Aero Inventory Share Discussion Threads

[waldron] 30/3/2023 06:26

1 Hour Ago

CNBC Pro: A less obvious area in A.I. could be an $11 billion opportunity, Bernstein says — and names 4 stocks

Microsoft and Google are among the most obvious beneficiaries of the artificial intelligence revolution. But AI’s applications extend beyond search engines and cloud computing.

One such application is manufacturing — an opportunity that Bernstein estimates will be worth $11 billion.


— Zavier Ong

[the grumpy old men] 29/3/2023 06:10

Apple acquires AI-driven video compression technology startup WaveOne



By NS Business Staff Writer 28 Mar 2023

Established in 2016, WaveOne is engaged in developing the next generation of digital media compression by making use of the latest advancements made in machine learning and deep learning
)

Apple has reportedly snapped up WaveOne, a US-based startup which develops algorithms for compressing video by leveraging artificial intelligence (AI).

According to various reports, the transaction has not been confirmed by the tech major. However, WaveOne’s ex-sales vice president Bob Stankosh confirmed the acquisition through a LinkedIn post published in February 2023.

The financial terms of the deal are not known.

Stankosh stated: “After almost 2 years at WaveOne, last week we finalised the sale of the company to Apple. We started our journey at WaveOne, realising that machine learning/deep learning video technology could potentially change the world. Apple saw this potential and took the opportunity to add it to their technology portfolio.

“I want to thank my co-founders, Lubomir Bourdev and Oren Rippel as well as Craig Lytle for helping to bring the idea to market and advancing this concept for commercial applications.”

Established in 2016, WaveOne is engaged in developing the next generation of digital media compression by utilising the latest advancements made in machine learning and deep learning.

The firm develops custom-tailored and context-dependent solutions.

In January this year, WaveOne’s website was taken down and some former staff members including one of the firm’s co-founders are now employed by other machine learning groups at Apple, reported TechCrunch.

The online newspaper stated that the company’s significant creation was a content-aware video compression and decompression algorithm which can operate on the AI accelerators built into various phones and a growing number of PCs.

By utilising AI-powered scene and object detection, the technology could basically interpret a video frame enabling it to, for instance, prioritise faces at the expense of other elements within a scene for saving bandwidth, said TechCrunch.

[hedgehog 100] 25/3/2023 11:07

Biometrics & AI now seem to be 'joined at the hip' ... or perhaps we should say 'joined at the face':-

"EXPLORING FACIAL RECOGNITION USE CASES IN 2023

Alexey Shalimov / 27th February, 2023 / Trends
The article was updated on February 27, 2023.

A few decades ago, the potential of biometric technology was shown in numerous Sci-Fi films. At that time, such AI technologies seemed mind-blowing and unachievable to be implemented in real life. Years passed, and we have witnessed the emergence of biometrics, which is widely applied in various spheres of our daily lives. These technologies are constantly developing and improving, making our existence more effortless and safer.

Thus, we utilize AI biometric technology when unlocking our smartphones, making payments or logging into our accounts with two-factor authentication, securely boarding a flight, etc. It has become commonplace to the point that many don’t even think twice about it. Such a tendency enables experts to forecast the world market for biometric verification and identification potentially reaching $99.63bn by 2027.

As a biometrics tool, facial recognition technology has attained overwhelming popularity across different industries due to its high efficiency in terms of safety and security. Growing demand for facial recognition algorithms accounts for an accuracy rate of 99.97%.

Face recognition can also be integrated with other authentication methods like fingerprint, voice, or iris detection, providing higher security for certain operations. Millions of people favor face recognition as a reliable way to secure their phones. Nevertheless, today’s possibilities of recognition systems are not limited to the safe use of mobile devices but go far beyond that.

In this article, we will explore how facial recognition works, reveal its main benefits, and present practical uses of facial recognition.

