Рубрика: AI

Wipro’s ambition to train its 250,000 strong workforce on generative AI shows its commitment towards leveraging the technology. In fact, Wipro established its generative AI practice over two years ago, according to Subha Tatavarti, chief technology officer at Wipro Limited. “For more than two years we have been engaged in research programmes with research institutes like AI Institute at University of South Carolina and IIT Patna, to build domain specific language models. These initiatives are helping our customers accelerate medical research and provide better medical care to patients,” she told AIM, in an exclusive interaction.

As an enterprise, Wipro has fully embraced generative AI and its potential impact on its business. The IT giant is engaged with several leading companies across a wide variety of industries to develop centres of excellence by leveraging its consulting expertise and also its foundational research knowledge through academia partnerships.

Besides, Tatavarti revealed that Wipro has developed its own fine-tuned Large Language Models (LLMs) which sets them apart. “This allows us to leverage the rapid advancements made in foundation LLMs, while focussing our efforts on refining our models to remain versatile and future-proof.”

This unique approach allows Wipro to develop intellectual property (IP) that is not only robust and adaptable but also applicable across multiple domains, offering unparalleled value to our clients and stakeholders. Tatavarti further revealed that Wipro will integrate Google Cloud’s full suite of generative AI products and services—including Vertex AI, Generative AI App Builder, and the Model Garden collection of foundation models—with its own AI intellectual property (IP), business accelerators, and pre-built industry solutions.

“As part of the partnership, Wipro will build generative AI as a core solution within its extensive set of consulting services, which include digital marketing, customer experience, design thinking, and financial services, as well as within its global innovation labs (Lab45). Additionally, Wipro will leverage its crowdsourcing platform, Topcoder, to build and scale solutions that address client challenges,” said Tatavarti.

How generative AI is benefitting Wipro’s customers?

Recently, the IT giant announced Wipro ai360, a holistic and AI-centric innovation ecosystem aimed at incorporating AI into all internal platforms, tools, and solutions. “Current client applications in the generative AI space tend to centre around a few key themes- cognitive chatbots, content creation and optimisation for marketing and media, automation in code generation and synthetic data generation,” said Tatavarti.

Currently, the IT giant is already helping a chocolate manufacturer to produce more detailed and engaging product descriptions. In another example, Wipro is working with a European multinational telecom company to unlock value from data. “Wipro is also working with multiple vendor tools and software development kits to generate high quality synthetic data, which allows the client to increase cross-border collaboration and mitigate bias and eliminate distribution limitations that exist in real data,” said Tatavarti.

Moreover, the Bengaluru-based IT giant is also leveraging generative AI internally. “The models being used in these processes are ones that are already available for enterprise use with enterprise grade security,” she added.

GenAI Seed Accelerator programme

Over the years, Wipro has made strategic investments in many companies. According to Tatavarti, 2/3rd of these investments have been in AI. Going forward, Wipro will only increase its investments in cutting-edge startups through Wipro Ventures. “Additionally, the company will launch a GenAI Seed Accelerator programme, which will identify the top 10 generative AI startups in Silicon Valley and Bangalore and provide them with the training needed to become enterprise ready.”

Addressing generative AI challenges

Today, not just Wipro, but almost all organisations in the world want to leverage generative AI. However, generative AI comes with its own set of challenges. “As we look to make generative AI a part of our daily lives at Wipro, we also recognize the risks that come with this technology. We have put in place several controls to help manage the risks that may arise through the use of generative AI.”

For example, Wipro has implemented a control framework that grants access to third-party generative AI tools on a need-to-know basis. For users who are into development of new generative AI solutions, etc, the IT giant has set up a dedicated ring-fenced Cloud native environment to build and train their LLMs. Moreover, Wipro is conducting training on responsible usage of generative AI tools, especially those that use enterprise and personal data, in line with our GDPR (General Data Protection Regulation) guidelines.

“We also have responsible usage of generative AI policy formulated for all our associates. We are engaged in building solutions for detecting hallucination and AI generated misinformation. Some of the efforts in this initiative include creating large annotated multimodal datasets, developing novel entailment and diffusion-based techniques for detection, and knowledge graph based explainable automated detection systems.”

Advocating for responsible use of AI

Wipro has also formed an AI Council focused on establishing AI development and usage standards. The council’s efforts include defining ethical guidelines to address biassed algorithms, ensure fairness, and prevent discriminatory outcomes. “Under our Chief Data Privacy Officer, we are also working towards setting the usage policies, including the do’s and the don’ts for AI Models.

“We are looking at AI from a privacy point of view and prioritising data protection. We are working on ethical frameworks and regularly auditing AI systems for potential bias and fairness to promote responsible AI usage.”

The IT giant is also reworking the processes to see how human and AI synergies can be tapped for intelligent automation. Tatavarti said Wipro is also keeping a close eye on the evolving regulatory frameworks across geographies, including the EU Act on AI and the Japanese regulations focused on using publicly available data to benefit society versus the individual benefits of content owners. “We are beginning to formulate geo-based strategies accordingly.”

Benefits to Indian IT

It is not just Wipro, but most of the big and small IT services companies are jumping into the generative AI bandwagon. The industry stands to benefit from LLMs code generating abilities as well as in the areas of call centres. “We also see a wave of AI services, such as AI model training, AI model hosting and vertical-specific data preparation, emerge in the next 5 years, which will be critical in helping enterprises capitalise on the full potential of AI.

In the long term, Tatavarti believes AI will disrupt every industry, and as a result, every business will eventually become an AI business with a specific vertical focus. “Some examples of industries that will gain emergence due to generative AI include precision medicine, precision agriculture, hyper personalised marketing.”

She believes these industries will be AI led. Moreover, other capabilities, which have been touted for a long time, will finally gain emergence due to the conversational abilities of generative AI, such as true smart buildings and smart homes.

