Jump to content

All things AI

Ditto

By Ditto
in AI in Education & Enterprise

 Share

Recommended Posts

Ditto

Ditto

Posted
  • Author
Posted
Not me, but I see on their website, the case studies page mentions they partnered with LGfL, so hopefully someone will be along soon?
Bankesy

Bankesy

Posted
Posted

I know of TeachMateAI because my partner works in a primary and they use it, some of you may have heard of ict_mrp on social media and he had a fair bit to do with the creation of TeachMateAi as I understand it.

 

He gave a talk at BETT last year about the platform and did mention the merger with LGfL, he even gave away free signed copies of his book and stopped for a selfie :)

  • Thanks 1
  • 2 weeks later...
6Foot2

6Foot2

Posted
Posted

Link: Apple AI tool transcribed the word 'racist' as 'Trump' - BBC News

 

Apple says it is working to fix its speech-to-text tool after some social media users found that when they spoke the word "racist" into their iPhones it typed it out as "Trump."

 

The tech giant has suggested the issue with its Dictation service has been caused by a problem it has distinguishing between words with an "r" in them.

 

"We are aware of an issue with the speech recognition model that powers Dictation and we are rolling out a fix today," an Apple spokesperson said.

 

However an expert in speech recognition told the BBC this explanation was "just not plausible."

 

Peter Bell, professor of speech technology at the University of Edinburgh, said it was more likely that someone had altered the underlying software that the tool used.

 

Videos shared online, external show people speaking the word "racist" into the Dictation tool.

 

Sometimes it is transcribed correctly - but on other occasions it is turned into "Trump", before being quickly restored to the correct word.

 

The BBC has not been able to replicate the mistake, suggesting Apple's fix is already taking effect.

 

Prof Bell said Apple's explanation of phonetic overlap did not make sense because the two words were not similar enough to confuse an artificial intelligence (AI) system.

 

Speech-to-text recognition models are trained by inputting clips of real people speaking alongside an accurate transcript of what they say.

 

They are also taught to understand words in context - for example, they could distinguish the word "cup" from "cut" if it was within the phrase "a cup of tea".

 

Prof Bell says the situation with Apple is unlikely to be a genuine mistake with its data because its English language model would be trained on hundreds of thousands of hours of speech, which should give it a high level of accuracy.

 

For "less well-resourced languages" he said it could be an AI training issue.

 

But he said in this case: "it probably points to somebody that's got access to the process."

 

A former Apple employee who worked on its AI assistant Siri told the New York Times:, external "This smells like a serious prank."

 

Apple had to row back on another AI-powered feature last month after complaints from the BBC and other news organisations.

 

It suspended its AI summaries of news headlines after it displayed false notifications on stories - including one where it said tennis player Rafael Nadal had come out as gay.

 

The company announced yesterday it would be investing $500bn (£395bn) in the US over the next four year, including on a large data centre in Texas to power Apple Intelligence.

 

The company's chief executive Tim Cook also said it may have to change its policies on diversity, equity and inclusion (DEI) after President Donald Trump has called for an end to DEI programmes.

6Foot2

6Foot2

Posted
Posted

Link: UK watchdog drops competition review of Microsoft’s OpenAI partnership | Microsoft | The Guardian

 

The UK’s competition watchdog will not hold a formal investigation into Microsoft’s partnership with the startup behind the artificial intelligence chatbot ChatGPT, stating that while the $2.9tn (£2.3tn) tech company has “material influence” over OpenAI it does not control it.

 

The Competition and Markets Authority (CMA) said Microsoft, OpenAI’s biggest financial backer with a $13bn investment, acquired material influence over the San Francisco-based business in 2019 but did not exercise de facto control over it – and therefore did not meet the threshold for an official inquiry.

 

The decision follows expressions of disquiet over the appointment of the former boss of Amazon UK, Doug Gurr, as the CMA’s interim chair. The organisation’s chief executive, Sarah Cardell, has also said the CMA does not want to create a “chilling effect” on business confidence, amid pressure from the UK government on regulators to produce pro-growth proposals.