Facial recognition explained

Facial recognition is a solution based on AI algorithms for face detection and recognition. The systems measure an individual’s eye shape, forehead length, and nose parameters, matching them with the biometrics previously obtained from individuals. If a face recognition engine finds a match in a database, it confirms a person’s identity. Face matching can be done in real-time or through images and video.

The technology has proven efficient in face analysis aimed at recognizing individual’s characteristics including gender, age, emotions, movements, mood, and face mask detection that helps notice people who are violating the rules of wearing a mask in public places amidst the coronavirus pandemic.

Unlike such authentication means as passwords, emails, and fingerprints, biometric facial identification applies unique dynamic patterns and mathematical equations that ensure the safety and efficiency of a face detection system.

Face identification technologies are currently experiencing large-scale application across different spheres, including retail and e-Commerce, media and entertainment, automobile and transportation, telecom and IT, government, health, and others. ...

1. Access control

Facial recognition can execute a selective ban on entering certain facilities or accessing some equipment. Examine a list of places where such systems can be beneficial.

Commercial and apartment buildings

In this case, facial identification either allows entrance for authorized workers, homeowners, family members, and previously enrolled guests or denies entry for unapproved individuals. Compelling uses include door entry systems, smart locks, and elevators.

Airports

Biometric technology makes travelers’ journey through an airport seamless and enjoyable. Facial recognition algorithms used by automated terminals and entrance control tourniquets prove efficient for immigration control.

For example, Ljubljana Airport successfully tested facial recognition technology for passengers’ identification at the boarding gates, which reduced boarding times at least by 75%.

Resources and machinery

Each industry uses specialized equipment which requires restricted access, operational monitoring, tracing, and regular reporting. Facial identification software helps protect a company’s resources. Thus, healthcare institutions are already applying smart medicine cabinets to identify their authorized personnel.

Another example can be taken from the Toyota Motor Corporation, which has adopted Vehicle Inspection Information System integrated with face recognition to ban uncertified inspectors from checking finished vehicles. ..."

[the grumpy old men] 25/3/2023 07:56

Conversational AI platform Character.AI raises $150m in Series A



By NS Business Staff Writer 24 Mar 2023

Established in 2021 by AI researchers and former Google engineers Noam Shazeer and Daniel De Freitas, Character.AI offers critical conversational AI technology to help people make their everyday tasks easier and solve complex societal issues
artificial-intelligence-g6eb2cea24_640

Character.AI secures $150m in a Series A funding round. (Credit: Gerd Altmann from Pixabay)

Character.AI has raised $150m in a Series A funding round at a valuation of $1bn for its conversational artificial intelligence (AI) platform.

The funding round was led by US-based venture capital firm Andreessen Horowitz (a16z).

It also saw the participation from the company’s existing investors, including GitHub former CEO Nat Friedman, SV Angel, Elad Gil, and A Capital.

A16z general partner Sarah Wang said: “Character.AI is rapidly and dramatically advancing generative AI, with the potential to transform how humans connect not just with AI, but more broadly reinvent how we interact with technology as a whole in our everyday lives.”

With the funding, Character.AI expects to expand its compute abilities to bring a more sophisticated model with advanced reasoning and increased accuracy.

It is anticipated to provide users with new features and capabilities as well as to grow the company’s engineering team.

Established in 2021 by AI researchers and former Google engineers Noam Shazeer and Daniel De Freitas, Character.AI offers critical conversational AI technology to help people make their everyday tasks easier and solve complex societal issues.

It provides users with the option to develop a fully-customisable and personalised AI companion with a unique personality and values.

In addition, the firm announced the release of an early preview of C1.2, which is its new, smarter, and more helpful model.

C1.2 is said to bring in new productivity capabilities, building on the entertainment, roleplay, and emotional support of the C1.1 model.

De Freitas said: “Our world-class team of engineers continue to develop new frontiers in large language models and AI, bringing together the EQ and IQ to deliver an unparalleled and personalized experience to each user.

“Whether you’re looking for productivity, entertainment, or a supportive friend, Character.AI is here for you, ready to assist and support you every step of the way.

