Tag: AI

Amazon Is Working On A ChatGPT Chatbot For Their Consumers

June 25, 2024

According to a Monday Business Insider article, Amazon is developing a consumer-facing chatbot in addition to the AI capabilities it intends to add to its Alexa voice assistant. The browser-based chatbot, internally code-named “Metis,” will reportedly mimic current chatbots provided by rivals like OpenAI and Microsoft and be powered by Amazon’s Olympus big language model.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

The U.S. Department of Justice (DOJ) and Federal Trade Commission (FTC) Launches Investigations

June 18, 2024

Government organizations carry out antitrust investigations to find out if businesses are participating in anticompetitive behavior that could hurt customers and impede competition.

Nvidia looked into being the dominant GPU. Concerns have been expressed over Nvidia’s hegemony in the GPU industry, according to the New York Times, “particularly how the company’s software locks customers into using its chips, as well as how Nvidia distributes those chips to customers.” Over the past week, Nvidia has surpassed Apple and Microsoft to become one of the three biggest corporations in the world by market capitalization; the three companies are valued almost equally.

Microsoft looked into possible connections to AI businesses. The close cooperation between Microsoft and OpenAI and the possibility of anticompetitive behavior will probably be the main topics of the investigation. There may be worries about competitors finding it more difficult to enter the AI business as a result of Microsoft’s $13 billion investment in OpenAI and the incorporation of OpenAI’s technology into Microsoft products.

More whistleblower protections from AI businesses are demanded by OpenAI staff. Nine current and former OpenAI employees wrote an open letter that was released last week, expressing worries that AI companies would hide hazards associated with their systems from the public and government. It then urges AI companies to grant permission to present and past workers to voice concerns about risks without fear of reprisal.

Former OpenAI board member Helen Toner explains why she is trying to remove Sam Altman. Although Altman was previously cited by the board for “not [being] consistently candid,” the public’s understanding of this has been limited. Toner provides three justifications in an interview for the board’s choice: Providing false information regarding the company’s safety procedures, failing to disclose to the board that he owned the OpenAI Startup Fund, and seeking to remove Toner from the board after she released “a research paper that angered him” are all examples of Altman’s misconduct.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Norwegian Startup Iris Has A Antidote For Hallucinations

June 4, 2024

Iris is a machine learning engine for scholarly research analysis. The tool is prompted by users to look for the information they require. Millions of papers are then categorized, summarized, and systematized by the system to produce insights that may be put to use.

 Iris tools enabled users to rapidly go through tens of thousands of research papers to pinpoint relevant information at the intersection of niche areas that would have taken researchers months to do manually.

Academics noticed AI’s tendency to generate inaccurate information. Researcher Michael Black, director of the prestigious Max Planck Institute for Intelligent Systems in Germany, warned the tool could “usher in an era of deep scientific fakes.”  

 Iris now plans to add extra tools to its tool. Up next is a chat assistant that will guide researchers through the system and personalizes the workflow.

Researchers Developed A Bio-inspired With AI Camera For Cars

May 29, 2024

Compared to existing automobile cameras, the combination of artificial intelligence (AI) and a revolutionary bio-inspired camera allows for 100-times faster identification of pedestrians and obstructions. The safety of automotive systems and self-driving automobiles can be considerably increased by the significant advancement in computer vision and artificial intelligence that University of Zurich researchers have made. A unique bio-inspired camera and artificial intelligence (AI) have been coupled by Daniel Gehrig and Davide Scaramuzza of the University of Zurich (UZH) Department of Informatics to create a system that can identify impediments surrounding an automobile far more quickly than existing systems and with less computer resources. Nature publishes the study. Daniel Gehrig et al, Low Latency Automotive Vision with Event Cameras, Nature (2024). DOI: 10.1038/s41586-024-07409-w

Researchers Find ChatGPT Inaccurate When Answering Computer Programming Queries

May 28, 2024

When it comes to answering questions about computer programming, ChatGPT, a popular LLM, is grossly wrong, according to a team of Purdue University computer scientists. The team explains in their work included in the Proceedings of the CHI Conference on Human Factors in Computing Systems how they took questions from the StackOverflow website, asked ChatGPT, and then gauged how accurate the system was in answering them.

The group also gave a presentation on their research at the May 11–16, 2024, Conference on Human Factors in Computing Systems (CHI 2024).