 

The CMA’s executive director for mergers, Joel Bamford, said: “We have found that there has not been a change of control by Microsoft from material influence to de facto control over OpenAI. Because this change of control has not happened, the partnership in its current form does not qualify for review under the UK’s merger control regime.”

 

However, Bamford added that the decision should “not be read as the partnership being given a clean bill of health on competition concerns”.

 

The CMA started investigating the OpenAI relationship after the dramatic sacking and reinstatement of Sam Altman as OpenAI’s chief executive – over a hectic few days in November 2023, Microsoft announced it had hired Altman, only for him to rejoin the startup.

 

The CMA highlighted a recent reduction in OpenAI’s reliance on Microsoft for computer power – a key factor in operating an AI business – as an influence over its decision.

 

A Microsoft spokesperson said the partnership with OpenAI and its continued evolution “promote competition, innovation, and responsible AI development … We welcome the CMA’s conclusion, after careful and prudent consideration of the commercial realities, to close its investigation.”

 

Last year, the CMA decided not to investigate Amazon’s investment in the AI firm Anthropic and further Microsoft partnerships with the AI firms Mistral and Inflection.

 

Microsoft recently contributed to a funding round that raised $6.6bn for OpenAI and valued the business at $157bn. OpenAI is run by a non-profit board but has a for-profit subsidiary, in which Microsoft is the biggest backer, with returns to investors and employees capped.

 

Despite concerns over Gurr’s appointment and the CMA’s efforts to avoid a “chilling effect” on the economy, the CMA did target big tech with investigations in January, the month that Gurr was appointed. Google is being investigated over its dominance in internet search and search advertising, while it is also conducting a separate inquiry on the impact of Apple and Google’s mobile platforms on consumers and businesses.

 

In January, Microsoft said the CMA was making a “fundamental mistake” in its investigation into the cloud market, after the watchdog said the tech firm was making it harder for Google and Amazon to compete effectively for customers in cloud computing, the term for providing IT services such as data storage and computing power over the internet with a pay-as-you-go pricing structure.

Ditto

Ditto

Posted
  • Author
Posted
ChatGPT (GPT-4) regularly reports it can not access external sites, it works on it's own knowledge. This seems like a significant weakness. Does it update it's knowledge day by day. How does that compare to other AI engines?
dmj

dmj

Posted
Posted
ChatGPT (GPT-4) regularly reports it can not access external sites, it works on it's own knowledge. This seems like a significant weakness. Does it update it's knowledge day by day. How does that compare to other AI engines?

 

I don't have this problem; ChatGPT and Gemini will both summarise external sites for me.

(is this because I have company paid GPT?)

Ditto

Ditto

Posted
  • Author
Posted
I don't have this problem; ChatGPT and Gemini will both summarise external sites for me.

(is this because I have company paid GPT?)

Could be. I use free version but it is still super useful.

dmj

dmj

Posted
Posted

yeah. Confirmed. If I log out I get:

It seems I can't access the page directly due to restrictions. However, if you could provide the key content or details from the page, I can help summarize or explain it for you!

 

I guess they are trying to claw back all the money they sank into it.

  • Thanks 1
6Foot2

6Foot2

Posted
Posted

Link: Grandmother gets X-rated message after Apple AI fail - BBC News

 

A woman from Dunfermline has spoken of her shock after an Apple voice-to-text service mistakenly inserted a reference to sex - and an apparent insult - into a message left by a garage.

 

Louise Littlejohn, 66, received a voicemail message on Wednesday from a Lookers Land Rover garage in Motherwell inviting her to an event.

 

An artificial intelligence (AI) powered service offered by Apple turned it into a text message which - to her surprise - asked if she been "able to have sex" before calling her a "piece of ****".

 

Mrs Littlejohn told BBC News: "Initially I was shocked - astonished - but then I thought that is so funny. The text was obviously quite inappropriate.

 

"The garage is trying to sell cars, and instead of that they are leaving insulting messages without even being aware of it. It is not their fault at all."

 

Apple and the garage both declined to comment.

 

An expert has told the BBC the AI system may have struggled in part because of the caller's Scottish accent, but far more likely factors were the background noise at the garage and the fact he was reading off a script.