“Character.AI is proud to be a full-stack AI company, meaning we do everything from training the language models to building the user interface. This gives Character.AI the power to create personalised experiences no other AI can match.”

[sarkasm] 22/3/2023 11:17

Flying autonomous robots take on orchard fruit picking

Wed, 22 Mar, 2023 - 11:07


Rachel Martin


Sowing and growing is one thing, but one of the most difficult tasks on any fruit farm is the arduous and often manual task of harvesting the crop.

Israeli-firm Tevel Aerobotics Technologies’ flying autonomous fruit-picking robots could offer just the solution.

Zipping along the rows of trees in an orchard, the drone-like machines can already pick anything from apples through to peaches, pears and nectarines.

The technology isn’t quite delicate enough for raspberries or strawberries at this point, but testing with larger fruit on commercial farms in California and Washington, USA, and the Piedmont region of Italy has already proved successful.

The technology uses specially-designed visual guidance and control algorithms capable of detecting fruit, foliage, and other objects, that enable a high level of accuracy and manoeuvrability.

It then uses artificial intelligence to choose the best trajectory and angle from which to execute the picking process.

It also continuously collects data on every single piece of fruit it picks, providing a level of real-time harvesting data never seen before in the industry.

Such data includes the total amount of fruit picked, weight, colour-grade, and size of each fruit, which combined can offer information on the distribution of weight, size, and colours in the bin.

The really clever bit is that this data is then fed into internal algorithms to optimise fleet management to improve the entire harvesting process.

And it has added benefits for further down the supply chain too; the technology is also capable of adding timestamps and geolocation tagging for traceability.

Tevel’s Flying Autonomous Robots have also manged to overcome one of the biggest practical challenges previously for automated orchard picking, interacting with moving foliage on the trees.

The innovation has been granted 11 patents worldwide and is designed to continuously fly in and out of the trees without harm.

The firm claims the advantages for farmers include improve yield and quality and more streamlined harvesting logistics, resulting in better control over produce and better profitability for the business.

[waldron] 22/3/2023 07:15

Breaking
Bard Vs. ChatGPT: The Major Difference Between The AI Chat Tools, Explained
Arianna Johnson
Forbes Staff

Mar 21, 2023,03:06pm EDT

Topline

On Tuesday, the battle between Google and Microsoft escalated as Google opened public access to Bard, its new AI chatbot tool and ChatGPT’s latest competitor, but the two have major differences, including the ability to remember past conversations and their knowledge base limits.

Key Facts

Only offered to select markets in the U.S. and U.K., hopeful users can sign up for the waitlist to gain access to Bard, which claims to help users plan a birthday party, understand complex topics and create a pros and cons list for a tough decision.

In an effort to separate the chatbot, which is designed to give human-like responses, from its popular search engine, Google has created a standalone website for Bard.

Bard runs on Google’s large learning model Language Model for Dialogue Applications (LaMDA), which it promises will be updated with new information over time.

PROMOTED

In November 2022, Microsoft-backed OpenAI released ChatGPT, an AI-powered chatbot that can answer questions and solve unique problems, like writing a college level essay or drafting a contract between artist and producer.

Google has been in a race to compete with OpenAI since December 2022, declaring a “code red” in response to the launch of ChatGPT.


Main Differences

Though both OpenAI and Google acknowledge their chatbots aren’t perfect and may say inaccurate or offensive things from time to time, the two have stark differences:

Coding: One of the most prominent things that gave ChatGPT its claim to fame was its ability to create complex code. It can even debug code. Researchers at Johannes Gutenberg University Mainz and University College London put the chatbot against industry “standard automated program repair techniques,'' and two common deep learning approaches and found ChatGPT “is competitive to the common deep learning approaches,” and produced “notably better” results than the standard program repair approaches, according to their paper published in arXiv. However, Google stated Bard is “still learning code,” so the feature isn’t available just yet.


