Täällä seurataan AI-ripuligeneraattoreiden maailmanvalloitusta

[pigra senlaborulo]

neuters.de
Exclusive: Reddit in AI content licensing deal with Google
2024-02-22 04:10 - By Anna Tong, Echo Wang and Martin Coulter
2–3 minutes

Social media platform Reddit has struck a deal with Google (GOOGL.O) to make its content available for training the search engine giant's artificial intelligence models, three people familiar with the matter said.

The contract with Alphabet-owned Google is worth about $60 million per year, according to one of the sources.

The deal underscores how Reddit, which is preparing for a high-profile stock market launch, is seeking to generate new revenue amid fierce competition for advertising dollars from the likes of TikTok and Meta Platform's (META.O) Facebook.

The sources were not authorized to speak to media and declined to be identified.

Reddit and Google declined to comment.

Bloomberg previously reported Reddit's content deal without naming the buyer.

Last year, Reddit said it would charge companies for access to its application programming interface (API) - the means by which it distributes its content. The agreement with Google is its first reported deal with a big AI company.

San Francisco-based Reddit, which has been looking at a stock float for more than three years, is preparing to make its initial public offering filing this week, which would detail its financials for the first time to potential IPO investors.

The filing could be available as early as Thursday, two of the sources said.

The company, which was valued at about $10 billion in a funding round in 2021, is seeking to sell about 10% of its shares in the offering, Reuters has previously reported.

Reddit's stock market launch would mark the first IPO of a major social media company since Pinterest floated its shares in 2019.

Makers of AI models have been busy clinching deals with content owners in recent months, aiming to diversify their training data beyond large scrapes of the internet. That practice is rife with potential copyright issues as many content creators have alleged that their content was used without permission.

Founded in 2005 by web developer Steve Huffman and entrepreneur Alexis Ohanian, Reddit is known for its manifold niche discussion groups, some of which boast tens of millions of members.

https://neuters.de/technology/reddit-ai ... 024-02-22/

[pigra senlaborulo]

mikrobitti.fi
ChatGPT:n kehittäjä halusi tavaramerkin kolmelle kirjaimelle – näin kävi
Petri Ranta
~2 minutes

ChatGPT-tekstitekoälystään tunnettu OpenAI-yhtiö on kokenut pettymyksen Yhdysvalloissa, sillä maan patentti- ja tavaramerkkiviranomainen ei myöntänyt ”GPT”-termille tavaramerkin asemaa.

GPT tulee sanoista generative pre-trained transformer. Se tarkoittaa, että tekoäly generoi eli luo sisältöä, se on esikoulutettu ja se on rakennettu transformer-teknologialla, jonka Googlen tutkijat keksivät vuonna 2017.

TechCrunchin mukaan OpenAI on yrittänyt perustella tavaramerkkiasiaa sillä, että yhtiö nosti toiminnallaan GPT-termin yleiseen tietoisuuteen. Yhtiö ei kuitenkaan ollut millään tavoin ensimmäinen termiä käyttänyt taho, vaan sitä käytettiin ennestään jo useissa muissa yhtiöissä eri yhteyksissä.

GPT-termi ei siis ollut OpenAI:n keksimä, eikä se ollut myöskään sen ainoa käyttäjä. Viranomaiset päätyivät hylkäämään OpenAI:n hakemuksen GPT-tavaramerkille.

”Haettu tavaramerkki pelkästään kuvaa hakijan tuotteiden ja palveluiden ominaisuutta tai käyttötarkoitusta”, patentti- ja tavaramerkkiviranomaisen päätöksessä lukee.

OpenAI oli hakenut termille tavaramerkkiä jo aiemmin, mutta hakemus oli hylätty lokakuussa 2023. Kolmatta kertaa todennäköisesti ei tule, sillä viranomaisen asiakirjassa päätös on merkitty lopulliseksi.