Since these apps were made available to the general public, ChatGPT and other LLMs have gained a lot of attention and popularity. Regretfully, in addition to a wealth of helpful information contained  along with a host of useful information included in many of the responses given by such apps, there are a bunch of inaccuracies. Furthermore, it is not always clear when the apps are giving wrong answers.

In this new study, the team at Purdue noted that many programming students have begun using LLMs to help write code for programming assignments, and answer queries related to programming. Example :a student could ask ChatGPT, what is the difference between a bubble sort and merge sort, or, more popularly, what is recursion?

 The researchers used questions from the StackOverflow website—it is a site built to help programmers learn more about programming by working with others in their field of interest.

Info From  Samia Kabir et al, Is Stack Overflow Obsolete? An Empirical Study of the Characteristics of ChatGPT Answers to Stack Overflow Questions, Proceedings of the CHI Conference on Human Factors in Computing Systems (2024). DOI: 10.1145/3613904.3642596

DeepFakes of Deceased Love Ones Is A Booming Business In China

May 8, 2024

AI Video

Many people are interested in using AI to interact, preserve, and bring deceased loved ones back to life while they grieve and work toward recovery. At least six companies in China are already offering these technologies, and thousands of customers have already paid for them, indicating the strength of the market there. As a matter of fact, Chinese people have long found comfort in confiding in the dead; the avatars are just the most recent expression of this cultural practice.

Some people wonder if communicating with artificial intelligence (AI) death mimics is a good way to deal with loss, and it’s unclear what the ethical and legal ramifications of this technology would be. Many folks are still uneasy with the idea as of right now. However, as CEO Sima Huapeng of Silicon Intelligence notes, “Even if only 1% of Chinese people can accept [AI cloning of the dead], that’s still a huge market.”

AI Resurrections
The technology utilized to mimic a living person and a deceased person are practically the same, making dead people’s avatars nothing more than deepfakes. An animated, talkative, and lifelike avatar is produced via diffusion models. Conversations can be generated by attaching large language models. The more information these models take in including photos, videos, audio recordings, and texts—the more closely the result will mimic that person, whether dead or alive.

China has shown to be a lucrative market for virtual twins of all stripes. For instance, the nation boasts a thriving e-commerce industry, and many live streamers are employed by consumer brands to promote their goods. These were genuine people at first, but many brands are moving to AI-cloned influencers that can stream around-the-clock, as the MIT Technology Review revealed last autumn.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Self Published Author Received Copyright Over Book With AI Generated Text.

April 18, 2024

When authoring the book, the author made considerable use of ChatGPT from OpenAI. Through her application, she sought to force the US Copyright Office to change its rules for artificial intelligence (AI) work, which typically mandate that prospective copyright holders remove machine-generated content. Following approval of her copyright application for AI Machinations: Typed Words and Tangled Webs, a Shupe’s colorful life, including her support for more inclusive gender recognition, is incorporated into the narrative. With more people integrating AI technologies into their creative work, the USCO’s response to artificial intelligence is becoming more apparent through its registration. It is one of the first works of art to be granted a copyright for the arrangement of words produced by artificial intelligence. self-published autobiographical piece on Amazon, written under the pen name Ellen Rae.

As is customary for written works, Shupe is not acknowledged as the author of the entire text in the USCO notice awarding copyright registration for her book. Rather, the “selection, coordination, and arrangement of text generated by artificial intelligence” is credited to her. This indicates that while the book itself cannot be copied without permission, individual phrases and paragraphs are not protected by copyright, therefore they might be changed or rearranged and published as a distinct book.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Two Researchers Terminated By Open AI For Leaking Information

April 12, 2024

Two researchers on a team committed to maintaining AI safe for reasoning and society were sacked by Open AI.

Although the nature of the material that the two researchers disclosed is unknown, MSPowerUser revealed that OpenAI was seeking to appoint a “insider risk investigator” in mid-February in order to prevent leaks and safeguard its ideas.

Of course, when the board fired OpenAI’s CEO, Sam Altman, in November, it created a highly public boardroom drama. However, Altman was brought back after criticism.

There had been rumors back then of internal conflict within OpenAI, purportedly over disagreements over the appropriate rate of AI development. Ilya Sutskever, the company’s top scientist, was said to be the cause of the unhappiness.

He was presumably concerned about the rate at which AI was developing. Later, Sutskever made an appearance in the BBC documentary iHuman, in which he predicted that artificial general intelligence models will “solve all the problems that we have today” but also held out the possibility that they would “create infinitely stable dictatorships.”