Conversation Retention: According to OpenAI, ChatGPT is able to remember what was said in previous conversations. But there are two caveats: the bot can only remember up to 3,000 words (anything beyond that isn’t stored), and it doesn’t use past conversation to form responses. Bard’s ability to retain context is “purposefully limited for now,” Google said, but the company claims the ability will grow over time.


Responses: One of the biggest differences between the two is Bard’s LaMDA can draw responses from the internet, so it will always have the latest responses. It’s also integrated into Google’s search engine and can provide direct links to websites when prompted. On the other hand, ChatGPT runs on Generative Pre-training Transformer-4 (GPT-4), so all of its responses come from its knowledge base, whose cutoff date ends in September 2021, so it’s limited in newer information and research.


Language: ChatGPT knows several languages, including Spanish, French, Arabic, Mandarin, Italian, Japanese and Korean, though its proficiency in responses varies by language and its primary language is English. Bard is only available and can only speak in English.


Drafts: Bard creates several different versions of every prompt (called “drafts”), allowing users to pick the best response. This was implemented because there’s a “sense of authoritativeness221; when only one response is used, Jack Krawczyk, a senior product director at Google, told MIT Technology Review. ChatGPT only produces one response to questions.

Surprising Fact

At Google’s pitch in February, Bard incorrectly answered a question during a promotional video, costing Google’s parent company Alphabet $100 billion in market value.


Further Reading

Google Releases ‘Bard’ AI Chatbot Amid Competition With ChatGPT (Forbes)

Here’s What To Know About OpenAI’s ChatGPT—What It’s Disrupting And How To Use It (Forbes)

Alphabet Stock Plunge Erases $100 Billion After New AI Chatbot Gives Wrong Answer In Ad (Forbes)

Arianna Johnson

I'm a Texas native covering the latest trends in tech, science and healthcare...

[florenceorbis] 20/3/2023 11:42

BYTE Acquisition, Airship AI sign LOI for potential merger deal


By NS Business Staff Writer 13 Mar 2023

Airship AI is an AI-driven surveillance video, sensor and data management platform that helps government agencies and enterprises to collect unstructured data from surveillance cameras and sensors


Airship AI has signed a non-binding letter of intent (LOI) with BYTE Acquisition, a special purpose acquisition corporation (SPAC), for a proposed business combination that would value the former at a pro forma enterprise value of $290m.

Established in 2006, Airship AI is a US-based artificial intelligence (AI)-driven surveillance video, sensor and data management platform. The deal with the SPAC will enable it to go public.

The firm helps government agencies and enterprises to collect unstructured data from surveillance cameras and sensors and apply artificial intelligence (AI) analytics to enhance public safety and operational efficiency.

As per the terms of the proposed deal, Airship AI’s shareholders would roll 100% of their equity into the combined entity.

The SPAC has committed to secure a $4m pre-transaction bridge.

Both firms plan to jointly raise further capital through a private investment in public equity (PIPE).

Airship AI president Paul Allen said: “We’re excited to partner with the BYTE team which brings substantial software and business expertise.

“Being a public company will bolster credibility as a financially strong partner with customers by demonstrating our financial stability and transparency.”

Airship AI also offers visualisation tools to improve decision making in mission critical environments.

Headquartered in Redmond, Washington, the technology firm performs predictive analysis of events before they happen and provides decision makers with meaningful intelligence.

The completion of the proposed business combination is subject to the fulfilment of due diligence, the negotiation of a definitive agreement providing for the deal, as well as the satisfaction of the conditions negotiated therein.

Besides, the proposed deal requires BYTE Acquisition and Airship AI’s boards’ and shareholders’ approvals.

BYTE Acquisition stated that there is no assurance that a definitive agreement will be signed by the companies or that the proposed business combination will be completed.