Tavaramerkkihakemuksen hylkäämisen johdosta OpenAI ei voi ryhtyä oikeustoimiin, vaikka markkinoille saapuisi kuinka monta GPT-lyhenteellä varustettua tekstitekoälyä tahansa. Niitä onkin jo nyt saatavilla useita, mikä kenties rohkaisi OpenAI:ta yrittämään tavaramerkin hankkimista.

OpenAI:lla on kuitenkin GPT-termin muista käyttäjistä ainakin toistaiseksi silti julkisuudessa niskalenkkiote.

Jos keskivertoihmiseltä kysytään GPT-termiin liittyviä mielleyhtymiä, enemmistö luultavasti mainitsee OpenAI:n ja ChatGPT:n. Yhtiö luultavasti pyrkii pitämään asian näin myös jatkossa.

https://www.mikrobitti.fi/uutiset/mb/b9 ... d4a03be2bf

[Ultra-lehti]

[pigra senlaborulo]

theverge.com
Google explains Gemini’s “embarrassing” AI pictures of diverse Nazis
~3 minutes

Google has issued an explanation for the “embarrassing and wrong” images generated by its Gemini AI tool. In a blog post on Friday, Google says its model produced “inaccurate historical” images due to tuning issues. The Verge and others caught Gemini generating images of racially diverse Nazis and US Founding Fathers earlier this week.

“Our tuning to ensure that Gemini showed a range of people failed to account for cases that should clearly not show a range,” Prabhakar Raghavan, Google’s senior vice president, writes in the post. “And second, over time, the model became way more cautious than we intended and refused to answer certain prompts entirely — wrongly interpreting some very anodyne prompts as sensitive.”



A picture of Gemini results promising “some images featuring diverse US senators from the 1800s,” featuring what appear to be three women of color and an Asian American man.

This led Gemini AI to “overcompensate in some cases,” like what we saw with the images of the racially diverse Nazis. It also caused Gemini to become “over-conservative.” This resulted in it refusing to generate specific images of “a Black person” or a “white person” when prompted.

In the blog post, Raghavan says Google is “sorry the feature didn’t work well.” He also notes that Google wants Gemini to “work well for everyone” and that means getting depictions of different types of people (including different ethnicities) when you ask for images of “football players” or “someone walking a dog.” But, he says:

However, if you prompt Gemini for images of a specific type of person — such as “a Black teacher in a classroom,” or “a white veterinarian with a dog” — or people in particular cultural or historical contexts, you should absolutely get a response that accurately reflects what you ask for.

Google stopped letting users create images of people with its Gemini AI tool on February 22nd — just weeks after it launched image generation in Gemini (formerly known as Bard).

Raghavan says Google is going to continue testing Gemini AI’s image-generation abilities and “work to improve it significantly” before reenabling it. “As we’ve said from the beginning, hallucinations are a known challenge with all LLMs [large language models] — there are instances where the AI just gets things wrong,” Raghavan notes. “This is something that we’re constantly working on improving.”

https://www.theverge.com/2024/2/23/2408 ... verse-nazi

[pigra senlaborulo]

gizmodo.com
Tinder Owner Signs ChatGPT Deal. Enjoy the AI Dating Tidal Wave
Thomas Germain
~4 minutes

Match Group, the international conglomerate that owns Tinder, Hinge, OkCupid, and almost every other popular dating app, just inked a major partnership with OpenAI. The company shared only a few hazy details, saying AI will help employees with “work-related tasks.” The dating giant says it long term plan is to squeeze artificial intelligence into “literally everything” in its apps. Incorporating ChatGPT into its worker’s daily responsibilities is the first major step towards making online dating an AI-focused enterprise.

The company celebrated its new partnership with a press release written by ChatGPT and edited by Match Group’s Corporate Communications team (meaning it was written by MatchGroup’s communications team). It includes quotes supposedly penned by the AI itself, as though it has its own feelings about corporate strategy. “I’m thrilled that Match Group matched with me,” ChatGPT said. “Together, we’re not just breaking the ice; we’re melting it, and reshaping the way work gets done.”

