do if they test it on the deals that came after September 2021. Overall, the limitations of AI when it comes to finance are still there are AI models are not ready at all to do high-profile financial tasks as they can have serious legal and financial consequences.

Image: DIW-Aigen

Read next: iPhone 16 Named Best Small Phone, Samsung S24 Ultra Wins Best Smartphone of 2024
by Arooj Ahmed via Digital Information World

Crypto Thefts On The Rise As More Than $2.2 Billion Stolen From Various Platforms in 2024

A new report by Chainalysis is shedding light on the growing rise of crypto theft incidents. In 2024 alone, more than $2.2 billion worth of crypto was stolen from these apps.

This is not the first time that such alarming reports made the air. It’s in fact the fifth time in the past decade that we’re hearing about more than $1B getting stolen. In other years, the number of theft reportings rose in 2018, 2021, 2022, and even 2023.

Another very interesting point in this report has to do with Chainalysis seeing the trend where crypto was stolen from in terms of exact location. It was commonly observed that DeFi platforms were a prime target for stealing including in the first quarter of this year. Meanwhile, during the second and third quarters, the focus tends to shift to centralized apps. Some noteworthy mentions include DMM Bitcoin and the WazirX where nearly $305M and $234M were stolen.

The growing rise in attacks on these apps has brought to attention another very interesting point related to private security features. Most of these hacking incidents arose due to compromised private keys that gave hackers the greatest access to such funds. So to prevent such thefts, apps must take this into consideration and better security.

There is another notable mention of countries like North Korea which is to blame for a huge number of crypto theft incidents. The country might be closed but it ended up stealing close to $1.3B worth of valuable crypto assets this year from a whopping 47 different incidents.

This is a massive 61% growth in crypto theft in 2024 alone and the matter made up more than 20% of all such incidents. It’s quite sophisticated in terms of its tactics and it uses them to produce nuclear weapons and ballistic missiles.

Remember, Bitcoin’s price continues to hit new highs in the past couple of weeks and hacks keep rising as it’s a simple way for many to make revenue. This is why platforms should invest in building up users’ trust through better security measures or it can be uneventful for all involved.

This is why the report highlights crypto regulatory networks and the need to keep getting better. With the right law enforcement in place and resources, the industry can become more trustworthy.

Read next: AI Competition Heats Up As Google Shares First Reasoning-Focused LLM Called Gemini 2.0 Flash Thinking
by Dr. Hura Anwar via Digital Information World

AI Competition Heats Up As Google Shares First Reasoning-Focused LLM Called Gemini 2.0 Flash Thinking

The AI battle continues to heat up as Android maker Google just shared its latest reasoning-focused large language model dubbed Gemini 2.0 Flash Thinking. The model is equipped to give OpenAI a run for its money as Google promises better performances in the world of science, coding, and even math.

The models are designed to solve the most complex tasks involving reasoning and while it’s still an experimental design, the anticipation is growing among tech experts. As per the Android maker, the latest model is for developers to benefit from inside the company’s AI Studio.

Back in September, the tech giant did share how it was trying to better its models so they gave out more intelligence and suitable responses. It seems like that time is here. Google shared how the results were certainly promising with enhanced inference timing for computation.

However, the company didn’t share the inference timing computation which it reportedly increased. No other benchmarks were available publicly to make the claims. But as per Chatbot Arena, this Flash Thinking model currently ranks at the number one position for all domains. Google also shared many use cases in the developer portal for users to test Gemini 2.0 Flash Thinking models.

The model shows support for text lengths greater than 128k and even comes with cutting-edge stats for knowledge. Furthermore, developers attain access to the latest Gemini reasoning models through a Gemini API inside Google’s AI Studio and the Vertex AI.

During the week’s start, OpenAI shared how its new o1 reasoning models are launching to developers on usage tier 5 inside the API. This latest update for the o1 model gives rise to the best results on different AI benchmarks.

Meanwhile, developers also get the chance to make use of it to create apps for better client support and optimize the entire supply chain decision. It’s also wonderful for matters such as making financial trend predictions.

The latest unveiling of the LLMs provides developers with more power systems to create the most innovative AI apps throughout different industries. We can see how this launch can really safeguard Google’s mark as a strong contender in the world of AI as it takes on OpenAI.

Image: DIW-Aigen

Read next: Instagram Shares A Glimpse of Exciting Video AI Elements Coming Soon To The App
by Dr. Hura Anwar via Digital Information World

Instagram Shares A Glimpse of Exciting Video AI Elements Coming Soon To The App

It’s clear that Meta wants to make a mark with AI and that’s why its recent target is introducing new AI video elements on the Instagram platform.