[adrian j boris] 20/3/2023 10:01

Microsoft adds AI-powered Microsoft 365 Copilot to Office suite



By NS Business Staff Writer 17 Mar 2023

The new offering is presently in testing with a small group of commercial customers in order to gather essential feedback needed to enhance the models as they scale and is integrated into the Microsoft 365 apps including Word, Excel, PowerPoint, Outlook, Teams, and others


Microsoft has introduced Microsoft 365 Copilot, an artificial intelligence (AI)-powered co-pilot designed to combine large language models (LLMs) and business data and Microsoft 365 applications to drive productivity in workplaces.

The new offering is presently in testing with a small group of commercial customers in order to gather essential feedback needed to enhance the models as they scale.

According to the company, Microsoft 365 Copilot is integrated into Microsoft 365 apps including Word, PowerPoint, Excel, Outlook, Teams, and others.

Microsoft has also announced a new experience called Business Chat, which has been designed to work across the LLM, the Microsoft 365 apps, and users’ emails, chats, calendar, documents, meetings, and contacts.

Based on the morning meetings, emails, and chat threads, Business Chat will produce a status update, said the tech major.

The new Co-pilot in Power Platform is expected to allow developers of all skill levels expedite and optimise development with low-code tools through the launch of two new capabilities within Power Apps and Power Virtual Agent.

In Word, Microsoft 365 Copilot will write, edit, summarise, and create along with users while they work.

The PowerPoint Co-pilot integration will allow the creation process by converting ideas into designed presentations by using natural language commands.

Microsoft modern work and business applications corporate vice president Jared Spataro said: “Copilot combines the power of large language models with your data and apps to turn your words into the most powerful productivity tool on the planet.

“By grounding in your business content and context, Copilot delivers results that are relevant and actionable. It’s enterprise-ready, built on Microsoft’s comprehensive approach to security, compliance, privacy and responsible AI.

“Copilot marks a new era of computing that will fundamentally transform the way we work.”

Earlier this month, Microsoft launched Microsoft Dynamics 365 Copilot, an AI-powered co-pilot in both customer relationship management (CRM) and enterprise resource planning (ERP) systems.

[florenceorbis] 15/3/2023 06:41

OpenAI announces ChatGPT successor GPT-4

Published

9 hours ago

By Ben Derico and Zoe Kleinman
BBC News

OpenAI has released GPT-4, the latest version of its hugely popular artificial intelligence chatbot ChatGPT.

The new model can respond to images - providing recipe suggestions from photos of ingredients, for example, as well as writing captions and descriptions.

It can also process up to 25,000 words, about eight times as many as ChatGPT.

Millions of people have used ChatGPT since it launched in November 2022.

Popular requests for it include writing songs, poems, marketing copy, computer code, and helping with homework - although teachers say students shouldn't use it.

ChatGPT answers questions using natural human-like language, and it can also mimic other writing styles such as songwriters and authors, using the internet as it was in 2021 as its knowledge database.

There are concerns that it could one day take over many jobs currently done by humans.

OpenAI said it had spent six months on safety features for GPT-4, and had trained it on human feedback. However it warned that it may still be prone to sharing disinformation.

GPT-4 will initially be available to ChatGPT Plus subscribers, who pay $20 per month for premium access to the service.

It's already powering Microsoft's Bing search engine platform. The tech giant has invested $10b into OpenAI.

In a live demo it generated an answer to a complicated tax query - although there was no way to verify its answer.

GPT-4, like ChatGPT, is a type of generative artificial intelligence. Generative AI uses algorithms and predictive text to create new content based on prompts.


GPT-4 has "more advanced reasoning skills" than ChatGPT, OpenAI said. The model can, for example, find available meeting times for three schedules.

OpenAI also announced new partnerships with language learning app Duolingo and Be My Eyes, an application for the visually impaired, to create AI Chatbots which can assist their users using natural language.

However, like its predecessors, OpenAI has warned that GPT-4 is still not fully reliable and may "hallucinate" - a phenomenon where AI invents facts or makes reasoning errors.

