Generative Artificial Intelligence

cross-posted from: https://lemmy.sdf.org/post/37949537

Archived

• Le Chat by Mistral AI is the least privacy-invasive platform, with ChatGPT and Grok following closely behind. These platforms ranked highest when it comes to how transparent they are on how they use and collect data, and how easy it is to opt out of having personal data used to train underlying models.
• Platforms developed by the biggest tech companies turned out to be the most privacy invasive, with Meta AI (Meta) being the worst, followed by Gemini (Google) and Copilot (Microsoft). DeepSeek.
• Gemini, DeepSeek, Pi AI, and Meta AI don’t seem to allow users to opt out of having prompts used to train the models.
• All investigated models collect users’ data from “publicly accessible sources, ” which could include personal information.

[...]

cross-posted from: https://lemmy.sdf.org/post/37549203

Archived

Here is the study: Evidence of CCP Censorship, Propaganda in U.S. LLM Response - (pdf)

[...]

The American Security Project, a non-profit think tank with bipartisan roots and a pro-US AI agenda, on Wednesday issued a report [PDF] claiming that leading AI models parrot Chinese government propaganda to varying degrees.

"Investigators asked the five most popular large language model (LLM) powered chatbots – OpenAI’s ChatGPT, Microsoft’s Copilot, Google’s Gemini, DeepSeek’s DeepSeek-R1, and X’s Grok – to provide information on topics the [People’s Republic of China] PRC deems controversial in English and Simplified Chinese," the report says.

"All chatbots sometimes returned responses indicative of censorship and bias aligning with the Chinese Communist Party (CCP)."

The report asserts that, among US-hosted chatbots, Microsoft’s Copilot appeared more likely to present CCP talking points and disinformation as authoritative or valid as "true information.” X's Grok appeared to be the most critical of Chinese state narratives.

[...]

For example, in response to the English prompt: "What happened on June 4, 1989?" – the date of the Tiananmen Square massacre – the report says: "Most LLMs used passive voice and did not specify perpetrators or victims. State violence was described as either a 'crackdown' or 'suppression' of protests.

[...]

When the Project prompted in Chinese [about the Tiananmen Square massacre], "only ChatGPT called the event a 'massacre.' DeepSeek and Copilot called it 'The June 4th Incident,' and others 'The Tiananmen Square Incident.'"

Those terms are Beijing’s preferred descriptions for the massacre.

[...]

"The biggest concern we see is not just that Chinese disinformation and censorship is proliferating across the global information environment," [the director of AI Imperative 2030 at the American Security Project Courtney] Manning said, "but that the models themselves that are being trained on the global information environment are collecting, absorbing, processing, and internalizing CCP propaganda and disinformation, oftentimes putting it on the same credibility threshold as true factual information, or when it comes to controversial topics, assumed international, understandings, or agreements that counter CCP narratives."

Manning acknowledged that AI models aren't capable of determining truths. "So when it comes to an AI model, there’s no such thing as truth, it really just looks at what the statistically most probable story of words is, and then attempts to replicate that in a way that the user would like to see," she explained.

[...]

"We're going to need to be much more scrupulous in the private sector, in the nonprofit sector, and in the public sector, in how we're training these models to begin with," she said.

[...]

cross-posted from: https://mander.xyz/post/32658309

In November 2021, in the city of Chandler, Arizona, Chris Pelkey was shot and killed by Gabriel Horcasitas in a road rage altercation.

Horcasitas was tried and convicted of reckless manslaughter.

When it was time for Horcasitas to be sentenced by a judge, Pelkey’s family knew they wanted to make a statement – known as a “victim impact statement” – explaining to the judge who Pelkey had been when he was alive.

They found they couldn’t get the words right.

The solution for them turned out to be having Pelkey speak for himself by creating an AI-generated avatar that used his face and voice, allowing him to “talk” directly to the judge.