Instagram’s chief just shared a glimpse of some new and exciting AI video features that are coming soon. Amongst those include the Movie Gen Engine where users get the chance to rollout wholesale edits of videos in-stream.

Similarly, it is launching the option to switch backgrounds, representations, outfits, and even Muppet variants of themselves. It’s all very interesting and might give rise to some more exciting and creative trends. However, it’s also one step away from the social aspect of the platform where users share connections and experiences with their followers.

AI only comprises pictures and videos that have very little to do with one’s life or even creativity. The video edits feel more aligned where it still involves captured content and only that can give rise to the most compelling clips. Remember, elements of visual storytelling come with enhancements that assist in building up a post.

The same cannot be said for image generation through AI. It’s all quite interesting and novel but we don’t see this as being something complimentary within the social media realm. You can produce images of aliens but it cannot symbolize things that are real, right? Users can roll out images of the beach but what happens when they’re not going in the first place?

• Also read: Are Surveillance Cameras Making Us More Aware of Our Actions and Surroundings?

Meta’s latest AI video tools do give the right enhancements and also lead to exciting creative forms inside the app. We feel that some would find it so valuable. Still, the main point worth considering is storytelling and that is quintessential right now. If social isn’t a part of it, how do you interact? It’s just plain images and videos flashing on your screen and brains, giving you utter confusion.

That might be the end result for Generative AI in the end with surreal creations designed to play with the brain. So the take-home message is that narratives are crucial in the world of social media and you’ve got to understand that.

This would happen as more and more people start to experiment with different options. How valuable they end up being, only time can tell. For now, it’s a great point to better communicate through human feelings or emotions.

Read next: AI in the Workplace: Majority of Workers Avoid Paying for Tools, Survey Shows
by Dr. Hura Anwar via Digital Information World

AI in the Workplace: Majority of Workers Avoid Paying for Tools, Survey Shows

According to a new survey by Tech Brew, most of the office workers do not pay for AI tools for their work. AI is getting quite common in workplaces and many office workers have started using it during their work too, but the results of the survey suggest that most of the office workers are not spending anything on AI.

Out of 878 people who were surveyed, 78% of them said that they do not pay anything for AI tools, while only 16% spend $1-$20 on AI tools every month. Only 5% of respondents said that they have allocated more than $20 every month for the AI tools that they use.

The survey also found that more than half of the respondents use AI at work, whether it's provided by their company or they use on their own. 23% of them use customized AI tools, 30% of them use third party AI tools and 38% of the respondents reported using AI tools on their own while working.

The people who are more likely to pay for AI tools are business owners(43%) and freelancers(31%) and they are the ones that pay for AI independently. Most of the other workers have their companies for any AI tools they use during their work. When the respondents were asked why they are not using paid AI tools, most of them said that cost was a problem. Sometimes, companies cannot handle the extra expense of paying for AI tools and other times, the workers do not have the capacity to pay on their own.

• Related: Study Reveals Prolonged Standing May Pose Health Risks, Challenging Desk Job Solutions

69% of the respondents said that they always use AI at work, followed by 32% who use them often and 26% who use AI tools only sometimes. 13% of the respondents said that they use AI rarely while 18% said that they have never used AI tools for work. There was an age factor when it came to using AI at work. Workers between the ages of 18-24 are more likely to use AI Tools than people of older ages. People with ages more than 44 are more likely to not use AI Tools at work at all. Gen-Z said that AI helps them sound professional at work and guides them and how to make emails and rewrite things to sound professional.

Respondents from marketing and advertising reported using AI persuasively, with 90% saying that they use AI tools sometimes. 78% of the respondents from technology and computer industry also reported using AI for work, with Microsoft CoPilot, ChatGPT, Perplexity and Adobe being the most used AI tools.

Image: DIW-Aigen

Read next: Which Countries Have The Best and Worst Work-Life Balance?
by Arooj Ahmed via Digital Information World

Google Warns Gmail Users of Rising Cyber Threats Ahead

Google is warning its 2.5 billion Gmail users about a looming second wave of cyber threats. Attackers, it seems, are gearing up for the final push of the year. While phishing attacks have dropped by 35% compared to last year’s holiday season, there’s been a massive increase in email traffic since mid-November, making it harder than ever to protect users’ inboxes. Google, as the world’s most popular email service, remains a top target for hackers, but the company is doing everything it can to shield users from these persistent threats.

