• Skip to main content

Biz Builder Mike

You can't sail Today's boat on Yesterdays wind - Michael Noel

  • Tokenomics is not Economics – Digital CX -The Digital Transformation
  • Resume / CV – Michael Noel
  • Contact Us
  • Featured

Blog

Mar 23 2023

Insurer Zurich experiments with ChatGPT for claims and data mining

Insurer Zurich is testing how it can use ChatGPT artificial intelligence technology in areas such as claims and modelling, as it responds to the challenge posed by start-ups and bigger rivals, including China’s Ping An.

It is the latest example of long-established businesses, from law firm Allen & Overy to newspaper publisher Reach, experimenting with the technology.

Ericson Chan, who was poached in 2020 from Ping An to be Zurich’s chief information and digital officer, told the Financial Times that AI could create “a huge amount of efficiency” in jobs such as extracting information from long documents and writing code for statistical models.

“You’re not going to replace a developer, it’s a co-pilot. Similarly, for underwriting, for claims, it is not going to replace [people] but it is going to make it a lot more efficient.”

Zurich is investigating applications of the technology such as extracting data from claims descriptions and other documents. It is feeding in the most recent six years of claims data in an attempt to identify the specific cause of loss across a whole section of claims, with the aim of improving its underwriting.

Under Chan, the insurer has also created a new patent programme to protect its intellectual property, focusing on areas such as automated risk inspection and AI systems for processing bills.

Insurers’ use of AI has caused concern among public policy groups and regulators because of the implications for privacy and the potential for bias. US start-up Lemonade caused controversy in 2021 when it said its AI could identify fraud by picking up on “non-verbal cues” in videos from customers making claims. The insurer later clarified that it was simply describing commonly used facial recognition software that aims to spot the same person making multiple claims.

Chinese lenders including Ping An’s banking arm have gone further, using micro-expression technology in video communications with customers that they say can detect facial indications of whether borrowers are telling the truth. European and US firms have been more reluctant to use this technology.

Chan said Zurich was “very, very open” about its use of AI tools such as facial recognition and did not use micro-expression technology. But, in general, he warned about focusing only on the “pitfalls” of AI. “There are so many areas where there is so much benefit.”

Some insurance start-ups, such as Lemonade, have built their claims processes using AI, which can result in payouts being made on the spot. But despite these technology advances, big underwriting losses have weighed on their valuations.

Chan drew a distinction between newer “challenger” insurers such as Lemonade, which have suffered big drops in valuation, and other start-ups that do not carry risk but instead provide services and technology to the insurance sector. He said it was “not easy” for the challengers because they lack historic data and underwriting experience.

The emergence of tech-focused insurance start-ups was “healthy to take the industry forward”, he said, but the question for some was “how long can you survive to make it work?”

[mailpoet_form id="1"]

Insurer Zurich experiments with ChatGPT for claims and data mining Republished from Source https://www.ft.com/content/45e5525c-ac45-4a49-a55c-8833d1a036b9 via https://www.ft.com/companies/technology?format=rss

Written by Ian Smith Insurance Correspondent · Categorized: entrepreneur, Technology · Tagged: entrepreneur, Technology

Mar 23 2023

FirstFT: Tense US hearing for TikTok chief

Good morning. TikTok’s chief executive was grilled today by US lawmakers, who called the Chinese-owned app a “cancer” and tool of surveillance.

In a tense hearing, Shou Zi Chew told legislators that the viral-video app will be kept “free from any manipulation by any government”, but he struggled to head off a potential US ban.

TikTok has become a flashpoint in rising tensions between the US and China, with Washington concerned that it could be used to steal sensitive US data.

Washington has demanded that the US arm of TikTok should be separated from its Chinese owners, but China is unlikely to co-operate.

“Forcing the sale of TikTok will seriously damage the confidence of investors from all over the world, including from China, on investing in the United States,” the commerce ministry said before the hearing, adding that it would “firmly” oppose such a move.

The other events I will be watching today include:

  • Japan inflation: February consumer price index inflation rate data will be released this morning.

  • UK-Israel meeting: Prime minister Rishi Sunak will host his Israeli counterpart Benjamin Netanyahu in London.

  • Melbourne food and wine festival: The annual festival kicks off today with events in Melbourne and across Victoria.

What did you think of today’s FirstFT? Let us know at firstft@ft.com. Thanks for reading.

Five more top stories

1. Toshiba’s board has approved a $15bn buyout proposal led by Japan Industrial Partners. The deal, in which the 147-year-old conglomerate will be taken private by a consortium backed by 17 domestic companies and six financial institutions, paves the way for the country’s biggest ever take-private deal.