[waldron] 13/3/2023 16:12

Conclusion

In the best-case scenario, outsourcing writing essays and testing knowledge to ChatGPT is a significant sign that traditional learning and teaching methods are already declining. The educational system remains largely unchanged, and it may be time to undergo necessary changes.

[waldron] 12/3/2023 09:30

At San Francisco expo, AI 'sorry' for destroying humanity

Issued on: 12/03/2023 - 08:48Modified: 12/03/2023 - 08:46


A new exhibition titled the Misalignment Museum opened in San Francisco on March 9, 2023, featuring funny and disturbing AI art works

San Francisco (AFP) – Advances in artificial intelligence are coming so hard and fast that a museum in San Francisco, the beating heart of the tech revolution, has imagined a memorial to the demise of humanity.


"Sorry for killing most of humanity person with smile cap and mustache," says a monitor welcoming a visitor to the "Misalignment Museum," a new exhibit on the controversial technology.

The pieces in this temporary show mix the disturbing with the comic, and this first display has AI disburse pithy observations to the visitors that cross into its line of vision.

"The concept of the museum is that we are in a post-apocalyptic world where artificial general intelligence has already destroyed most of humanity," said Audrey Kim, the show's curator.

"But then the AI realizes that was bad and creates a type of memorial to the human, so our show's tagline is 'sorry for killing most of humanity,'" she said.

Artificial General Intelligence is a concept that is even more nebulous than the simple AI that is cascading into everyday life, as seen in the fast emergence of apps such as ChatGPT or Bing's chatbot and all the hype surrounding them.

AGI is "artificial intelligence that is able to do anything that a human would be able to do," integrating human cognitive capacities into machines.


All around San Francisco, and down the peninsula in Silicon Valley, startups are hot on the trail of the AGI holy grail.

Sam Altman, the founder of ChatGPT creator OpenAI, has said AGI, done right, can "elevate humanity" and change the "limits of possibilities."


But Kim wants to trigger a reflection on the dangers of going too far, too quickly.


Curator Audrey Kim talks about the piece "Paperclip Embrace" at the Misalignment Museum


"There have been lots of conversations about the safety of AI in pretty niche intellectual tech circles on Twitter and I think that's very important," she said.

But those conversations are not as easily accessible to the general public as concepts that you can see or feel, she added.

Kim is particularly fond of a sculpture called "Paperclip Embrace": two busts of humans holding each other, made entirely of paperclips.

The work refers to a metaphor by philosopher Nick Bostrom, who in the 2000s imagined what would happen if artificial intelligence was programmed to create paper clips.

"It could become more and more powerful, and constantly optimize itself to achieve its one and only goal, to the point of destroying all of humanity in order to flood the world with paper clips," Kim said.

Weighing the pros and cons of AI is a subject that became close to Kim's heart in an earlier job working for Cruise, an autonomous vehicle company.

There she worked on an "incredible" technology, which "could reduce the number of accidents due to human error," but also presented risks, she said.

The exhibit occupies a small space in a street corner building in San Francisco's hip Mission neighborhood.

The lower floor of the exhibition is dedicated to AI as a nightmarish dystopia where a machine powered by GPT-3, the language model behind ChatGPT, composes spiteful calligrams against humanity, in cursive writing.


One exhibit is an AI-generated -- and totally fake -- dialogue between the philosopher Slavoj Zizek and the filmmaker Werner Herzog, two of Europe's most respected intellectuals.

This "Infinite Conversation" is a meditation on deep fakes: images, sound or video that aim to manipulate opinion by impersonating real people and that have become the latest disinformation weapon online.

"We only started this project five months ago, and yet many of the technologies presented here already seem almost primitive," Kim said, astonished.

She hopes to turn the exhibit into a permanent one with more space and more events.

[hedgehog 100] 08/3/2023 19:26

"Investors Target AI Winners and They're Not All Household Names

As AI investment funds sizzle, firms large and small provide the heat

By LYLE NIEDENS Published February 27, 2023

In the investment world, AI is on fire.