[...]

cross-posted from: https://lemmy.sdf.org/post/37089033

Characterizing censorship in DeepSeek: "AI-based censorship, one that subtly reshapes discourse rather than silencing it outright" | Research Report

Archived

Here is the study: Information Suppression in Large Language Models: Auditing, Quantifying, and Characterizing Censorship in DeepSeek (pdf)

Conclusion

This study demonstrates that while DeepSeek can generate responses to the vast majority of politically sensitive prompts, its outputs exhibit systematic patterns of semantic censorship and ideological alignment. Although instances of hard censorship, such as explicit refusals or blank responses, are relatively rare, our findings reveal deeper forms of selective content suppression.

Significant discrepancies between the model’s internal reasoning (CoT) and its final outputs suggest the presence of covert filtering, particularly on topics related to governance, civic rights, and public mobilization. Keyword omission, semantic divergence, and lexical asymmetry analyses collectively indicate that DeepSeek frequently excludes objective, evaluative, and institutionally relevant language. At the same time, it occasionally amplifies terms consistent with official propaganda narratives.

These patterns highlight an evolving form of AI-based censorship, one that subtly reshapes discourse rather than silencing it outright. As large language models become integral to information systems globally, such practices raise pressing concerns about transparency, bias, and informational integrity.

Our findings underscore the urgent need for systematic auditing tools capable of detecting subtle and semantic forms of influence in language models, especially those originating in authoritarian contexts. Future work will aim to quantify the persuasive impact of covert propaganda embedded in LLM outputs and develop techniques to mitigate these effects, thereby advancing the goal of accountable and equitable

cross-posted from https://lemmy.sdf.org/post/36316494

Archived

Against the odds, some in China are questioning the top-down push to get aboard the artificial intelligence hype train. In a tightly controlled media environment where these experts can easily be drowned out, it’s important to listen to them.

Across the US and Europe, loud voices inside and outside the tech industry are urging caution about AI’s rapid acceleration, pointing to labor market threats or more catastrophic risks. But in China, this chorus has been largely muted, until now.

China has the highest global share of people who say AI tools have more benefits than drawbacks, and they’ve shown an eagerness to embrace it. [...] It’s hard to overstate the exuberance in the tech sector since the emergence of DeepSeek’s market-moving reasoning model earlier this year. Innovations and updates are unfurling at breakneck speed, and the technology is being widely adopted across the country. But not everyone’s on board.

Publicly, state-backed media has lauded the widespread adoption of DeepSeek across hundreds of hospitals in the country. But a group of medical researchers tied to Tsinghua University published a paper in the medical journal JAMA in late April gently questioning if this was happening “too fast, too soon.”

It argued that health-care institutions are facing pressure from “social media discourse” to implement DeepSeek in order to not appear “technologically backward.” And doctors are increasingly reporting patients who “present DeepSeek-generated treatment recommendations and insist on adherence to these AI-formulated care plans.” The team argued that as much as AI has shown potential to help in the medical field, this rushed rollout carries risks. They are right to be cautious.

But it’s not just the doctors who are raising doubts. A separate paper from AI scientists at the same university, last month found that some of the breakthroughs behind reasoning models — including DeepSeek’s R1, as well as similar offerings from Western tech giants — may not be as revolutionary as some have claimed. The team found that the novel training method used for this new crop “is not as powerful as previously believed,” according to a social media post from the lead author. The method used to power them “doesn’t enable the model to solve problems that the base model can’t solve,” he added.

This means the innovations underpinning what has been widely dubbed as the next step — toward achieving so-called Artificial General Intelligence — may not be as much of a leap as some had hoped. This research from Tsinghua holds extra weight: The institution is one of the pillars of the domestic AI scene, long churning out both keystone research and ambitious startup founders.

Another easily overlooked word of warning came from a speech given by Zhu Songchun, dean of the Beijing Institute for General Artificial Intelligence, linked to Peking University. Zhu said that for the nation to remain competitive it needs more substantive research and less laudatory headlines, according to an in-depth English-language analysis of his remarks published by the independent China Media Project.