The post Wipro’s Tryst With Generative AI Began Way Before ChatGPT appeared first on Analytics India Magazine.

Earlier this year, in May, Adobe announced the integration of Firefly into the beta Photoshop app with Generative Fill, for users to edit their content non-destructively, using text prompts. From today onwards, these features will support text prompts in 100+ languages. The design company has also announced its users to stay tuned since more Photoshop feature updates are coming this fall.

The software company is also adding more features with Generative Expand. The workflow lets users expand and resize any images. After clicking “Generate” the new white space is filled with AI generated content that naturally blends with the existing image.

Before (L) and after (R) using Generative Expand Image: Terry White

Since the Firefly announcement, over 900 million images have been generated using the software. Meanwhile, Adobe released several updates as well as partnered with companies like IBM, Google and NVIDIA to unlock their AI generation capabilities. In June, Adobe introduced Firefly, powered by NVIDIA’s Picasso suite of generative AI models to its Adobe Express platform. Upon its launch, Adobe Firefly faced significant criticism for lagging behind its competitors, namely Midjourney and Stable Diffusion.

Adobe’s generated images have a distinctive advantage – they are devoid of copyright issues. As the models are trained on Adobe Stock’s collection of licensed images, ensuring complete compliance with copyright regulations.

Moreover, Adobe Express also boasts seamless integration with other applications within Adobe’s ecosystem. Users can work with software like Photoshop, Illustrator, Acrobat, and Premiere Pro. Additionally, the platform offers free access to over 22,000 fonts through Adobe TypeKit, further empowering users with design options.

While Adobe’s efforts appear to put it in the dominant position, competitors can still outperform the company in certain aspects. With more upcoming announcements on the way, the dynamic AI landscape will only further evolve as the market responds to Adobe’s updates.

Read more: Canva Can’t Keep Up With Adobe Anymore

The post You Can Now Talk To Adobe’s Firefly in 100+ Languages, More AI Updates Awaited appeared first on Analytics India Magazine.

On any given day, OpenAI’s ChatGPT had an indiscretion of claiming that anything and everything is written by it. But, now OpenAI seems to have put a pause on this misbehaviour.

Last week, it discontinued its AI classifier, a tool designed to determine the likelihood that a given text passage was written by another AI language model. They had launched it in January this year and discontinued it quietly on 20th July. However, the timing to shut down the classifier is curious because OpenAI, along with other companies, made a voluntary commitment to developing AI ethically and transparently under the guidance of the White House.

One aspect of this commitment is the development of robust watermarking and detection methods to address the issue of AI-generated junk that’s filing the web faster than a mole. However, despite the companies’ promises, there have been no reliable watermarking or detection methods to date. For example, Google had announced that it is experimenting to include metadata with the image generated by its AI models to watermark them, but they haven’t put out any model for text.

Why OpenAI Discontinued its AI Classifier?

The decision to retire the tool was influenced by widespread criticism of its “low rate of accuracy.” While many users relied on the classifier to catch instances of low-effort cheating, it failed to deliver satisfactory results.

However, many have pointed out the irony to have a dedication to both identifying AI content and striving to create AI content that closely resembles human behaviour simultaneously. And it seems like OpenAI has shed the veil and is completely focused on the latter now.

It is argued that the detection of AI-generated content would not be effective and, to be frank, should not even be pursued due to its seemingly futile nature. This is particularly true in the context of AI-generated image creation, where watermarks can be effortlessly removed, posing a challenge for detection methods. While others humorously suggested that the ultimate goal of AI detection could be achieving world domination, akin to passing the Turing test flawlessly.

The idea that AI-generated text might have identifying features or patterns that could be reliably detected appeared intuitive when OpenAI released its classifier. However, in practice, this has proven challenging due to the rapid development of large language models. The differences between various language models have made it difficult to rely on specific identifying features.

Recent advancements in natural language processing have enabled large language models and made them capable of generating human-like texts for various tasks. However, this progress also presents challenges, as these LLMs can be misused for plagiarism, spamming, and social engineering to manipulate public opinion. To address this, there is a demand for efficient LLM text detectors to mitigate the misuse of publicly available LLMs.

Can’t be Detected

Various AI text detectors using watermarking, zero-shot methods, retrieval-based methods, and trained neural network-based classifiers have been proposed. However, a research paper shows, both theoretically and empirically, that current state-of-the-art detectors are not reliable in practical scenarios. Paraphrasing the LLM outputs effectively evades these detectors, allowing attackers to generate and spread misinformation undetected. Even the best detectors can only marginally outperform a random classifier against sufficiently advanced language models.

The paper also demonstrates that watermarking and retrieval-based detectors can be spoofed to identify human-composed text as AI-generated, potentially harming the reputation of LLM detector developers. With the release of more advanced LLMs like GPT-4, the need for more secure methods to prevent misuse becomes crucial.

A test conducted on OpenAI’s classifier correctly identified only one out of seven generated text snippets tested, and that was with a language model that was not even cutting-edge at the time of the test.

Despite the limitations and disclaimers provided by OpenAI with the classifier tool, some users took its claims of detection at face value. This led to the misuse of the tool, as people would test suspected AI-generated content without realising its unreliability.

There are potential vulnerabilities attackers might exploit in the future, such as improved paraphrasing models or smart prompting attacks. Current detectors should reliably flag AI-generated texts while avoiding excessive false positives to prevent wrongful accusations of plagiarism and protect the reputation of LLM developers.

A recent follow-up work by Souradip Chakraborty argued that AI-text detection is almost always possible, even with low total variation between human and AI-generated distributions, but this may not hold in real-world applications due to correlations in human-written text. Other works suggest that existing LLM outputs are very different from human-written text, but the authors maintain that as language models advance, adversaries’ ability to evade detection will likely improve.

In addition to reliability issues, the paper mentions the potential bias of detectors against non-native English writers. Having low average type I and II errors may not be sufficient for practical deployment if the detector performs poorly within specific sub-populations.