Since the introduction of OpenAI's ChatGPT last fall, the technology's transformative potential has enthralled investors, who've sent AI-oriented investment funds soaring this year as they also scrutinize which individual companies might benefit most from the technology.

The iShares Robotics and Artificial Intelligence Multisector ETF (IRBO) has gained 15% year-to-date, with rival Global X Robotics & Artificial Intelligence ETF (BOTZ) surging 13%. 1 2

While obvious suspects such as Amazon, Google, and Apple stand to reap AI's rewards, likely beneficiaries also include smaller, more obscure companies in industries ranging from finance and human resources to travel and transportation.

Wealth management firm Baird last week issued a research report detailing its top 50 AI and machine learning stock picks. That followed a Forbes list of best AI stocks to buy in 2023. 3

Four companies appeared on both lists: Alphabet (Google's parent company), Adobe, Amazon, and Nvidia. Forbes identified such old-school and high-profile firms as IBM, Oracle, and Microsoft, while Baird targeted well-known names such as Pinterest and Accenture.

The lists also highlighted numerous smaller companies, all that someday may have the chops to strike AI gold. A few examples: ...

Paycor HCM (PYCR): Payroll and so-called human capital management (HCM) providers reduce service and implementation costs via AI bots that answer and fulfill service requests. AI also can aid recruiting: Paycor recently purchased Talenya, which helps hiring managers efficiently find employment candidates. 4 ... "




Paycor HCM (NASDAQ: PYCR) is the largest North American partner of Grosvenor Technology, a subsidiary of Newmark Security (NWT):-

"Insights | Human Capital Management
Paycor & Grosvenor Technology: A Powerful Pair
Paycor’s Product Manager, David Goodwin, explains why they selected Grosvenor Technology as their hardware provider
“ ... The Grosvenor Technology team is an amazing partner. The collaboration and communication as we worked on the development of our software was great. ...” ..."

[sarkasm] 06/3/2023 07:06

Europe’s AI weaknesses could matter less in generative world, says Insight Partners
Insight Partners'

Lonne Jaffe gives his take on the state of European AI innovation, which he says could benefit from generative models

By Tim Smith 6 March 2023


Europe is far behind in the AI race; of the 10 best-funded startups in the space, none are European, while the biggest heavyweights in the space — Google and Microsoft — are US companies.

But, according to New York-based Insight Partners’ managing director Lonne Jaffe, Europe may not be at such a disadvantage in this new era of generative AI.

Insight has backed big AI startups like JasperAI, Weights & Biases and Granulate (which got acquired by Intel in 2022) and, while these big-ticket investments are US-based, Jaffe has his fair share of skin in the European game.

His investments on this side of the pond include AI-powered insurtech Tractable, cybersecurity platform Featurespace and trading platform TradingView — but he makes no bones about Europe’s shortcomings when it comes to building AI.


Historically hampered

Europe has had two structural problems when it comes to AI innovation, says Jaffe.

The first is regulation around the use of data — which is much stricter in Europe, making it harder to train machine learning models on lots of information than in the States. And then there’s the shortage of talent (much of which has relocated to the US).

But Jaffe believes that both of these problems will be less of an issue in the generative AI context.
Fine tuning

That’s because much of the value from generative AI going forward will be created by companies fine-tuning pre-existing models, he says.

This, he argues, is a task that’s far less technically demanding than building previous types of AI functionality into a business. And it’s something where European companies could take a lead, even if they might struggle to catch up with the likes of OpenAI when it comes to training massive models.

“It doesn’t seem to be that hard to fine-tune these models. I had one example recently of talking to a large multinational corporate that you would not expect to have a deeptech team of any kind,” he explains. “In this generative AI use case their people — who were relatively not very technically sophisticated — fine-tuned a version of GPT-3 in [Microsoft] Azure, in about two weeks.”

Jaffe says this company was able to build a useful generative tool for navigating its document database, without needing to hire machine learning specialists.