The deal includes 1000 ChatGPT Enterprise licenses purchased for an undisclosed sum. The company says it plans to use the AI for communications, coding, design, analysis, building templates, and eliminating other repetitive tasks.

Continuing with the company’s cutesy tone, Match Group’s Chief Technology Officer Will Wu said that using AI “has been like adding a pinch of magic into our daily tasks.” Wu promises it will amount to a “full-on work culture revolution.”

What that means for Match Group’s hundreds of millions of lovelorn users is less clear. However, Match Group has been explicit that it thinks it will be good for its business and the world if just about every part of its users’ interactions is shaped and filtered through the algorithmic lens of artificial intelligence.

“I believe that AI is existential to the future of Match Group and our business,” Match Group CEO Bernard Kim said on a January earnings call. “I envision AI to be felt through the entire experience, influencing everything from profile creation to matching to connecting for dates. Literally everything.”

In one sense, that’s the status quo. A Match Group spokesperson said the company has used AI to power trust and safety efforts, enhance matching algorithms, and a variety of other purposes for most of its recent history.

But adding AI to “literally everything” is a major shift to a company that has astronomical power over our romantic lives. And so far, that shift includes a surprisingly tolerant policy towards people plugging AI into their dating profiles and letting it run amok on other users. In early February, Gizmodo reported that a man used ChatGPT to talk to 5,000 women on Tinder.

This, apparently, does not violate Match Group’s policies, including its number one AI principle: promoting authenticity. At the time, a company spokesperson told Gizmodo that “Tinder and Match Group believe AI is a great tool that we believe will make dating better.”

In fact, Match Group wants to explore building its own AI tools that could perform tasks like writing your conversation starters or suggesting date ideas. That means the next time you’re chatting with someone on a dating app, you might want to ask yourself whether you’re just flirting with a robot.

Update 02/22/2024, 5:03 p.m. ET: This story has been updated with a comment from Match Group.


https://gizmodo.com/tinder-owner-signs- ... 1851275100

[bad grankulla]

Tällaisen kattelin just ja huolestuin vähän lisää, mutta ilmeisesti tää homma ei ole tosiaan katoamassa minnekään lähiaikoina:

[38911 BASIC BYTES FREE]

Tumblrissa kiertää nyt enemmän tai vähemmän huhun asteella olevaa juttua että Tumblr olis tekemässä sopparia Midjourneyn kanssa siitä että Tumblriin postattujen / postattavien kuvien dataa vois käyttää Midjourneyn AI-treeniaineistona.

Tää on erittäin perseestä jos näin tapahtuu, koska Tumblrissa on paljon epäaktiivisia tilejä joille on postattu taidetta tai valokuvia, ja joiden omistajat ei varmaankaan seuraa tilannetta niin tarkkaan että kävisivät poistamassa teoksensa tuolta sit kun toi muutos yhtäkkiä tapahtuu.

Spoiler:

[Aloysius Kärppä]

<s>kai sä tiesit että jos pistää jotain nettiin se on siellä ikuisesti?</s>

[pigra senlaborulo]

theregister.com
Judge slaps down law firm using ChatGPT to justify six-figure trial fee
Brandon Vigliarolo
4–5 minutes

You'd think lawyers – ostensibly a clever group of people – would have figured out by now that relying on ChatGPT to do anything related to their jobs could be a bad idea, but here we are, yet again, with a judge rebuking a law firm for doing just that.

The legal eagles at New York-based Cuddy Law tried using OpenAI's chatbot, despite its penchant for lying and spouting nonsense, to help justify their hefty fees for a recently won trial, a sum the losing side is expected to pay.

NYC federal district Judge Paul Engelmayer, however, rejected the submitted amount, awarded less than half of what Cuddy requested, and added a sharp rebuke to the lawyers for using ChatGPT to cross-check the figures. The briefs basically cited ChatGPT's output to support their stated hourly rate, which does depend on things like the level and amount of research, preparation, and other work involved.