Sutskever’s close ally Aschenbrenner, a 19-year-old Columbia University graduate, was regarded as a rising star at OpenAI and a strong proponent of safe AI research.

The “primary fiduciary duty is to humanity,” not investors, as stated in OpenAI’s charter. However, Altman converted the AI lab into a for-profit business in 2019 and fully commercialized it following the release of ChatGPT in late 2022.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

ChatGPT Users No Longer Have To Sign Up

April 1, 2024

The company that created ChatGPT, OpenAI, said on Monday that users will no longer need to register in order to access their popular AI chatbot, GPT 3.5, for free.

The firm released an update saying, “Start using ChatGPT today for anyone that has been curious about AI’s potential but didn’t want to go through the steps to set-up an account.”

In order to access the bot, users wishing to utilize the “free-to-use AI system” currently needed to register for an OpenAI account.

It seems that creating an account is still required for individuals who wish to upgrade to the Plus membership level and get access to the more sophisticated ChatGPT-4, which features OpenAI’s DALL·E 2 text-to-image generating model.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Gemini AI Not Allowed To Answer Questions Concerning The Elections

March 12, 2024

Google is limiting the queries that its Gemini chatbot can respond to on the election in order to curb the dissemination of false information.

The tech giant discussed the limitation in a blog post on endorsing the Indian elections.

“We have started to roll out restrictions on the types of election-related queries for which Gemini will return responses,” the business stated, expressing their abundance of care on such a significant matter. “We continuously work to improve our protections, and we take our responsibility to provide high-quality information for these types of queries seriously.”

However it appears that the limitations apply outside of India. The chatbot Gemini will merely respond, “I’m still learning how to answer this question,” whether you inquire about elections that are taking place in the US or other nations.

More Here

Digital Biomarkers

March 12, 2024

Digital indicators of health, have the potential to significantly improve healthcare by enabling ongoing, remote patient monitoring, giving instant feedback, customizing treatment regimens, and promoting early disease identification. They also have the potential to improve clinical trials by enabling more effective and economical drug development processes and by providing objective, reliable measures of outcomes.

Numerous physiological, behavioral, or environmental data points, such as heart rate variability, sleep patterns, physical activity levels, voice/speech parameters, cognitive performance, and more, might be considered biomarkers. Additionally, the data used to create them comes from digital technologies including wearables, smartphones, sensors, and other software-based medical equipment.

Accepting new technologies in the medical field cutting-edge technologies—such as virtual reality and artificial intelligence—will be essential to the development of health technology in the future.

Numerous physiological, behavioral, or environmental data points, such as heart rate variability, sleep patterns, physical activity levels, voice/speech parameters, cognitive performance, and more, can be included in these biomarkers. Additionally, the data used to create them comes from digital technologies including wearables, smartphones, sensors, and other software-based medical equipment.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

The Split Between Open AI & Musk

March 6, 2024

In a joint statement released Tuesday night, five of OpenAI’s top executives refuted Elon Musk’s assertions that the AI startup had forsaken its original goals when it increased its attempts to market its technology the previous year. Additionally, it was reported that the CEO of Tesla severed connections with OpenAI after requesting total authority over the company’s for-profit division and proposing that it combine with his manufacturer of electric cars.

Musk, one of OpenAI’s original backers, filed a lawsuit against the business last week in a San Francisco court, claiming that its founders violated the organization’s founding agreement by expanding the company’s commercial reach and maintaining a tight relationship with Microsoft, a significant investor.

OpenAI published emails from 2015 to 2016 in a blog post that attempted to demonstrate Musk’s complete support for the company’s decision to establish a for-profit company.

Additionally, Musk contended that OpenAI would need to combine with Tesla in order to compete with Google. Even in that case, there is little chance of opposing Google. It simply isn’t zero,” one of the emails purportedly wrote from Musk.

CEO Sam Altman and the other OpenAI executives stated on their blog post, “We’re sad that it’s come to this with someone whom we’ve deeply admired.”

When Altman and associates founded OpenAI as a nonprofit research facility in 2015, Musk contributed more than $44 million. He subsequently stepped down from the OpenAI board just before the company created a for-profit division in 2019.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

ChatGPT vs Gemini

March 2, 2024

Two of the top generative AI programs on the market right now are Gemini and ChatGPT. Both offer simple-to-use interfaces for content creation, chat, questioning, and troubleshooting.