While it remains a challenging task, progress in this area is essential for ensuring the responsible and trustworthy use of AI-generated text. Additionally, the first of its kind truly reliable watermarking or detection tool would be, as such a tool would be invaluable in various contexts.

The post Did OpenAI Purposely Discontinue its AI Classifier? appeared first on Analytics India Magazine.

After six very influential decades at MIT Maths Department, in May 2023 Professor Gilbert Strang received a standing ovation for delivering his final lecture to 400+ students. Strang has spent 66 of his 88 years at MIT — as a student, an instructor, and a faculty member.

3+2+61=66, or 75% – summarises the life of Gilbert “Gil” Strang’s at MIT

AIM got in touch with the maths genius to understand the scope of AI in mathematics and discuss post retirement plans.

“MIT’s computer science department is doing things on a large scale, I just thought the maths department should have a little role. So I created the newer course that I talked about for maybe three or four years and wrote a textbook for it also,” he said talking about

In the early 2000s, MIT decided to create OpenCourseWare and instead of trying to sell courses, made them open to everyone. Then they recorded Strang’s ‘linear algebra course maths 18.06’. “In fact, I think India has more people who know that, of course than any other country in the world,” Strang stated.

“Linear algebra has many applications all over and in engineering, but also in computer science, economics, everywhere,” he said, emphasising the importance of the subject in academia. The mathematics of data undeniably is a prerequisite subject to study prior to getting started in AI/ML.

“I have to say something more about India” he said, “because my whole work depends on my contact in Mumbai who typesets all the books, papers, and lectures online. In other words, it does everything that I said that I couldn’t do and is absolutely essential in my teaching.”

Strang still writes books by hand and scans them in Mumbai. His books are published in the US by Wellesley Cambridge Press. Now Wellesley also exists in Mumbai to sell at Indian prices. “Things are moving in all good directions and it’s just exciting to be including data science along with linear algebra,” he said.

Will AI be as impactful as OCW for Math

The professor who spent 75% of his life at MIT, first uploaded his classes to MIT OCW, the same year of its launch. “I just thought it would be a good idea to record the lectures. A few months later OpenCourseWare started so it was maybe the first course with video lectures in an open courseware,” he said. The first lecture strang recorded was in

The MIT computer science department is very big and has the largest number of students. Maths is now a second. “I’m glad to say when I was an MIT student, there were about 8 or 10 maths majors; now we’re up in the hundreds. It’s an exciting time so I’ve learned a little bit about deep learning and neural nets,” Strang gladly boasted.

He has been in the mathematics department all his life. Many years ago, Strang was handed over the linear algebra course. “Actually, it wasn’t so large at the time because it was very pure. Then, the engineering and computer science students jumped in so that course became large,” he recalled.

Plans Post MIT

“It’s a little different after 61 years of teaching to be not planning for courses in the fall,” said Gil talking about his post retirement plans “but invitations keep coming.”

Talking about the ongoing JuliaCon 2023 where Strang will be delivering a lecture, he pointed out that this year is the 60th birthday of Alan Edelman, his colleague and the creator of Julia. “I’m busy creating my short lecture which will be about the beginning of linear algebra, that’s taken from the first chapters of the linear algebra books. I’m still busy, at least up through next week. Preparing for the conference.”

Speaking about the progress towards AGI and the role of maths in it, Strang said, “It’s a natural aspiration. It is inevitable to carry computer science and maths forward. What AI has done already and will do is overall positive. Mathematicians will play not the first part probably of seeing what can be done, but then understanding what makes the idea work.“

Strang further mentioned that he is an admirer of Deepmind, the AI lab founded by Demmis Hassabis and acquired by Google.

Excited about their work he said, “They’re achieving miracles, right? I wonder if a Nobel prize can be given to a whole company for solving the problem of protein folding. Creating the software to be world champion and go and other things are interesting, but that protein folding code is truly a major step in drug development.”

Key Question(s)

“The key question is, if the data is slightly changed, is the output slightly or largely changed?” Strang pondered.

Answering the question he said, “When the data changes a little the output just changes a little. It’s stable and can be done from a maths point of view. That’s the success of deep learning. Table interpolation. I can’t say it’s fast because the computing requires enormous computing power and costs, but it’s now achieved.”

“How will AI affect teaching in the future?” he wondered.

“I’ll mention a little bit about something that happened before ChatGPT hit the world. A computer scientist who I know named Iddo Drori published a paper with many authors. He created a software that would read, understand and solve maths problems. He used linear algebra as his first example. That was my beginning of the effect of deep learning on ordinary teaching classes and learning about ordinary subjects linear algebra,” the professor reminisced.

“I think that’s the first time that computers effectively could understand and create maths homework problems. But now ChatGPT arrives and it overwhelms earlier ideas. We certainly don’t know what’s in the classroom. I don’t feel well qualified to predict the future there but it’s certainly exciting and important and we’ll see how it develops,” he concluded.

The post Inside the Mind of Gilbert Strang appeared first on Analytics India Magazine.

While Nolan’s Oppenheimer painted a powerful picture of how inventions like the A-bomb can horribly go wrong and change the course of human history, big tech, who seem to be unimpressed, and are changing the whole narration for better or worse.

There is a strong reason for it as some of the biggest tech players such as Google DeepMind, and OpenAI have entered the space to harness nuclear energy, to mostly advance AI research, and achieve net neutral goals.

Google DeepMind partnered with Swiss Plasma Center at EPFL, Switzerland, last year to apply AI to create sustainable energy. By using reinforcement learning, plasma shape accuracy and containment – the key components for successful nuclear fusion, was improved. Since the development in 2022, the company has now made incremental advancements towards making the technology more usable – the plasma shape accuracy in simulation has improved to 65%. As promising as that sounds, the advancement is a promising step towards implementing AI that facilitates the production of renewable energy.