To this end, Google has invested heavily in new technology. They’ve developed AI models specifically designed to detect phishing, malware, and spam. According to Andy Wen, Gmail’s senior director of product management, these new systems have already blocked 20% more spam than before. One of the latest models is an AI tool that acts as a supervisor, reviewing hundreds of potential threats in an instant. It’s all part of Google’s effort to stay ahead of cybercriminals, but even with these innovations, the fight is far from over.

As we head into the busiest time of year, Google is warning users about three specific scams that have resurfaced. First, the “We know where you live” extortion scam. Attackers send emails with personal information about the victim—like their home address—along with threats of physical harm or blackmail. It’s a tactic designed to frighten people into responding quickly without thinking.

Image: Google

Second, there’s the Gmail invoice scam. It’s as old as the internet, but still effective. Fraudulent invoices are sent to victims, claiming they owe money for a service or product. The scammer then urges the victim to call a phone number to dispute the charges. Once the victim is on the line, the scammer uses pressure tactics to trick them into paying fees or handing over personal information.

• Related: New Report Reveals Rising Trend of Fake Captchas and Malvertising on the Internet

Lastly, we have the celebrity impersonation scam. In this case, attackers send emails pretending to be famous personalities, either claiming they’re endorsing a product or trying to sell something in their name. It’s a clever trick that preys on trust, making people more likely to click or buy something they wouldn’t otherwise.

The good news is, you don’t have to fall for these scams if you stay alert and think critically. Google’s advice is simple, slow down and think. Don’t let the sense of urgency get the best of you. If something feels off or sounds too good to be true, it probably is. Take a minute to double-check the email address of the sender. And remember, no reputable organization will demand money or personal details via email. An important tip is to search key parts or the entire suspicious email or message on Google, which might reveal clues about its authenticity. 

If you get a suspicious email, don’t hesitate to report it. By marking these messages as spam, you’re not just cleaning up your inbox; you’re also helping Google improve its AI defenses, making Gmail safer for everyone.

In the end, protecting your Gmail account is just like locking your front door. You can have all the security systems in place, but you still need to take care in how you open it. Cybercriminals are always looking for ways in. The key is to stay alert, think before you act, and use the tools available to keep your inbox safe.

Read next: Human Rights Watch Labels Israel's Gaza Tactics Genocide, Intensifying Global Calls for Justice
by Web Desk via Digital Information World

Human Rights Watch Labels Israel's Gaza Tactics Genocide, Intensifying Global Calls for Justice

Israel's actions in Gaza have come under renewed focus as Human Rights Watch (HRW) published a scathing report, which has pointed to the denial of basic resources like clean water for Palestinians.

Illustration: HRW

The report directly labels these actions as acts of genocide, which underscores the mass killing of civilians as a violation of international law. HRW claims that these policies amount to extermination and genocide, as defined by the 1948 Genocide Convention. While proof of intent remains an important part of legal classification, statements made by Israeli officials might suggest the intent to destroy the Palestinian population, thereby strengthening the case for genocide.

This assessment follows similar accusations by Amnesty International, marking the second major rights group to use the term genocide in this context. Israel has repeatedly rejected these accusations, asserting its right to defend itself following the October 7, 2023, Hamas-led attack, which resulted in over a thousand Israeli deaths and hundreds taken hostage. In response, Israel launched a large-scale air and ground campaign, which has reportedly killed over 45,000 Palestinians, displaced millions, and left much of Gaza in ruins.

The legal ramifications are significant. The International Criminal Court has issued arrest warrants for Israeli leaders over alleged war crimes, including genocide. Israel denies these allegations, maintaining its stance that its actions are necessary for national security.

The blockade on Gaza has crippled essential infrastructure, with water, electricity, and fuel supplies severely restricted. This has led to catastrophic conditions for Palestinians, as the Human Rights Watch report details the extent of the crisis. With genocide now being formally attributed to Israel’s actions, international discussions on accountability and justice are more urgent than ever.

Information is everywhere, yet too often, it doesn’t lead to action. We have the tools to see the truth, but the stories we hear are shaped by the powerful, not the vulnerable. Technology should inform, but it too often distorts, clouding the human cost of conflicts.

Powerful nations and digital platforms decide what we focus on, leaving lives, dignity, and morality on the sidelines. The real question is what we choose to do next. When knowledge is within reach, inaction becomes a choice. And that choice — more than anything — defines us.

Top featured image: Dixit Dhinakaran / Unsplash

Read next: Supreme Court Steps In as TikTok Challenges U.S. Sale Order Deadline

by Asim BN via Digital Information World