 

Mrs Littlejohn said she initially thought the call was a scam, but then recognised that the associated number was from the Motherwell area.

 

She had bought a car from the same garage a few years ago.

 

The BBC has listened to the audio left by the garage worker and confirmed it was a conventional business call.

 

The transcription is so jumbled it's hard to decipher where it went wrong, but the reference to "sex" may have in fact been when the caller mentioned the "sixth" of March.

 

The BBC has removed the name of the garage employee:

 

"Hi Mrs Littlejohn, it is ____ here from Lookers Land Rover in Lanarkshire. I hope you are well. Just a wee call to see if you have received your invite to our new car INAUDIBLE event that we do have on between the sixth and tenth of March.

 

"Just a wee call to see if it is something you were looking to come along to, and to see if we can confirm an appointment slot that would be suitable for yourself. If it is something you would be interested in, feel free to give me a call on ____, ask for myself ____ INAUDIBLE. Thank you."

 

What went wrong?

 

On Apple's website, it details how their voicemail transcription, external is limited to voicemails in English received on an iPhone with iOS 10 operating system or later, and the transcription "depends on the quality of the recording".

 

Peter Bell, a professor of speech technology at the University of Edinburgh, listened to the message left for Mrs Littlejohn.

 

He suggested it was at the "challenging end for speech-to-text engines to deal with".

 

He believes there are a number of factors which could have resulted in rogue transcription:

 

  • The fact it is over the telephone and, therefore, harder to hear
  • There is some background noise in the call
  • The way the garage worker speaks is like he is reading a prepared script rather than speaking in a natural way

"All of those factors contribute to the system doing badly, " he added. "The bigger question is why it outputs that kind of content.

 

"If you are producing a speech-to-text system that is being used by the public, you would think you would have safeguards for that kind of thing."

 

Did a Scottish accent make a difference?

 

Many people in Scotland will remember a sketch from the BBC comedy show Burnistoun, when two Scottish men get trapped after the voice-activated lift can't understand their accents.

 

It has hundreds of millions of views online, and tapped into a feeling some Scots have about voice technology.

 

In 2016, Scots were asked to work with voice recognition technology in a bid to help mobile phones decipher the accent better.

 

For Prof Bell, the Scottish accent may have had an impact here along with all other factors, but - with ideal audio conditions - problems that technology will have with the Scottish accent are "a thing of the past".

 

However, this not the first time that Apple's speech tools have recently hit the headlines for the wrong reasons.

 

A few weeks ago, the tech giant said it was working to fix its speech-to-text tool after some social media users found that when they spoke the word "racist" into their iPhones it typed it out as "Trump".

 

Apple also suspended its AI summaries of news headlines after it displayed false notifications on stories.

6Foot2

6Foot2

Posted
Posted

Link: McDonald’s Gives Its Restaurants an AI Makeover - WSJ

 

McDonald’s is giving its 43,000 restaurants a technology makeover, starting with internet-connected kitchen equipment, artificial intelligence-enabled drive-throughs and AI-powered tools for managers.

 

The goal? To drive better experiences for its customers and workers who today contend with issues ranging from broken machines to wrong orders, according to Brian Rice, the Chicago-based burger giant’s chief information officer.

 

“Our restaurants, frankly, can be very stressful. We have customers at the counter, we have customers at our drive-through, couriers coming in for delivery, delivery at curbside. That’s a lot to deal with for our crew,” Rice said in an interview. “Technology solutions will alleviate the stress.”

 

The investments are coming at a key time for McDonald’s, whose U.S. sales were sluggish in January—reflective of a broader slump in the fast-food industry. McDonald’s said consumers remain pressured, particularly low-income diners and families. The restaurant is hoping that a better tech-enabled experience will help it deliver on its goal of growing its loyalty patrons from 175 million to 250 million by 2027.

 

To accomplish that, McDonald’s tapped Google Cloud in late 2023 to bring more computing power to each of its restaurants—giving them the ability to process and analyze data on-site. The setup, known as edge computing, can be a faster, cheaper option than sending data to the cloud, especially in more far-flung locations with less reliable cloud connections, said Rice.