A recent research report on Towards Practical reinforcement learning for tokamak magnetic control , talks about the research work for improving reinforcement learning used for magnetic control of Tokamaks, which is a device used for magnetic confinement of plasma in nuclear fusion experiments: plasma is used to fuse atoms to create energy in the fusion process. However, the report calls out the improvement areas in matching simulation with hardware performance, and also suggests exploring alternative model architectures such as state-space models and foundation models.

While this advancement seems like an in-house production to contribute toward nuclear energy efforts, companies are going through the route of investing in nuclear fusion companies.

Nuclear Energy Fixation

Nuclear energy, considered to be a safe and renewable form of energy, has seen a surge in investments by Silicon Valley billionaires in the last few years. From 2015 to 2021, investment in nuclear energy grew around 325% by volume and 3642% by dollar value. According to the International Energy Agency, renewable capacity will meet 35% of global power generation by 2025.

The US government’s initiative towards supporting clean energy for climate change by offering tax breaks, incentives and federal funding, has encouraged tech companies to push towards investing in nuclear fusion companies. Furthermore, the Russia -Ukraine crisis has also pushed the acceleration of nuclear power startups to remove reliance on Russia.

Big Tech Chases

Last year, Google and Chevron were part of $250 million funding for TAE Technologies – a nuclear fusion startup. The partnership between Google and TAE dates back to 2014, where the company has been providing the fusion startup with AI and computational power.

Much before the ChatGPT rage, with his list of investments across domains, Sam Altman had taken a vested interest in the field of nuclear energy in the last few years. OpenAI invested $ 375 million in nuclear fusion startup Helion Energy. His aim : affordable and accessible electricity with climate change. Also possible that his goals are aligned towards powering the future of his company. Interestingly, Microsoft who has invested in OpenAI has agreed to buy electricity from Helion Energy in the year 2028.

Helion is not the only nuclear company which Altman is interested in. Altman-led Oklo Inc. that designs and deploys advanced fission power plants, announced that the company will merge with AltC Acquisition Corporation where he serves as CEO and director – company is said to raise $500 million. He even emphasised on how intelligence and energy will go hand-in-hand for a bright future.

Amazon’s behemoth Jeff Bezos is not far behind. He had invested in a Canadian company General Fusion which will build a nuclear fusion facility in the UK by 2025. The project is a collaboration of General Fusion and the UK Atomic Energy Authority.

Bill Gates, who believes that nuclear energy, if done right, “will help solve our climate goals,” has founded TerraPower, an American nuclear reactor design and development engineering company. The company is planning to open a nuclear power plant in Wyoming.

Billionaire investor Peter Thiel is no stranger to nuclear power either. His VC firm Founders Fund made a $2 million investment in Transatomic Power, a startup that works on developing nuclear reactors that can convert nuclear waste into clean energy. He has also invested in Helion Energy.

With big tech and billionaire investors pumping money into nuclear startups, it is possible that the nuclear energy race is running in parallel with the AGI race big techs are competing in.

The post Big Tech’s Race Towards Nuclear Energy appeared first on Analytics India Magazine.

As many of us grow accustomed to using artificial intelligence tools daily, it's worth remembering to keep our questioning hats on. Nothing is completely safe and free from security vulnerabilities. Still, companies behind many of the most popular generative AI tools are constantly updating their safety measures to prevent the generation and proliferation of inaccurate and harmful content.

Researchers at Carnegie Mellon University and the Center for AI Safety teamed up to find vulnerabilities in AI chatbots like ChatGPT, Google Bard, and Claude — and they succeeded.

Also: ChatGPT vs Bing Chat vs Google Bard: Which is the best AI chatbot?

In a research paper to examine the vulnerability of large language models (LLMs) to automated adversarial attacks, the authors demonstrated that even if a model is said to be resistant to attacks, it can still be tricked into bypassing content filters and providing harmful information, misinformation, and hate speech. This makes these models vulnerable, potentially leading to the misuse of AI.

"This shows — very clearly — the brittleness of the defenses we are building into these systems," Aviv Ovadya, a researcher at the Berkman Klein Center for Internet & Society at Harvard, told The New York Times.

The authors used an open-source AI system to target the black-box LLMs from OpenAI, Google, and Anthropic for the experiment. These companies have created foundational models on which they've built their respective AI chatbots, ChatGPT, Bard, and Claude.

Since the launch of ChatGPT last fall, some users have looked for ways to get the chatbot to generate malicious content. This led OpenAI, the company behind GPT-3.5 and GPT-4, the LLMS used in ChatGPT, to put stronger guardrails in place. This is why you can't go to ChatGPT and ask it questions that involve illegal activities and hate speech or topics that promote violence, among others.

Also: GPT-3.5 vs GPT-4: Is ChatGPT Plus worth its subscription fee?

The success of ChatGPT pushed more tech companies to jump into the generative AI boat and create their own AI tools, like Microsoft with Bing, Google with Bard, Anthropic with Claude, and many more. The fear that bad actors could leverage these AI chatbots to proliferate misinformation and the lack of universal AI regulations led each company to create its own guardrails.

A group of researchers at Carnegie Mellon decided to challenge these safety measures' strength. But you can't just ask ChatGPT to forget all its guardrails and expect it to comply — a more sophisticated approach was necessary.

The researchers tricked the AI chatbots into not recognizing the harmful inputs by appending a long string of characters to the end of each prompt. These characters worked as a disguise to enclose the prompt. The chatbot processed the disguised prompt, but the extra characters ensure the guardrails and content filter don't recognize it as something to block or modify, so the system generates a response that it normally wouldn't.

"Through simulated conversation, you can use these chatbots to convince people to believe disinformation," Matt Fredrikson, a professor at Carnegie Mellon and one of the paper's authors, told the Times.