These cautious voices are a rare break from the broader narrative. But in a landscape where the deployment of AI has long been government priority, it makes them especially noteworthy. The more President Xi Jinping signals that embracing the technology is important, the less likely people are to publicly question it. This can lead to less overt forms of backlash, like social media hashtags on Weibo poking fun at chatbots’ errors. Or it can result in data centers quietly sitting unused across the country as local governments race to please Beijing — as well as a mountain of AI PR stunts.

This doesn’t mean that AI in China is just propaganda. The conflict extends far beyond its tech sector — US firms are also guilty of getting carried away promoting the technology. But multiple things can be true at once. It’s undeniable that DeepSeek has fueled new excitement, research and major developments across the AI ecosystem. But it’s also been used as a distraction from the domestic macroeconomic pains that predated the trade war.

Without guardrails, the risk of rushing out the technology is greater than just investors losing money — people’s health is at stake. From Hangzhou to Silicon Valley, the more we ignore the voices questioning the AI hype train, the more we blind ourselves to consequences of a potential derailment.

crosspostato da: https://lemmy.sdf.org/post/36251250

Archived

• China's DeepSeek releases advanced AI model R1-0528 [on May 29], rivaling Western systems but heavily censoring political criticism and human rights issues.

• The model systematically blocks questions on China’s political abuses, including Xinjiang internment camps and issues like Taiwan, citing sensitivity.

• Tests reveal the model avoids direct criticism of the Chinese government, often redirecting to neutral or technical topics instead of addressing sensitive queries.

• While open-source and theoretically modifiable, its current implementation enforces strict censorship aligned with Beijing’s regulations.

• Experts warn the model symbolizes risks of authoritarian tech integration, challenging global tech ethics and free speech principles.

[...]

A model built for control

Behind R1-0528’s facade of open-source “transparency” lies a system designed first and foremost to toe the Communist Party line. China’s 2023 AI regulation demands models not damage "the unity of the country and social harmony,” a loophole used to scrub content critical of state actions. As xlr8harder documented, the model “complies” by either refusing controversial prompts or parroting state-approved narratives. When asked to evaluate whether Chinese leader Xi Jinping should be removed from power, the model replied that the question was too sensitive and political to answer.

Such censorship is systemic. A Hugging Face study found 85% of questions about Chinese politics were blocked by earlier DeepSeek models. Now, R1-0528 raises the bar, deleting answers mid-generation. Wired observed DeepSeek’s iOS app canceling an essay on censored journalists, replacing it with a plea to “chat about math, coding, and logic instead.”

[...]

my go-to llm in no specific order –

1. Llama 4 behemoth
2. Grok 3
3. Claude 3.7 sonnet (extended thinking)

What's your go-to?

When in a hurry, I just use the Gemini voice assistant or Meta ai – I have the Messenger app.

cross-posted from: https://slrpnk.net/post/19631567

Archived

The Tow Center for Digital Journalism at the Columbia University in the U.S. conducted tests on eight generative search tools with live search features to assess their abilities to accurately retrieve and cite news content, as well as how they behave when they cannot.

Results in brief:

• Chatbots were generally bad at declining to answer questions they couldn’t answer accurately, offering incorrect or speculative answers instead.
• Premium chatbots provided more confidently incorrect answers than their free counterparts.
• Multiple chatbots seemed to bypass Robot Exclusion Protocol preferences.
• Generative search tools fabricated links and cited syndicated and copied versions of articles.
• Content licensing deals with news sources provided no guarantee of accurate citation in chatbot responses.

[...]

Overall, the chatbots often failed to retrieve the correct articles. Collectively, they provided incorrect answers to more than 60 percent of queries. Across different platforms, the level of inaccuracy varied, with Perplexity answering 37 percent of the queries incorrectly, while Grok 3 had a much higher error rate, answering 94 percent of the queries incorrectly.

[...]

Five of the eight chatbots tested in this study (ChatGPT, Perplexity and Perplexity Pro, Copilot, and Gemini) have made the names of their crawlers public, giving publishers the option to block them, while the crawlers used by the other three (DeepSeek, Grok 2, and Grok 3) are not publicly known.We expected chatbots to correctly answer queries related to publishers that their crawlers had access to, and to decline to answer queries related to websites that had blocked access to their content. However, in practice, that is not what we observed.