Even though ChatGPT and Gemini both offer helpful features, they each have unique advantages as well as room for development. Before we compare these AI-based features in-depth, let’s quickly review the main advantages of each tool:

Gemini: The best option if you’re looking for a less expensive alternative to ChatGPT that nevertheless produces excellent conversational and informative material for both business and leisure customers. In contrast to ChatGPT, Gemini offers a more responsible and transparent AI, links directly to the internet and other Google resources across all plans, and has a more efficient mechanism for managing the quality of the information.
ChatGPT: The best option for a well-known AI chatbot and content creation platform that provides both business and personal users with a variety of helpful features. In contrast to Gemini, ChatGPT is more widely used across many channels, has more third-party commercial apps integrated into it, and provides a number of more sophisticated generative AI multimodal content production features.

The comparison table:

PriceCore FeaturesEase of Use and ImplementationOutput Quality and RelevanceEnterprise Use Cases
GeminiWinnerDependent on Use CaseDependent on Use Case
ChatGPTWinnerWinnerDependent on Use CaseDependent on Use Case
credit: eWeek

More here

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Google’s Gemini & Racial Errors

March 1, 2024

credit The Verge


A significant obstacle facing artificial intelligence is the occurrence of hallucinations. In short, artificial intelligence delusion occurs when a chatbot or algorithm purposefully produces erroneous responses in order to fulfill the user’s request. The goal of an AI algorithm to please users frequently backfires, as the ethnically varied Nazis and Founding Fathers demonstrate.

Google’s Gemini did satisfy its user prompts, technically although it produced historically inaccurate responses.

Generally speaking, the best way to prevent hallucinations is to train algorithms on a variety of trustworthy datasets. Another strategy is to train algorithms using factual data and personally identifiable traits like sex and race. Naturally, this has its own set of issues, like the potential for AI to be used for racial profiling.

Alternatively, AI programmers can regulate information by imposing strict guidelines on specific themes. For example, an AI machine may be trained to preserve historical truth and refrain from deleting remarks that promote modern values. However, more laws would be required to ensure that offensive content is not created in the process of eradicating modern ideas.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

Google’s New Genie AI Text -to-Video Games

February 28, 2024

The “Genie” AI model, developed by Google’s DeepMind team, has the potential to completely transform both the gaming and creative industries.

Genie, a cutting-edge platform that can create interactive 2D video games with just an image cue or word description, represents a major advancement in artificial intelligence. Created by the Open-Endedness Team at Google DeepMind, this innovative initiative is a marriage of state-of-the-art technology with creative potential.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly

ChatGPT Histories From Other Users Resulted From An ChatGPT Account Compromised.

February 1, 2024

The user’s reported ChatGPT records indicate that his ChatGPT account was compromised. Sri Lanka was the source of the illegal logins, according to an Open AI official. According to the user, he accesses his account from Brooklyn, New York.

The spokesperson wrote, “Based on our findings, we regard it as an account takeover because it aligns with behaviors we observe where an individual is adding to a “pool” of identities that an outside community or proxy server uses to grant unrestricted access.” “The inquiry found that discussions were started recently from Sri Lanka. These discussions are taking place concurrently with Sri Lankan logins that are successful.

However the User Disagrees with Open AI’s Explanation of His Account Being Compromised

The user, reset his password after that, but he didn’t think his account was compromised. He claimed to have used a nine-character password consisting of special characters, lower- and uppercase letters. He said that he only used it for his Microsoft account. He claimed that on Monday morning, while taking a quick break from using his account, the conversation history of other users suddenly emerged all at once.

New England Journal Of Medicine ‘s AI-Focused Peer Review Journal

January 31, 2024

The articles new peer-reviewed journal from the New England Journal of Medicine is devoted to the new technology craze that will be written by actual doctors and healthcare professionals.

The journal’s idea is to demonstrate how doctors and researchers are utilizing AI in order to take advantage of new developments in healthcare and medicine.

The  initial article, article, an editorial on the journal’s founding, emphasizes the significance of conducting research that “actively shapes the ethical integration of AI into a health care system that respects patient autonomy.” A  report report concentrates on the exponential growth of AI-enabled medical devices over the previous ten years. Both reports address issues pertinent to seniors and senior living.

One-Time
Monthly
Yearly

Make a one-time donation

Make a monthly donation

Make a yearly donation

Choose an amount

$5.00
$15.00
$100.00
$5.00
$15.00
$100.00
$5.00
$15.00
$100.00

Or enter a custom amount

$

Your contribution is appreciated.

Your contribution is appreciated.

Your contribution is appreciated.

DonateDonate monthlyDonate yearly