2. Swiss financial regulator Finma has defended its decision to wipe out a huge swath of risky subordinated bonds as part of the Credit Suisse rescue deal. The move, which rendered SFr16bn ($17bn) of investments worthless, has become one of the most controversial elements of the shotgun marriage between Credit Suisse and UBS.

3. The crypto fugitive Do Kwon has been arrested in Montenegro, according to interior minister Filip Adzic. In a statement on Twitter, Adzic said the crypto entrepreneur behind the $40bn implosion of the terraUSD and luna digital tokens had been detained by police at Podgorica Airport with falsified documents.

4. Indian opposition figure Rahul Gandhi has been sentenced to two years in jail for remarks he made about Narendra Modi by a court in the prime minister’s home state of Gujarat, raising the temperature of Indian politics a year ahead of a national election. Here’s what else we know about the sentencing of the best-known figure in the Congress party.

5. China’s population of super-rich fell more than 14 per cent last year as President Xi Jinping’s zero-Covid policy, regulatory crackdowns and a property collapse took their toll on the nation’s largest fortunes. The number of billionaires in China fell by 164 to 969 compared with a decline of 25 to 691 in the US, according to the 2023 M3M Hurun Global Rich List.

How well did you keep up with the news this week? Take our quiz.

The Big Read

© FT montage/AFP/Getty Images

The EU wants to make solar power its single biggest source of energy by 2030. That would mean almost tripling its solar power generation capacity over the next seven years. The problem is more than three-quarters of the EU’s solar panel imports in 2021 came from a single country: China.

We’re also reading . . .

  • Iran-Russia: Moscow has become Iran’s largest foreign investor over the past year, as the two heavily sanctioned nations have increased co-operation since Russia invaded Ukraine.

  • ‘Sensible’ crypto: It’s not the crypto bros we have to fear, writes Jemima Kelly, but the strait-laced, earnest types that are snake-oil salesmen in sensible clothing.

  • Starbucks in China: The US coffee chain plans to open a store in China every nine hours to reach 9,000 locations by 2025.

Chart of the day

There could hardly be a less auspicious time for US business to attend Beijing’s flagship investment conference. But this weekend Americans including former secretary of state Henry Kissinger and investor Ray Dalio will head to Beijing. Recent earnings calls out of the US show that awareness of the geopolitical landscape is tempered by optimism over the Chinese market.

You are seeing a snapshot of an interactive graphic. This is most likely due to being offline or JavaScript being disabled in your browser.



Take a break from the news

As shares of Credit Suisse hover under $1, the price of its memorabilia is on the rise. When the storied Swiss bank ceased to exist practically overnight baseball caps, ski hats and other items bearing its logos were posted on auction sites and racked up bids, as collectors tried to grab pieces of fashion, and financial, history.

Additional contributions by Tee Zhuo and Emily Goldberg

[mailpoet_form id="1"]

FirstFT: Tense US hearing for TikTok chief Republished from Source https://www.ft.com/content/4fdd6e3b-afe4-4ded-86f9-000218038b04 via https://www.ft.com/companies/technology?format=rss

Written by Emily Goldberg · Categorized: entrepreneur, Technology · Tagged: entrepreneur, Technology

Mar 23 2023

TikTok chief faces hostile Congress in bid to fight off US ban

TikTok’s chief executive has told hostile US legislators that the viral-video app will be kept “free from any manipulation by any government”, as he struggled to head off a potential US ban.

Shou Zi Chew received a bipartisan pummelling from a congressional committee on Thursday that threatened to ban the Chinese-owned app, calling it a “cancer” and tool of surveillance.

TikTok has become a flashpoint in rising tensions between the US and China, uniting Republicans and Democrats in Washington who are concerned that it could be used to steal sensitive US data.

Before the hearing, China’s commerce ministry hit out at a US demand that the US arm of TikTok should be separated from its Chinese owners.

“Forcing the sale of TikTok will seriously damage the confidence of investors from all over the world, including from China, on investing in the United States,” it said, adding that it would “firmly” oppose such a move.

In a series of tense exchanges on Capitol Hill over five hours, Chew attempted to address concern over the social media app’s links to its Chinese parent company ByteDance but met a sceptical audience.

“We do not trust TikTok will ever embrace American values,” said the Republican chair of the House energy and commerce committee, Cathy McMorris Rodgers. “TikTok has repeatedly chosen a path for more control, more surveillance and more manipulation. Your platform should be banned.”