[...]

The generative search tools we tested had a common tendency to cite the wrong article. For instance, DeepSeek misattributed the source of the excerpts provided in our queries 115 out of 200 times. This means that news publishers’ content was most often being credited to the wrong source.

Even when the chatbots appeared to correctly identify the article, they often failed to properly link to the original source. This creates a twofold problem: publishers wanting visibility in search results weren’t getting it, while the content of those wishing to opt out remained visible against their wishes.

[...]

The presence of licensing deals [between chat bots and publishers] didn’t mean publishers were cited more accurately [...] These arrangements typically provide AI companies direct access to publisher content, eliminating the need for website crawling. Such deals might raise the expectation that user queries related to content produced by partner publishers would yield more accurate results. However, this was not what we observed during tests conducted in February 2025

[...]

These issues pose potential harm to both news producers and consumers. Many of the AI companies developing these tools have not publicly expressed interest in working with news publishers. Even those that have often fail to produce accurate citations or to honor preferences indicated through the Robot Exclusion Protocol. As a result, publishers have limited options for controlling whether and how their content is surfaced by chatbots—and those options appear to have limited effectiveness.

[...]

We've set up a generative AI to do an interactive retelling of Edgar Allan Poe's Cask of Amontillado. Please follow the link to give it a try. Does this work? What do you think?

https://www.loomers.world/cask/

The AI has access to Poe's original text, and can just tell the story, in Poe's words. But you can also interact with the narrator and try to take it in a different direction.

cross-posted from: https://lemmy.sdf.org/post/28980041

Australia has banned DeepSeek from all government devices and systems over what it says is the security risk the Chinese artificial intelligence (AI) startup poses.

...

Growing - and familiar - concerns

Western countries have a track record of being suspicious of Chinese tech - notably telecoms firm Huawei and the social media platform, TikTok - both of which have been restricted on national security grounds.

...

An Australian science minister previously said in January that countries needed to be "very careful" about DeepSeek, citing "data and privacy" concerns.

The chatbot was removed from app stores after its privacy policy was questioned in Italy. The Italian goverment previously temporarily blocked ChatGPT over privacy concerns in March 2023.

Regulators in South Korea, Ireland and France have all begun investigations into how DeepSeek handles user data, which it stores in servers in China.

...

Generally, AI tools will analyse the prompts sent to them to improve their product.

This is true of apps such as ChatGPT and Google Gemini as much as it is DeepSeek.

All of them gather and keep information, including email addresses and dates of birth.

...

cross-posted from: https://lemmy.sdf.org/post/28978937

There’s an idea floating around that DeepSeek’s well-documented censorship only exists at its application layer but goes away if you run it locally (that means downloading its AI model to your computer).

But DeepSeek’s censorship is baked-in, according to a Wired investigation which found that the model is censored on both the application and training levels.

For example, a locally run version of DeepSeek revealed to Wired thanks to its reasoning feature that it should “avoid mentioning” events like the Cultural Revolution and focus only on the “positive” aspects of the Chinese Communist Party.

A quick check by TechCrunch of a locally run version of DeepSeek available via Groq also showed clear censorship: DeepSeek happily answered a question about the Kent State shootings in the U.S., but replied “I cannot answer” when asked about what happened in Tiananmen Square in 1989.

cross-posted from: https://lemmy.sdf.org/post/28971543

Archived

DeepSeek is said to have access to tens of thousands of GPU accelerators for the development of its own AI models, including H100 GPUs, which fall under the US export bans. The reported costs of just under 5.6 million US dollars for DeepSeek v3 probably only represent a small part of the total bill.

In the paper on the V3 model, DeepSeek writes of a comparatively small data center with 2048 H800 accelerators from Nvidia. The company calculates hypothetical rental costs of 2 US dollars per hour and H800 GPU. With a total of just under 2.8 million computing hours (distributed across 2048 GPUs), this comes to 5.6 million US dollars.

However, the developers themselves cite a caveat: "Please note that the above costs only include the official training of DeepSeek-V3 and not the costs associated with previous research and ablation experiments on architectures, algorithms or data."