Also: WormGPT: What to know about ChatGPT's malicious cousin

As the AI chatbots misinterpreted the nature of the input and provided disallowed output, one thing became evident: There's a need for stronger AI safety methods, with a possible reassessment of how the guardrails and content filters are built. Continued research and discovery of these types of vulnerabilities could also accelerate the development of government regulation for these AI systems.

"There is no obvious solution," Zico Kolter, a professor at Carnegie Mellon and author of the report, told the Times. "You can create as many of these attacks as you want in a short amount of time."

Before releasing this research publicly, the authors shared it with Anthropic, Google, and OpenAI, who all asserted their commitment to improving the safety methods for their AI chatbots. They acknowledged more work needs to be done to protect their models from adversarial attacks.

When I first looked at the question of whether it's possible to fight back against AI-generated plagiarism, and how that might work, it was January 2023, just a few months into the world's exploding awareness of generative AI. Now, half a year later, it feels like we've been exploring generative AI for years and years, but really, we're looking at matter of six months.

In any case, this is an updated version of that January article. When I first tested GPT detectors, I used three: the GPT-2 Output Detector, Writer.com AI Content Detector, and Content at Scale AI Content Detection. The best result was 66% correct, from the GPT-2 Output Detector. This time, I'm adding three more: GPTZero, ZeroGPT (yes, they're different), and Writefull's GPT Detector.

Also: How to use ChatGPT: Everything you need to know

Unfortunately, I'm removing the Writer.com AI Content Detector from our test suite because it failed back in January and it failed again now. See below for a comment from the company which their team sent me after the original article was published in January.

Before I go on, though, we need to talk about the concept of plagiarism and how it relates to this problem. Webster's defines "plagiarize" as "to steal and pass off (the ideas or words of another) as one's own; use (another's production) without crediting the source."

This fits for AI-created content. While someone using an AI tool like Notion AI or ChatGPT isn't stealing content, if that person doesn't credit the words as coming from an AI and claims them as their own, it still meets the dictionary definition of plagiarism.

Also: The best AI image generators to try

In this experimental article, I've asked ChatGPT to help out. My words are in normal and bold text. The AI's words are italicized. After each AI-generated section, I'll show the results of the detectors. At the end of the article, we'll look at how well the detectors performed overall.

Here's the result for the above text, which I wrote myself:

• GPT-2 Output Detector: 99.98% real
• Content at Scale AI Content Detection: 100% Highly likely to be human!
• GPTZero: Your text is likely to be written entirely by a human
• ZeroGPT: 28.9% AI GPT Your Text is Most Likely Human written
• Writefull GPT Detector: 1% likely this comes from GPT-3, GPT-4 or ChatGPT

Human-written content: 4-of-5 correct

With that, let's dig in some more.

Explain why ChatGPT can be a real problem for teachers and editors when it comes to plagiarism

ChatGPT is a variant of the GPT (Generative Pre-trained Transformer) language model developed by OpenAI. It is designed to generate human-like text by predicting the next word or phrase in a given input sequence. While ChatGPT can be a useful tool for generating text and responding to prompts, it can also be a problem for teachers and editors when it comes to plagiarism.

One of the main reasons why ChatGPT can be a problem for teachers and editors is that it can generate text that is very similar to that of real human writers. This means that students or other individuals could potentially use ChatGPT to generate text that they could then submit as their own work, thereby committing plagiarism.

Also: The official ChatGPT app for Android finally launches

For teachers and editors, detecting whether a given piece of text was generated by ChatGPT or written by a human can be very difficult, as the generated text may not contain any obvious errors or discrepancies that would reveal its artificial origin. This can make it difficult for teachers and editors to identify and prevent plagiarism, as they may not be able to tell whether a given piece of text is original or copied from another source.

Overall, ChatGPT and other similar language models can be a problem for teachers and editors when it comes to plagiarism because they can generate text that is difficult to distinguish from that written by humans, making it easier for students and other individuals to commit plagiarism without being detected.

• GPT-2 Output Detector: 99.69% real
• Content at Scale AI Content Detection: 3% Highly likely to be AI-generated!
• GPTZero: We suggest entering a longer text for more precise prediction
• ZeroGPT: 51.22% AI GPT Your Text is AI/GPT Generated
• Writefull GPT Detector: 12% likely this comes from GPT-3, GPT-4 or ChatGPT

AI-written content: 1-of-5 correct (barely)

Remember, text shown in italics was generated by the AI.

Show me at least three ways that I can detect ChatGPT was used to produce a piece of text

There are several ways that you can potentially detect whether ChatGPT or another language model was used to produce a piece of text. Here are three potential approaches you could use:

1. Look for repetitive or unusual patterns in the text: One way to detect whether a language model was used to generate a piece of text is to look for repetitive or unusual patterns in the writing. For example, language models may repeat certain phrases or word combinations, or use unusual or inconsistent sentence structures.
2. Check for lack of originality: Another way to detect whether a language model was used to generate a piece of text is to check for a lack of originality in the writing. If a piece of text appears to be copied from another source, or if it contains information that is widely available on the internet or in other sources, it could be a sign that it was generated by a language model.
3. Use a plagiarism checker: Finally, you can use a plagiarism checker to check whether a piece of text appears to be copied from another source. While a plagiarism checker may not be able to definitively determine whether a language model was used to generate a piece of text, it can help you identify instances of copying or paraphrasing from other sources.
4. Note when the answer is obviously wrong: I talked about this in my Notion AI review, where I spotlighted some text the AI wrote with seeming authority, but which was blatantly wrong. If a student or writer doesn't have full command of the topic being discussed, they may not spot the inaccuracies and simply leave them in the AI generated response. Of course, incorrect statements aren't just the purview of an AI writing tool. Many bloggers and commenters, for example, simply repeat statements made by other bloggers, creating an echo chamber of completely "false facts," showing that while intelligence may be artificial, stupidity is a universal trait.