Jamaal Bowman speaking at a pro-TikTok rally
US congressman Jamaal Bowman leads a rally to defend TikTok at the US Capitol on Wednesday © J Scott Applewhite/AP

TikTok has taken measures to respond to the US concerns, spending about $2bn on a partnership with Oracle designed to safeguard data and content from its 150mn American users from Chinese influence.

“Our commitment is to move data into the United States to be stored on American soil by an American company overseen by American personnel,” Chew said. “So the risk will be similar to any government going to an American company asking for data.”

But Chew faced overt suspicion from politicians. “I still believe that the Beijing communist government will still control and have the ability to influence what you’re doing,” said Frank Pallone, the panel’s top Democrat.

In addition to concerns about national security, legislators also claimed that the app did not have adequate moderation processes and was unsafe for children, described by one as a “cancer, like fentanyl, another China export, that causes addiction and death”.

The hearing comes at a critical moment for TikTok. This month, Senate Democrats and Republicans introduced a White House-backed bill that would give the administration powers to ban Chinese apps that pose security threats, including TikTok.

Mark Warner, the Democratic chair of the Senate select committee on intelligence who was one of the lead sponsors of the bill, said after the hearing that Chew had said “nothing” to assuage fears that TikTok was at the behest of the Chinese Communist party.

Rogers responded to the Chinese commerce ministry’s statement, telling Chew he had “zero confidence” that ByteDance and TikTok were not “beholden” to the Chinese Communist party. “The CCP believes they have the final say over your company.”

Chew, a former Goldman Sachs banker, emphasised that he was from Singapore and had a US-born wife. He conceded that in the past, employees at Beijing-based ByteDance were subject to Chinese law that would have compelled them to co-operate with state intelligence work.

However, he added that its partnership with Oracle, known as Project Texas, would create a firewall, meaning that was no longer the case for TikTok staff.

In one exchange, Chew evaded commenting on whether he believed that the Chinese government had persecuted the Uyghur Muslim population, as has been alleged by the UN, prompting the committee’s vice-chair, Kelly Armstrong, to accuse him of “absolutely [tying himself] in knots to avoid criticising” the Communist party.

Chew, appearing increasingly exhausted by the questioning, quibbled with the characterisation by legislators of the recent news that TikTok had inappropriately obtained the data of US journalists, as well as a Financial Times reporter, as “spying”.

Ahead of the hearing, TikTok has made a frantic public relations push, including inviting to Washington certain creators, many of whom make livelihoods advertising on behalf of brands to their large followings on the platform, to help lobby politicians against a ban.

In the wake of the of hearing, TikTok said in a statement: “Shou came prepared to answer questions from Congress, but, unfortunately, the day was dominated by political grandstanding that failed to acknowledge the real solutions already under way through Project Texas or productively address industry-wide issues of youth safety.”

Additional reporting by James Politi in Washington

[mailpoet_form id="1"]

TikTok chief faces hostile Congress in bid to fight off US ban Republished from Source https://www.ft.com/content/6f0b931f-10c3-4f85-9f30-5d9977aab459 via https://www.ft.com/companies/technology?format=rss

Written by Hannah Murphy in San Francisco Ryan McMorrow in Beijing and · Categorized: entrepreneur, Technology · Tagged: entrepreneur, Technology

Mar 23 2023

Visual language maps for robot navigation

Posted by Oier Mees, PhD Student, University of Freiburg, and Andy Zeng, Research Scientist, Robotics at Google

People are excellent navigators of the physical world, due in part to their remarkable ability to build cognitive maps that form the basis of spatial memory — from localizing landmarks at varying ontological levels (like a book on a shelf in the living room) to determining whether a layout permits navigation from point A to point B. Building robots that are proficient at navigation requires an interconnected understanding of (a) vision and natural language (to associate landmarks or follow instructions), and (b) spatial reasoning (to connect a map representing an environment to the true spatial distribution of objects). While there have been many recent advances in training joint visual-language models on Internet-scale data, figuring out how to best connect them to a spatial representation of the physical world that can be used by robots remains an open research question.

To explore this, we collaborated with researchers at the University of Freiburg and Nuremberg to develop Visual Language Maps (VLMaps), a map representation that directly fuses pre-trained visual-language embeddings into a 3D reconstruction of the environment. VLMaps, which is set to appear at ICRA 2023, is a simple approach that allows robots to (1) index visual landmarks in the map using natural language descriptions, (2) employ Code as Policies to navigate to spatial goals, such as “go in between the sofa and TV” or “move three meters to the right of the chair”, and (3) generate open-vocabulary obstacle maps — allowing multiple robots with different morphologies (mobile manipulators vs. drones, for example) to use the same VLMap for path planning. VLMaps can be used out-of-the-box without additional labeled data or model fine-tuning, and outperforms other zero-shot methods by over 17% on challenging object-goal and spatial-goal navigation tasks in Habitat and Matterport3D. We are also releasing the code used for our experiments along with an interactive simulated robot demo.