He adds that the data regulation issue is also less important in the generative AI context, as a small amount of proprietary data can be used to fine-tune a pre-existing model, with great results.

“In the generative AI world many of these use cases require much less data to be able to take advantage of the systems,” says Jaffe. “You can just take some of your own private data, fine tune a model from one of the large foundation model companies and you can get a pretty powerful system. You don’t need to have any specialised talent or to manage any specialised machine learning infrastructure.̶1;


Picking battles

Jaffe also says that there might be some advantages that come from the open source strategies being led by European companies like Stability AI and Aleph Alpha, with one example being Apple’s integration of Stable Diffusion (Stability AI’s image generation tool) into its own processor.

“They took a version of Stable Diffusion and got it to work really well on their specialised chips on the iPhone. Having it be open source means that Apple can really get into the internals and play around with the nuance and subtlety of how it’s working to make sure that it’s good,” he says.

Use cases like this might give open source models like Stability’s penetration into places that would be harder for OpenAI, but Jaffe adds that, ultimately, whether open or closed source, it will be the foundation models (like GPT-3) that deliver the best results and outputs that will win out.

“The real question about open source versus closed source in the foundation model universe will be how much better are the closed source models, because they have these massive war chests of investment going into them,” he says. “People are going to really care about quality here.”

So, while the pursuit of building the best foundation models in generative AI might be somewhat out of Europe’s hands, Jaffe still thinks European companies — even incumbents — can compete by building products that leverage the technology.

It will now be up to Europe’s entrepreneurs working with generative AI to execute strategies that create value for the consumer in smart ways — something Jaffe says European’s have shown no shortage of talent in.

Tim Smith is senior reporter at Sifted. He tweets from @timmpsmith

[misca2] 01/3/2023 08:17

rté

Advertising watchdog to use AI to monitor social media

Updated / Wednesday, 1 Mar 2023 06:59

AI will evaluate activity by influencers and assess if they are disclosing content correctly

By Brian O'Donovan

Work & Technology Correspondent

The Advertising Standards Authority for Ireland (ASAI) is to use artificial intelligence (AI) to identify social media posts by influencers that are in breach of the advertising code.

The tools will evaluate activity by influencers and assess if they are disclosing content correctly.

Under Irish law, commercial content must be labelled as advertising.

New research by ASAI shows that inauthenticity, photoshopping, too many paid ads and influencers misrepresenting real life are the top traits that cause annoyance amongst Irish consumers.

The study found that only one in ten people have trust in what influencers post on social media, while over 62% believe that influencers post too much sponsored content.

The research also shows that not all consumers are familiar with the various hashtags used to label commercial content, such as #sponsored, #ad, #sp and #iworkwith.

ASAI said the results of its research indicate that there is a need for additional guidance for both consumers and influencers, particularly around labelling.

The authority said it is working with the Competition and Consumer Protection Commission (CCPC) on updated guidance to ensure there is increased clarity in relation to responsibilities and requirements from influencers when advertising on social media.

"As we can see from this research, over half of those surveyed remain bothered by both the lack of transparency in influencer marketing and not being able to distinguish content from advertising, which echoes similar results to the Social Media Influencer Report released by the CCPC last December," said Orla Twomey, Chief Executive of the ASAI.

"To help with this we are planning to continue implementing the use of AI tools and working with the CCPC to develop further guidance," Ms Twomey said.

CCPC member Kevin O'Brien said the commission welcomed the ASAI's on-going work on consumers’ understanding of how influencers and brands operate on social media platforms.

"Similar to our own research, published late last year, the ASAI’s research reinforces that platforms and brands must take greater responsibility for educating and informing their users and consumers, and must support influencers in clearly and consistently labelling paid content so that consumers are not misled," Mr O'Brien said.

[hazl] 21/2/2023 12:03

Yes we definitely seem to be following the Chinese model everywhere.
More surveillance likely leading to less freedom.

Strange I looked at this thread exactly a year ago....well on the 22nd.