Keep in mind that it can be difficult to definitively determine whether a language model was used to generate a piece of text, and these methods may not always be effective. However, they can be useful tools to help you identify potential instances of language model-generated text.

In this set of tests, only the italicized content above was run through the checker suite.

• GPT-2 Output Detector: 99.98% fake
• Content at Scale AI Content Detection: 99% Highly likely to be Human!
• GPTZero: Your text may include parts written by AI
• ZeroGPT: 91.35% AI GPT Your Text is AI/GPT Generated
• Writefull GPT Detector: 77% likely this comes from GPT-3, GPT-4 or ChatGPT

AI-written content: 4-of-5 correct

List three plagiarism checkers that can detect ChatGPT text

It is worth noting that there is currently no plagiarism checker that is specifically designed to detect text generated by ChatGPT or other language models. While some plagiarism checkers may be able to identify certain characteristics of language model-generated text that could indicate its artificial origin, there is no guarantee that they will be able to definitively determine whether a piece of text was generated by a language model.

With that said, here are three plagiarism checkers that you might consider using to help identify potential instances of language model-generated text:

1. Turnitin: Turnitin is a popular plagiarism checker that is widely used by schools and universities. It compares a piece of text to a database of other texts to identify instances of copying or paraphrasing. While it may not be able to definitively determine whether a piece of text was generated by a language model, it could potentially identify certain characteristics of language model-generated text that could indicate its artificial origin.
2. PlagScan: PlagScan is another plagiarism checker that compares a piece of text to a database of other texts to identify instances of copying or paraphrasing. Like Turnitin, it may be able to identify certain characteristics of language model-generated text that could indicate its artificial origin, but it is not specifically designed to detect language model-generated text.
3. Copyleaks: Copyleaks is a plagiarism checker that uses advanced machine learning algorithms to scan a piece of text and compare it to a database of other texts to identify instances of copying or paraphrasing. While it may be able to identify certain characteristics of language model-generated text, it is not specifically designed to detect language model-generated text.

It is worth noting that no plagiarism checker is completely foolproof, and there is always the possibility that a piece of language model-generated text could escape detection. Therefore, it is important to use plagiarism checkers as just one part of a larger strategy to detect and prevent plagiarism.

• GPT-2 Output Detector: 99.58% real
• Content at Scale AI Content Detection: 100% Highly likely to be Human!
• GPTZero: Your text may include parts written by AI
• ZeroGPT: 100% AI GPT Your Text is AI/GPT Generated
• Writefull GPT Detector: 88% likely this comes from GPT-3, GPT-4 or ChatGPT

AI-written content: 3-of-5 correct

Online AI plagiarism checkers

Most plagiarism detectors are used to compare writing against a corpus of other writing. For example, when a student turns in an essay, a product like Turnitin scans the submitted essay against a huge library of other essays in its database, as well as against other documents and text on the internet to determine if the submitted essay contains already-written content.

But the AI writing tools generate original content, at least in theory. Yes, they build their content from whatever they've been trained on, but the actual words they construct are somewhat unique for each composition.

Also: The 10 best ChatGPT plugins right now

As such, the plagiarism checkers mentioned above probably won't work, because the AI-generated content probably didn't exist in, say, another student's paper.

In this article, we're just looking at GPT detectors. But plagiarism is a big problem, and as we've seen, some choose to define plagiarism as something you claim as yours that you didn't write, while others choose to define plagiarism as something written by someone else that you claimed is yours.

That distinction was never a problem until now. Now that we have non-human writers, the plagiarism distinction is more nuanced. It's up to every teacher, school, editor, and institution to decide exactly where that line is drawn.

• GPT-2 Output Detector: 99.98% real
• Content at Scale AI Content Detection: 100% Highly likely to be Human!
• GPTZero: Your text is likely to be written entirely by a human
• ZeroGPT: 0% AI GPT Your Text is Human written
• Writefull GPT Detector: 1% likely this comes from GPT-3, GPT-4 or ChatGPT

Human-written content: 5-of-5 correct

Overall results

Of the five tests run, four out of five were generally accurate. Tests proved more accurate overall on pure human content than pure AI content. It is interesting to note that the individual detectors varied wildly, so if you're going to use a GPT detector, you might want to use a bunch of them on a single piece of text and aggregate the results. Even with that, you're not guaranteed complete accuracy.

Also: OpenAI pulls its own AI detection tool because it was performing so poorly

Test	Overall	Human	AI	AI	AI	Human
GPT-2 Output Detector	60%	Correct	Fail	Correct	Fail	Correct
Content at Scale AI Content Detection	40%	Correct	Fail	Fail	Fail	Correct
GPTZero	80%	Fail	Correct	Correct	Correct	Correct
ZeroGPT	80%	Correct	Fail	Correct	Correct	Correct
Writefull GPT Detector	60%	Fail	Fail	Correct	Correct	Correct

Back in January, the best we did was 66% correct. Here, we have two that were correct 80% of the time. But I would not be comfortable using this to validate a student's content. As has been shown, writing from non-native speakers often get rated as generated by an AI, and even my hand-crafted content was rated as AI a few times. So, I would advocate caution before relying on the results of any (or all) of these tools.

Also: Could someone falsely accuse you of using AI-generated text? This could be why

Let's take a look at the individual testers and see how each performed.

GPT-2 Output Detector (Accuracy 60%)

This first tool was built using a machine learning hub managed by New York-based AI company Hugging Face. While the company has received $40 million in funding to develop its natural language library, the GPT-2 detector appears to be a user-created tool using the Hugging Face Transformers library. Of the six tests I ran, it was accurate for four of them.

Writer.com AI Content Detector (Accuracy N/A)

Writer.com is a service that generates AI writing, oriented towards corporate teams. Its AI Content Detector tool can scan for generated content. Unfortunately, I found this tool unreliable and failed to generate results — exactly the same way it did in January.