VLMaps can be built by fusing pre-trained visual-language embeddings into a 3D reconstruction of the environment. At runtime, a robot can query the VLMap to locate visual landmarks given natural language descriptions, or to build open-vocabulary obstacle maps for path planning.

Classic 3D maps with a modern multimodal twist

VLMaps combines the geometric structure of classic 3D reconstructions with the expression of modern visual-language models pre-trained on Internet-scale data. As the robot moves around, VLMaps uses a pre-trained visual-language model to compute dense per-pixel embeddings from posed RGB camera views, and integrates them into a large map-sized 3D tensor aligned with an existing 3D reconstruction of the physical world. This representation allows the system to localize landmarks given their natural language descriptions (such as “a book on a shelf in the living room”) by comparing their text embeddings to all locations in the tensor and finding the closest match. Querying these target locations can be used directly as goal coordinates for language-conditioned navigation, as primitive API function calls for Code as Policies to process spatial goals (e.g., code-writing models interpret “in between” as arithmetic between two locations), or to sequence multiple navigation goals for long-horizon instructions.

# move first to the left side of the counter, then move between the sink and the oven, then move back and forth to the sofa and the table twice.
robot.move_to_left('counter')
robot.move_in_between('sink', 'oven')
pos1 = robot.get_pos('sofa')
pos2 = robot.get_pos('table')
for i in range(2):
   robot.move_to(pos1)
   robot.move_to(pos2)
# move 2 meters north of the laptop, then move 3 meters rightward.
robot.move_north('laptop')
robot.face('laptop')
robot.turn(180)
robot.move_forward(2)
robot.turn(90)
robot.move_forward(3)


VLMaps can be used to return the map coordinates of landmarks given natural language descriptions, which can be wrapped as a primitive API function call for Code as Policies to sequence multiple goals long-horizon navigation instructions.

Results

We evaluate VLMaps on challenging zero-shot object-goal and spatial-goal navigation tasks in Habitat and Matterport3D, without additional training or fine-tuning. The robot is asked to navigate to four subgoals sequentially specified in natural language. We observe that VLMaps significantly outperforms strong baselines (including CoW and LM-Nav) by up to 17% due to its improved visuo-lingual grounding.

Tasks    Number of subgoals in a row       Independent
subgoals
     
   1 2 3 4   
LM-Nav    26 4 1 1       26   
CoW    42 15 7 3       36   
CLIP MAP    33 8 2 0       30   
VLMaps (ours)      59 34 22 15       59   
GT Map    91 78 71 67       85   

The VLMaps-approach performs favorably over alternative open-vocabulary baselines on multi-object navigation (success rate [%]) and specifically excels on longer-horizon tasks with multiple sub-goals.

A key advantage of VLMaps is its ability to understand spatial goals, such as “go in between the sofa and TV” or “move three meters to the right of the chair”. Experiments for long-horizon spatial-goal navigation show an improvement by up to 29%. To gain more insights into the regions in the map that are activated for different language queries, we visualize the heatmaps for the object type “chair”.

The improved vision and language grounding capabilities of VLMaps, which contains significantly fewer false positives than competing approaches, enable it to navigate zero-shot to landmarks using language descriptions.

Open-vocabulary obstacle maps

A single VLMap of the same environment can also be used to build open-vocabulary obstacle maps for path planning. This is done by taking the union of binary-thresholded detection maps over a list of landmark categories that the robot can or cannot traverse (such as “tables”, “chairs”, “walls”, etc.). This is useful since robots with different morphologies may move around in the same environment differently. For example, “tables” are obstacles for a large mobile robot, but may be traversable for a drone. We observe that using VLMaps to create multiple robot-specific obstacle maps improves navigation efficiency by up to 4% (measured in terms of task success rates weighted by path length) over using a single shared obstacle map for each robot. See the paper for more details.

Experiments with a mobile robot (LoCoBot) and drone in AI2THOR simulated environments. Left: Top-down view of an environment. Middle columns: Agents’ observations during navigation. Right: Obstacle maps generated for different embodiments with corresponding navigation paths.