Cuddy told the court "its requested hourly rates are supported by feedback it received from the artificial intelligence tool 'ChatGPT-4,'" Engelmayer wrote in his order [PDF], referring to the GPT-4 version of OpenAI's bot.

"It suffices to say that the Cuddy Law Firm's invocation of ChatGPT as support for its aggressive fee bid is utterly and unusually unpersuasive."

For Judge Engelmayer – who also disputed the proposed charges for other reasons, including the use of "dubious resource(s)" to arrive at a final bill of $113,484.62 – the use of ChatGPT to justify steep fees was a final straw.

"As the firm should have appreciated, treating ChatGPT's conclusions as a useful gauge of the reasonable billing rate for the work of a lawyer with a particular background carrying out a bespoke assignment for a client in a niche practice area was misbegotten at the jump," Judge Engelmayer said.

One need not look far for ways in which the legal community has been led astray by generative AI of late, and Judge Engelmayer doesn't - he cites two cases from the US Court of Appeals Second Circuit (under which NYC falls) to make his case.

"In two recent cases, courts in the Second Circuit have reproved counsel for relying on ChatGPT, where ChatGPT proved unable to distinguish between real and fictitious case citations," the judge wrote, referring to the Mata v Avianca and Park v. Kim cases. Those cases involved ChatGPT being used to generate fake judicial opinions and fake authorities, respectively.

And the problem is far from confined to those examples. Even ex-Trump fixer Michael Cohen has been caught relying on bogus case law generated by AI.

In short, there's a long way to go between ChatGPT being good for a laugh and good to replace a paralegal.

Benjamin Kopp, a lawyer at Cuddy Law, told The Register his firm's use of AI wasn't anything like cases that fabricated court documents; this particular situation had nothing to do with influencing the legal process.

"The 'cross-check' referenced was a manner of saying that, in addition to support for our fees based upon the evidence provided, the rates were consistent with the range of rates and typical reasons for such rates that a parent ... might find if using ChatGPT while researching what rates to expect," Kopp told us.

Simply using ChatGPT wasn't the sole hangup the judge had with Cuddy's fee schedule; the firm apparently didn't identify the inputs it used to get ChatGPT's fee feedback or whether any of the data used was synthetic, the judge said in his order.

"The court therefore rejects out of hand ChatGPT's conclusions as to the appropriate billing rates here," Judge Engelmayer said. "Barring a paradigm shift in the reliability of this tool, the Cuddy Law Firm is well advised to excise references to ChatGPT from future fee applications."

Cuddy was ultimately awarded just $53,050.13. As to its future use of ChatGPT as a tool for anything to do with legal work, "we do not anticipate using ChatGPT barring, similar to as the court noted, a significant change or shift," Kopp said. ®


https://www.theregister.com/2024/02/24/ ... egal_fees/

[pigra senlaborulo]

Samsung on perustanut uuden kehitysorganisaation, joka sijaitsee Yhdysvaltain Piilaaksossa. Sen tehtävänä on kehittää seuraavan sukupolven AGI-piirejä.

AGI-piirit liittyvät AGI (Artificial General Intelligence) -tyyppiseen tekoälyyn. Se on autonominen tekoälyn muoto, joka selviytyy tehtävistä ihmismäiseen tapaan tai jopa paremmin.

Tekoälyn nousun aikakaudella Samsung ei ole vielä ottanut merkittäviä askelia verrattuna kilpailijoihinsa. Nyt näyttää siltä, että Samsung haluaa olla valmiina seuraavassa tekoälyyn liittyvässä harppauksessa.