 

Rice joined McDonald’s in 2022 from Cardinal Health, where he also served as chief information officer. Before his tenure at the healthcare-supplies distributor, Rice served in senior technology positions at cereal maker Kellogg, candy-maker Mars and carmaker General Motors.

 

Edge computing will enable applications like predicting when kitchen equipment—such as fryers and its notorious McFlurry ice cream machines—is likely to break down, Rice said. The burger chain said its suppliers have begun installing sensors on kitchen equipment that will feed data to the edge computing system and give franchisees a “real-time” view into how their restaurants are operating. AI can then analyze that data for early signs of a maintenance problem.

 

McDonald’s is also exploring the use of computer vision, the form of AI behind facial recognition, in store-mounted cameras to determine whether orders are accurate before they’re handed to customers, he said.

 

“If we can proactively address those issues before they occur, that’s going to mean smoother operations in the future,” Rice added.

 

McDonald’s began rolling out edge computing to some U.S. restaurants in 2024, the company said, and plans to include more this year. It declined to say how many restaurants currently have edge cloud connectivity, and when the process will be completed.

 

Additionally, the ability to tap edge computing will power voice AI at the drive-through, a capability McDonald’s is also working with Google’s cloud-computing arm to explore, Rice said. The company has been experimenting with voice-activated drive-throughs and robotic deep fryers since 2019, and ended its partnership with International Business Machines to test automated order-taking at the drive-through in 2024.

 

Edge computing will also help McDonald’s restaurant managers oversee their in-store operations. The burger giant is looking to create a “generative AI virtual manager,” Rice said, which handles administrative tasks such as shift scheduling on managers’ behalf. Fast-food giant Yum Brands’ Pizza Hut and Taco Bell have explored similar capabilities.

 

McDonald’s is an early mover in its industry when it comes to using edge computing, but it’s by no means alone, according to Jose Gomes, a managing director of retail and consumer goods at Google Cloud. The technology is expected to be used by any sector with distributed physical locations, including hospitals and factories.

 

Despite its first-mover advantage, McDonald’s will still face challenges including cost and the difficulty of rolling out the same technology across franchises and corporate-owned locations, said Sandeep Unni, a retail analyst at market research and IT consulting firm Gartner. McDonald’s declined to say how much it is spending on its technology initiatives.

 

But, compared with some of its quick-service restaurant peers, McDonald’s has been relatively aggressive at investing in new digital technologies, Unni said. That, combined with the vast amount of data it has collected on its customers, gives the fast-food giant a leg up on figuring out how to improve customer loyalty.

 

AI will be able to help McDonald’s tailor its promotions and offers by using customer data such as prior purchasing history, and even linking it with weather data, Rice said. “A customer who we know loves our sweet treats could get an offer through the app for a McFlurry on a hot summer day,” he said.

sigma

sigma

Posted
Posted (edited)

AI Search Has A Citation Problem

 

https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php

 

Building on our previous research, the Tow Center for Digital Journalism 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.

We found that…

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.

Our findings were consistent with our previous study, proving that our observations are not just a ChatGPT problem, but rather recur across all the prominent generative search tools that we tested

. Edited by sigma
Ditto

Ditto

Posted
  • Author
Posted
No, but for any app, that has unusual good reviews. I can see from a teachers perspective why there is interest Even if it does half of what is suggested, I'd imagine it is worth the $99 per year - but who's budget! Does it rely on you being a Google Workspace tool - it looks very much like an app developed for that platform.
Ditto

Ditto

Posted
  • Author
Posted

Deep within the FAQs there is a couple of helpful bits - see https://www.briskteaching.com/faq#:~:text=Privacy%20%26%20Security,-Protecting%20student%20and

 

Within that section there is a link to a Google Form to get help on achieving IT Approval - this is the Google Form https://docs.google.com/forms/d/e/1FAIpQLScfzmU53SHspDG9oZ1pgMw1-RpTD5_13XoZ_3iKTEKSLGYyww/viewform?pli=1

 

You'll see the statement "If you live outside of the Unites States, we are unable to accommodate international requests at this time, but are working on additional solutions. Thank you for using and promoting Brisk at your school/ district."