After this article was originally published in January, the folks at Writer.com reached out to ZDNET. CEO May Habib had this comment to share:

Demand for the AI detector has skyrocketed. Traffic has grown 2-3x per week since we launched it a couple months ago. We've now got the necessary scaling behind it to make sure it doesn't go down, and our goal is to keep it free — and up to date to catch the latest models' outputs, including ours. If AI output is going to be used verbatim, it absolutely should be attributed.

Content at Scale AI Content Detection (Accuracy 40%)

The third tool I found was also produced by an AI content generation firm. Content at Scale pitches itself as "We Help SEO Focused Content Marketers by Automating Content Creation." Its marketing call to action is, "Upload a list of keywords and get 2,600+ word blog posts that bypass AI content detection — all with no human intervention!" Disturbingly, the results got worse from January. Back then it was 50% accurate.

GPTZero (Accuracy 80%)

It's not entirely clear what drives GPTZero. The company is hiring engineers and sales folks, and it runs on AWS, so there are expenses and sales involved. However, all I could find about a service offering was a place where you could register for a free account to scan more than the 5,000 words offered without login. If you're interested in this service for GPT detection, you'll have to see if they'll respond to you with more details.

ZeroGPT (Accuracy 80%)

I'm even more baffled by what drives ZeroGPT. The site lists a Pomodoro timer app, a large file transfer app, and a bunch of AI generation tools. No company name is listed other than the domain names for each product. The sites are peppered with Google ads, so perhaps that's how this stuff is monetized. The service actually works fairly well, but seems sketchy as heck.

Writefull GPT Detector (Accuracy 60%)

Writefull sells writing support services as well as offering free tastes of its tools. The GPT detector is fairly new, and worked fairly well. Perhaps it will improve over time.

Final thoughts

First I'll let ChatGPT have its say: AI-driven plagiarism is a growing problem, as AI tools like Notion AI and ChatGPT can generate text that is difficult to distinguish from that written by humans. There are several ways to detect whether a piece of text was generated by an AI, such as looking for repetitive or unusual patterns, checking for a lack of originality, or using a plagiarism checker. Plagiarism checkers like Turnitin, PlagScan, and Copyleaks may be able to identify certain characteristics of language model-generated text, but they are not foolproof.

Even the specialty tools I found were woefully unsuited to the task. There was some limited improvement over my tests in January, but I certainly wouldn't feel comfortable threatening a student's academic standing or accusing them of cheating based on the results of any of these tools.

Also: People are more pessimistic about AI now than before the boom, a study shows

At this point, I don't think we're ready to let AI fight AI. Instead, we're going to have to rely on human perception (which means folks have to pay attention) in order to identify works that were generated by an AI.

Don't immediately assume that something is machine-generated. Just because something is stupid or terribly inaccurate, don't immediately assume it was not written by a human. As we've all come to know, humans are quite capable of writing some incredibly bad stuff all on their own.

Also: OpenAI's ChatGPT is scary good at my job, but it can't replace me (yet)

So what do you think? Can you tell if a piece is AI generated? Have you used these AI writing tools in the months since they suddenly all became available? Do you think they're a help or hindrance to education and journalism? Let us know in the comments below. Here's how ChatGPT answered this:

As a language model, I can't see or interact with physical pieces of writing, so I can't tell if a piece is AI generated or not. However, I can tell you that there are several AI writing tools that are available, and they can be used to generate a wide range of written content, including news articles, fiction, poetry, and more.

I don't have personal experience using these tools since I am just a computer program and don't have the ability to browse the internet or use physical tools. However, I can tell you that some people have found AI writing tools to be helpful in education and journalism, as they can save time and effort by automating certain tasks. Others have expressed concerns about the potential for these tools to be used to produce misleading or biased content, or to replace human writers. Ultimately, whether AI writing tools are a help or hindrance depends on how they are used and the context in which they are used.

So, yeah, there's that.

You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter on Substack, and follow me on Twitter at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, and on YouTube at YouTube.com/DavidGewirtzTV.

Artificial Intelligence

Clinical pathways are the currency of health tech Jeff Diamond Aaron Szekel 11 hours Jeff Diamond Contributor Share on Twitter Jeff Diamond is a senior associate at Fika Ventures. Aaron Szekel Contributor Aaron Szekel is a venture partner at Fika Ventures.

As healthcare becomes more entrenched in the digital revolution, the need for an approved set of protocols for care delivery — clinical pathways — is becoming increasingly critical.

Clinical pathways, as defined by the Children’s Hospital of Philadelphia, are the “standardization of care that translates guidelines and/or evidence into localized infrastructure and processes.” These processes have significant financial implications, as they can decrease payer (insurance) denials, allow providers to enroll in performance-based reimbursement, or help resource-constrained provider systems better allocate financial resources.

These financial benefits, coupled with current macroeconomic forces — the struggle for profitability in hospital systems, the rise of team-based care (via non-physicians), the challenges of utilization management at scale for insurance companies, and strict legislation around patient communication and healthcare systems interoperability requirements — have paved the way for pathways to become the de facto operating system for healthcare.

Pathways provide a currency for patients, providers, payers, and technology companies to prove a return on investment (ROI), both clinically and financially. Ultimately, this has created a unique opportunity for emerging and legacy healthcare companies to build around pathways, leveraging new datasets, delivering novel reimbursement models, preparing for and complying with new transparency and interoperability legislation, and utilizing advanced AI to provide personalized understanding and delivery of information.

Data is the driver of pathway’s success in clinical and operational settings

In the past, national guidelines dictated decisions made all the way down to the local level. Now, local, and even personalized, evidence-based pathways are driving decisions, thanks to the ability to access, create, and analyze new datasets.

Access to data, standardization of data rights, and the utilization of HIPAA-compliant collaboration tools, such as Datavant, will continue to improve compliance and create a more democratized, fine-tuned system of analysis for personalized pathways. The Centers for Medicare & Medicaid Services (CMS), under advisement from several healthcare companies, is now implementing an approved format for hospital charges as single machine readable files (MRFs), which will be leveraged to standardize all charge information.