Samsungin uusi organisaatio on nimeltään AGI Computing Lab. Myös Softbank ilmoitti vastikään tavoitteistaan AGI-tekoälyn saralla. AGIn nousulla voi olla mullistavia vaikutuksia, ja Softbankin toimitusjohtaja Masayoshi Son uskookin, että AGI johtaa turvallisempaan, terveellisempään ja onnellisempaan maailmaan.

https://muropaketti.com/tietotekniikka/ ... -maailman/

[pigra senlaborulo]

tomshardware.com
Jensen Huang says kids shouldn't learn to code — they should leave it up to AI
Mark Tyson
4–5 minutes

At the recent World Government Summit in Dubai, Nvidia CEO Jensen Huang made a counterintuitive break with what he thinks is a long tradition of tech CEOs advising youngsters to learn how to code. Huang argued that, even at this early stage of the artificial intelligence (AI) revolution, programming is no longer a vital skill. With coding taken care of by AI, humans can instead focus on more valuable expertise like biology, education, manufacturing, or farming, reasoned the Nvidia head.

You can see the exchange if you expand the above tweet. In the minute-long segment of Huang’s stage time shared on social media, the Nvidia CEO said that for 10-15 years almost every person sitting on a tech forum stage would have insisted that it is “vital” for young people to learn computer science, to learn how to program computers. “In fact, it’s almost exactly the opposite,” according to Huang’s counterintuitive sense.

“It is our job to create computing technology such that nobody has to program. And that the programming language is human,” Jensen Huang told the summit attendees. “Everybody in the world is now a programmer. This is the miracle of artificial intelligence.”

After his grand counterintuitive pronouncement, the Nvidia CEO surmised that people could instead learn the skills to become experts in more useful fields. Experts in domains like biology, education, manufacturing, farming, and so on could save the time they might have needed to learn computer programming for more productive pursuits. Thus, the only language people would need is the language they were born and raised to speak, and are already experts in.

However, people still need to know how and when to apply AI programming. Therefore Huang asserts at the end of the brief clip that “It is vital that we upskill everyone, and the upskilling process I believe will be delightful, surprising.”

With the above clip shared widely on social media, tech industry analyst Patrick Moorhead was stirred enough to comment. The top analyst pointed out to his Twitter / X followers, "For over 30 years, I’ve heard “XYZ will kill coding” yet we still don’t have enough programmers.” Moorhead listed several programming languages and tools that he said were supposed to kill coding – but clearly didn't.

Moorhead also drew parallels with the computer DTP revolution. He said that AI isn’t going to kill coding, but put it in the hands of more people. “Just like desktop publishing didn’t kill “creativity” it just expanded it.” While I agree that DTP and other digital art tools didn’t kill creativity, I don’t remember anyone suggesting moving from scalpels, Spray Mount, and scraps of paper to DTP would actually stunt creativity.
The impacts of AI on the (freelancing) job market - as seen so far

Only time will tell the actual impacts of the wave of AI applications that are set to appear in the coming months and years. However, a recent study of the volume of freelance work available since ChatGPT arrived was published by Bloomberry. This research shows that writing and translation freelancers have been hit hardest by AI rivals. Meanwhile, the numbers suggest software development jobs are up 6% since ChatGPT hit the scene.

https://www.tomshardware.com/tech-indus ... t-up-to-ai

[pigra senlaborulo]

mikrobitti.fi
Google varoittaa tekoälystään: älä kerro mitään arkaluontoista
Petteri Uusitalo
2–3 minutes

Tekoälyteknologioiden muodostuessa yhä jokapäiväisemmäksi osaksi työarkea ovat monet tulleet syöttäneeksi kaikenlaisiin tekstikenttiin sellaisia arkaluontoisia tietoja, joiden käsittelyssä olisi ehkä pitänyt miettiä kahteen kertaan.

Merkittävin tapaus lienee keväältä 2023, jolloin kävi ilmi, että Samsungin työntekijät olivat yrittäneet tehostaa työskentelyään ChatGPT-tekoälyllä. Seurauksena yrityksen ohjelmistokoodia vuoti julkisuuteen, ja se kielsi ulkopuolisten tekoälypalveluiden käytön työntekijöiltään.