 

I'm not sure if that means English schools will hit a bump or whether they just can't/won't support IT request for approval assistance. It reminds me of when I wanted to make use of Clever Badges. Some English schools did early on, but UK support ultimately was withdrawn, if I recall correctly, to do with GDPR concerns.

 

It would need a careful review of the privacy policy and get DPO sign off, and reassurance from the provider, if you do go forward.

  • Thanks 1
Bankesy

Bankesy

Posted
Posted

We recently had a Careers Fair here and a Tom Hudson from Google Deepmind was in to talk to the students but I didn't get to catch him.

 

I see LGfL SchoolProtect is going to start blocking AI Chat & Image Generators by default.

6Foot2

6Foot2

Posted
Posted

Link: AI-driven weather prediction breakthrough reported | Artificial intelligence (AI) | The Guardian

 

Quote

A single researcher with a desktop computer will be able to deliver accurate weather forecasts using a new AI weather prediction approach that is tens of times faster and uses thousands of times less computing power than conventional systems.

 

Weather forecasts are currently generated through a complex set of stages, each taking several hours to run on bespoke supercomputers, requiring large teams of experts to develop, maintain and deploy them.

 

Aardvark Weather provides a blueprint to replace the entire process by training an AI on raw data from weather stations, satellites, weather balloons, ships and planes from around the world to enable it to make predictions.

 

This offers the potential for vast improvements in forecast speed, accuracy and cost, according to research published on Thursday in Nature from the University of Cambridge, the Alan Turing Institute, Microsoft Research and the European Centre for Medium-Range Weather Forecasts (ECMWF).

 

Richard Turner, a professor of machine learning at the University of Cambridge, said the approach could be used to quickly provide bespoke forecasts for specific industries or locations, for example predicting temperatures for African agriculture or wind speeds for a renewable energy company in Europe.

 

This contrasts to traditional weather prediction systems where creating a customised system takes years of work by large teams of researchers, while supercomputers take hours to process measurements from the real world in order to build forecasting models.

 

“This is a completely different approach to what people have done before. The writing’s on the wall that this is going to transform things, it’s going to be the new way of doing forecasting,” Turner said. He said the model would eventually be able to produce accurate eight-day forecasts, compared with five-day forecast at present, as well as hyper-localised predictions.

Dr Scott Hosking, the director of science and innovation for environment and sustainability at the Alan Turing Institute, said the breakthrough could “democratise forecasting” by making powerful technologies available to developing nations around the world, as well as assisting policymakers, emergency planners and industries that rely on accurate weather forecasts.

Dr Anna Allen, the lead author of the paper, from the University of Cambridge, noted that the findings paved the way for better forecasts of natural disasters such as hurricanes, wildfires and tornadoes, as well as other climatic issues such as air quality, ocean dynamics and sea ice predictions.


Aardvark builds on recent research by Huawei, Google, and Microsoft demonstrating that one step of the weather prediction process known as the numerical solver, which calculates how weather evolves over time, can be replaced with AI to produce faster and more accurate predictions. This approach is already being deployed by the ECMWF.

 

The researchers said that using just 10% of the input data that existing systems required, Aardvark could already outperform the US national GFS forecasting system in certain respects, and was competitive with United States Weather Service forecasts.

 

Andrew_C

Andrew_C

Posted
Posted

I read that someone is planning to shut down the United States Weather Service anyway, as you can get a weather forecast on your phone these days...

  • Like 2
mrstrong

mrstrong

Posted
Posted

I’ve only recently started using ChatGPT and would now say it has replaced google as my first port of call.

I do like the “interaction” especially as a lone IT worker its like having a colleague to “discuss” IT issues with.

I know that might sound a bit sad but perhaps it is more common than you might think e.g. in Eastenders there

was a recent storyline about Tommy whose online friend turns out to be an AI bot.

I still don’t think ChatGPT will ever replace the real human interaction / intelligence you can get on forums like Edugeek,

you guys are real right ?

  • Like 1

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share
Go to topic listing


×
×
  • Create New...