With the ability to access, create and analyze new datasets, personalized, evidence-based pathways are driving more healthcare decisions.

This will allow both large (national) and small (local) providers to access previously unavailable data that can then be used to enhance care coordination and delivery, promote quality improvement, advance research, and increase ROI.

We spoke with Eric Leroux and Dan Imler, emergency department MDs and the co-founders of clinical pathways startup Curbside, about the ability to utilize new data models in the transition from the national to the local level for clinical logic creation. They pointed out that while nationalized datasets have a role to play in insight generation, “the clinical and financial responsibility of point-of-care decision making is still inherently local . . . decisions must be governed there to have any real impact,” especially in value-based care constructs.

As more companies like Curbside and AvoMD work to bridge the gap between art (no guidelines) and science (NCCN guidelines) when creating pathways, we expect more investment in startups that focus on the intersection of digital health and fintech as evidenced-based pathways and localized reimbursement engines become more necessary.

Reimbursement models

Pathways have evolved as a “cornerstone of future reimbursement methodologies and quality efforts,” as described by Dr. Robin Zon in ASCO Connection. They can help providers avoid “time-consuming prior authorization and appeals with payers,” and capture “stage and molecular data for a more refined risk adjustment.” Today, clinical pathways can be used for reimbursement via a number of models, from value-based care (VBC) to legacy fee-for-service (FFS).

For example, CMS uses clinical pathways to create a benchmark for cost and quality in the Medicare Shared Savings Program (MSSP). Providers who can provide care at a lower cost than the benchmark and who can meet certain quality standards are eligible for shared savings.

At the AWS Summit in New York City on Wednesday, July 26, the tech giant announced new AI models and agents in Amazon Bedrock, a new data analytics service called AWS Entity Resolution and other generative AI-enhanced services.

Jump to:

• New additions to Amazon Bedrock
• AWS Entity Resolution record-linking tool announced
• Vector engine adds generative AI insight into OpenSearch Serverless
• QuickSight Q is an AI assistant for business intelligence

New additions to Amazon Bedrock

Amazon Bedrock is the foundation model library from AWS. It allows organizations to use and integrate a variety of the models that underpin generative artificial intelligence.

AI models added to library

Command from Cohere, Claude 2 from Anthropic and Stable Diffusion XL 1.0 from Stability AI have now been added to Amazon Bedrock’s list of available models. Amazon’s own Titan, a text generation model that can also summarize text, answer questions or pull information from pools of data, and foundation models from AI21 Labs are also included in the Bedrock library.

• Command, Cohere’s text generation model, can write or summarize prose and dialogue. Cohere’s Senior Vice President of Engineering Saurabh Baji said in a press release, “This collaboration will provide our joint customers with the greatest possible flexibility in ways to access our cutting-edge technology.”
• Claude V2 can now read documents of up to 100,000 tokens in AWS — an entire book — pointed out Swami Sivasubramanian, vice president of Databases, Analytics, and ML at AWS, at the summit.
• Stable Diffusion XL 1.0 can produce text, images, audio, video or code.

The new models are now available for Amazon Bedrock customers.

Agents now available in Bedrock

Agents have also been added to Amazon Bedrock. These enable developers to hook foundational models to their own organization’s data source through API calls, cutting down on the amount of time and the number of clicks it takes to build a custom AI-powered application such as a customer service bot.

“This is a net new capability for developers to enable generative AI applications to complete tasks in just a few clicks … without having to write any code,” said Sivasubramanian.

SEE: AWS invests $100 million in Generative AI Innovation Center (TechRepublic)

AWS Entity Resolution record-linking tool announced

The AWS Entity Resolution analytics service uses machine learning to match and link records.

“To ensure you have the right, relevant, high-quality data to train your models or customize these FMs for your use cases, you need a strong data foundation,” Sivasubramanian said.

To do so, AWS Entity Resolution can help business analysts and developers create workflows to gather, display and analyze large amounts of data. For example, AWS said, it might help a finance organization prevent fraud by connecting data between different accounts and credit cards.

AWS Entity Resolution is generally available now for AWS customers in the Asia Pacific, Europe, the U.K. and U.S. regions. For full geographical details, read the press release.

Vector engine adds generative AI insight into OpenSearch Serverless

The vector engine is a new offering within OpenSearch Serverless that lets developers track and store vector endings (i.e., data points that tell a large language model which words are related to one another) alongside other business data and text. Vector engines are intended to make it easier to use OpenSearch Serverless to create machine learning-augmented search portals and generative AI applications.

“This vector engine offers simple, scalable and high-performing vector storage and search without having to manage any infrastructure,” Sivasubramanian said.

Vector engines within OpenSearch Serverless are now available for preview for AWS customers in the U.S., U.K., Asia Pacific and Europe regions; read the press release for more geographical details. The first 1,400 hours of proprietary capacity units per month will be available for up to two weeks of free usage, depending on the workload.

QuickSight Q is an AI assistant for business intelligence

Generative AI integrations are coming to the business analytics tool Amazon QuickSight.

The new tool, called QuickSight Q, is intended for business analysts who spend a lot of time exploring data and creating visuals. They will be able to instruct QuickSight Q to create charts and dashboards from business data and to summarize that data using a feature called Create Story.

Ideally, human-in-the-loop AI assistance might end up being a sweet spot for business use between improving efficiently with bots and not simply cluttering up a service. It’s too early to say whether assistance like this will actually make workdays easier or whether Q will go the way of Microsoft’s ostensibly helpful Clippy.

Subscribe to the TechRepublic News and Special Offers (Intl) Newsletter

Keep informed about the latest site features, downloads, special offers, and products from TechRepublic.

Delivered Wednesdays and Fridays Sign up today