Conclusion

VLMaps takes an initial step towards grounding pre-trained visual-language information onto spatial map representations that can be used by robots for navigation. Experiments in simulated and real environments show that VLMaps can enable language-using robots to (i) index landmarks (or spatial locations relative to them) given their natural language descriptions, and (ii) generate open-vocabulary obstacle maps for path planning. Extending VLMaps to handle more dynamic environments (e.g., with moving people) is an interesting avenue for future work.

Open-source release

We have released the code needed to reproduce our experiments and an interactive simulated robot demo on the project website, which also contains additional videos and code to benchmark agents in simulation.

Acknowledgments

We would like to thank the co-authors of this research: Chenguang Huang and Wolfram Burgard.

[mailpoet_form id="1"]

Visual language maps for robot navigation Republished from Source http://ai.googleblog.com/2023/03/visual-language-maps-for-robot.html via http://feeds.feedburner.com/blogspot/gJZg

crowdsourcing week

Written by Google AI · Categorized: AI · Tagged: AI

Mar 23 2023

Hindenburg shorts Jack Dorsey’s payments group Block

Hindenburg Research, the short seller that recently targeted India’s Adani Group, has accused payments group Block of artificially inflating its user numbers and facilitating fraudulent transactions.

In a statement on Thursday, Hindenburg said it had been investigating Block, which is led by Twitter co-founder Jack Dorsey, for two years and had taken a short position in the shares.

It said the accusations were based on “dozens of interviews with former employees, partners, and industry experts, extensive review of regulatory and litigation records” as well as Freedom of Information Act requests and “public records requests”.

Block’s two primary businesses are its Square payments services for merchants and its Cash App mobile payments product, which competes with rivals such as PayPal’s Venmo and Zelle. Hindenburg alleged Block “obfuscates” the actual number of people on its Cash App platform by reporting “misleading . . . metrics filled with fake and duplicate accounts”.

Block said it would work with the Securities and Exchange Commission to explore legal action against Hindenburg “for the factually inaccurate and misleading report” about its Cash App business.

“We have reviewed the full report in the context of our own data and believe it’s designed to deceive and confuse investors,” the company said.

Block’s shares were down more than 14 per cent on Thursday afternoon. As of Wednesday’s closing price, the company had a market capitalisation of nearly $44bn.

New York-based Hindenburg first came to prominence in 2020 when it claimed truck start-up Nikola’s business was a fraud. Earlier this year, it accused Indian billionaire Gautam Adani, at the time the third-richest person in the world, of “pulling the largest con in corporate history”. The report helped wipe about $100bn off his empire’s market value. Adani has denied the allegations.

Hindenburg on Thursday alleged Block had overstated its genuine user counts while also understating the amount it costs them to attract new customers.

Hindenburg cited “former employees” as estimating that 40 to 75 per cent of accounts they reviewed were fake, involved in fraud or were users holding multiple accounts.

Hindenburg said that in investigating the matter, it had opened several accounts under fake names including “Donald Trump” and “Elon Musk”.

The short seller also claimed that what was portrayed as Block’s “wild west” approach had led it to become a popular service for criminals to process transactions.

The report alleged Block has no obvious advantage over competitors such as PayPal or Zelle.

Hindenburg cited the inclusion of Cash App in hip-hop songs where rappers brag about using it for criminal activities.

“A review of those songs shows that the artists are not generally rapping about Cash App’s smooth user interface,” Hindenburg stated, adding that many describe using it “to scam” or for other illicit activity.

[mailpoet_form id="1"]

Hindenburg shorts Jack Dorsey’s payments group Block Republished from Source https://www.ft.com/content/65ec9e18-d09a-4f6b-a9fb-67cdb43ecff6 via https://www.ft.com/companies/technology?format=rss

Written by Joshua Franklin and Ortenca Aliaj in New York and Oliver Ral · Categorized: entrepreneur, Technology · Tagged: entrepreneur, Technology

  • « Go to Previous Page
  • Go to page 1
  • Go to page 2
  • Go to page 3
  • Go to page 4
  • Go to page 5
  • Interim pages omitted …
  • Go to page 438
  • Go to Next Page »
  • Twitter
  • Facebook
  • About Us
  • LinkedIn
  • ANTI-SPAM POLICY
  • Google+
  • API Terms and Conditions
  • RSS
  • Archive Page
  • Biz Builder Mike is all about New World Marketing
  • Cryptocurrency Exchange
  • Digital Millennium Copyright Act (DMCA) Notice
  • DMCA Safe Harbor Explained: Why Your Website Needs a DMCA/Copyright Policy
  • Marketing? Well, how hard can that be?
  • Michael Noel
  • Michael Noel CBP
  • Noels Law of decentralization

Copyright © 2023 · Altitude Pro on Genesis Framework · WordPress · Log in