Merkittävä rooli monella työpaikalla on myös Googlella, joka hiljattain uudelleenbrändäsi aiemmin nimellä Bard tunnetun tekoälytuotteensa Geminiksi. Gemini-sovellusten yksityisyystietojen tietokannassaan Google muistuttaakin varmuuden vuoksi käyttäjiä siitä, että Gemini kerää talteen aivan kaiken minkä käyttäjä sille syöttää.

”Älä syötä keskusteluihisi luottamuksellisia tietoja tai mitään dataa, jonka et haluaisi arvioijan näkevän tai Googlen käyttävän kehittämään tuoteitamme, palveluitamme tai koneoppimisteknologioitamme”, tietokannan johdannossa tähdennetään erikseen lihavoituna.

”Google kerää Gemini-sovellusten keskustelusi, aiheeseen liittyvät tiedot tuotteiden käytöstä, tietoa sijainnistasi sekä palautteesi”, tietokanta kertoo. ”Google käyttää tätä dataa – yksityisyydensuojakäytäntöjemme mukaisesti – Googlen tuotteiden, palveluiden ja koneoppimisteknologioiden tuottamiseen, parantamiseen ja kehittämiseen, mukaanl ukien Googlen yrityspalvelut kuten Google Cloud.”

Asiasta kirjoittava ForbesinZak Doffman huomauttaa myös, että toistaiseksi Google ei käytä Gemini-sovellusten kanssa käytettyjä keskusteluja mainosten kohdentamiseen. Mikäli asia joskus muuttuu, siitä tullaan tiedottamaan selkeästi, Google tähdentää.

Kun Gemini-sovelluksia käytetään yhdessä Googlen muiden palveluiden kanssa, nämä voivat tietysti niin ikään kerätä Gemini-sovelluksille annettuja tietoja.

Gemini-sovellusten keräämiä tietoja säilytetään oletusarvoisesti 18 kuukauden ajan. Käyttäjät voivat halutessaan vaihtaa asetuksista tämän ajan pituudeksi joko 3 tai 36 kuukautta.

https://www.mikrobitti.fi/uutiset/mb/c0 ... 1d8242744d

[Spandau Mullet]

https://techcrunch.com/2024/02/27/jolla-mind2/

Mobile OS maker Jolla is back and building an AI device

Forthcoming private cloud and AI router will power privacy-safe 'adaptive digital assistant'

On the fringes of MWC, the world’s biggest mobile phone show, TechCrunch is getting a demo of a new kind of AI device. The small black box sitting in front of us on a table in a hotel lobby in Barcelona is designed to tap into the generative AI boom — making it another example of how the wheel of innovation is spinning up again in mobile.

Its maker — Jolla, a born-again startup with the combined forces of around 40 seasoned staff behind this new push — is building what’s being tentatively described at this early product design stage as a private and secure “blackbox for your life”.

The device is intended as a private-cloud-cum-AI-router that makes the user’s data accessible for running AI queries. So this device-in-the-making is akin to the personal server concept that’s been kicking around in privacy circles for years — but reconfigured for the generative AI age.

Jne jne, en todellakaan pasteta kokonaan

[pigra senlaborulo]

[pigra senlaborulo]

Tumblr and Wordpress are preparing to sell user data to Midjourney and OpenAI, according to a source with internal knowledge about the deals and internal documentation referring to the deals.

The exact types of data from each platform going to each company are not spelled out in documentation we’ve reviewed, but internal communications reviewed by 404 Media make clear that deals between Automattic, the platforms’ parent company, and OpenAI and Midjourney are imminent.

The internal documentation details a messy and controversial process within Tumblr itself. One internal post made by Cyle Gage, a product manager at Tumblr, states that a query made to prepare data for OpenAI and Midjourney compiled a huge number of user posts that it wasn’t supposed to. It is not clear from Gage’s post whether this data has already been sent to OpenAI and Midjourney, or whether Gage was detailing a process for scrubbing the data before it was to be sent.

https://www.404media.co/tumblr-and-word ... -ai-tools/