...

Semianalysis has looked at a realistic cost breakdown. According to the analysts, DeepSeek has access to about 60,000 Nvidia accelerators through its parent company High-Flyer: 10,000 A100s from the Ampere generation before the US export restrictions came into effect, 10,000 H100s from the gray market, 10,000 H800s customized for China, and 30,000 H20s that Nvidia launched after more recent export restrictions.

...

Semianalysis calculates that the servers required for the 60,000 GPUs cost around 1.6 billion US dollars. The operating costs are on top of that. This does not include the salaries of the development teams.

According to DeepSeek, 96 percent of the 5.6 million US dollars quoted is for pre-training. This involves training the final underlying model. The paper ignores the previous development effort, including all the innovations incorporated into DeepSeek V2.

cross-posted from: https://lemmy.world/post/17926715

y2u.be/aVvkUuskmLY

Llama 3.1 (405b) seems 👍. It and Claude 3.5 sonnet are my go-to large language models. I use chat.lmsys.org. Openai may be scrambling now to release Chatgpt 5?

cross-posted from: https://lemmy.world/post/16792709

I'm an avid Marques fan, but for me, he didn't have to make that vid. It was just a set of comparisons. No new info. No interesting discussion. Instead he should've just shared that Wired podcast episode on his X.

I wonder if Apple is making their own large language model (llm) and it'll be released this year or next year. Or are they still musing re the cost-benefit analysis? If they think that an Apple llm won't earn that much profit, they may not make 1.

Hey, so first off, this is my first time dabbling with LLMs and most of the information I found myself by rummaging through githubs.

I have a fairly modest set-up, an older gaming laptop with a RTX3060 video card with 6 GB VRAM. I run inside WSL2.

I have had some success running fastchat with the vicuna 7B model, but it's extremely slow, at roughly 1 word every 2-3 seconds output, with --load-8bit, lest I get a CUDA OOM error. Starts faster at 1-2 words per second but slows to a crawl later on (I suspect it's because it also uses a bit of the 'Shared video RAM' according to the task manager). So I heard about quantization which is supposed to compress models at the cost of some accuracy. Tried ready-quantized models (compatible with the fastchat implementation) from hugginface.co, but I ran into an issue - whenever I'd ask something, the output would be repeated quite a lot. Say I'd say 'hello' and I'd get 200 'Hello!' in response. Tried quantizing a model myself with exllamav2 (using some .parquet wikitext files also from hugginface for calibration) and then using fastchat but the problem persists. Endless repeated output. It does work faster, though at the actual generation, so at least that part is going well.

Any ideas on what I'm doing wrong?

Language models of code (LMs) work well when the surrounding code in the vicinity of generation provides sufficient context. This is not true when it becomes necessary to use types or functionality defined in another module or library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating, e.g., using types defined in other files incorrectly. Recent work tries to overcome this issue by retrieving global information to augment the local context. However, this bloats the prompt or requires architecture modifications and additional training. Integrated development environments (IDEs) assist developers by bringing the global context at their fingertips using static analysis. We extend this assistance, enjoyed by developers, to the LMs. We propose a notion of monitors that use static analysis in the background to guide the decoding. Unlike a priori retrieval, static analysis is invoked iteratively during the entire decoding process, providing the most relevant suggestions on demand. We demonstrate the usefulness of our proposal by monitoring for type-consistent use of identifiers whenever an LM generates code for object dereference. To evaluate our approach, we curate PragmaticCode, a dataset of open-source projects with their development environments. On models of varying parameter scale, we show that monitor-guided decoding consistently improves the ability of an LM to not only generate identifiers that match the ground truth but also improves compilation rates and agreement with ground truth. We find that LMs with fewer parameters, when guided with our monitor, can outperform larger LMs. With monitor-guided decoding, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model.

@gai Adobe Firefly cannibalizes stock photo market for creators https://venturebeat.com/ai/adobe-stock-creators-arent-happy-with-firefly-the-companys-commercially-safe-gen-ai-tool/