The latest technology news and information on startups

Via files confidentially to go public
▻https://techcrunch.com/2021/12/21/via-files-confidentially-to-go-public

Livin’ la Via loca
Natasha Mascarenhas, Alex Wilhelm / 7:58 PM GMT+1•December 21, 2021

On-demand shuttle service and software company Via has confidentially filed to go public, the company said in a statement. The company has not yet determined how many shares will be offered or what the price range for the proposed offering will look like, as is typical for this sort of release.

Via follows Reddit in filing privately for a public listing before the end-of-year market freeze. Both companies should debut in early 2022

Why file now given the impending holiday period? Via has posted impressive results in recent quarters, so getting its ducks in a row – shuttles in a line? – for an early-early IPO is far from aggressive. It’s reasonable, frankly, given that tech valuations are still strong despite recent declines. And companies that can go public may want to take advantage of the IPO window while it’s open.

Per Crunchbase, Via has raised $777.1 million in known funding to date, from investors including Macquarie Capital, Mori Building, Shell, 83North, Broadscale Group, Ervington Investments, Hearst Ventures, Planven Ventures, Pitango and RiverPark Ventures.

Last month Via announced a $130 million round that pushed the on-demand shuttle and software company to a $3.3 billion valuation. We suppose that that was a pre-IPO round, in retrospect. The capital event came on the heels of TransitTech, Via’s software platform, doubling year over year to exceed an annual run rate of $100 million.

More simply, Via has reached the threshold of IPO scale in revenue terms, not taking into account other possible income streams at the company. So its IPO timing once again makes sense.

Past the software revenue result, after originally taking five years to land its first city partnership, the company today has over 500 partners, including Los Angeles Metro, Jersey City and Miami.

Life has not proved pothole free for Via. The company saw mixed demand for its business amidst the pandemic. While ridership initially dropped due to the disease, cities focused more on emergency services — helping generate demand for Via’s software platform. From a business perspective, that’s good news considering the stickiness of a city contract.

Via has beefed up its operations over the past few years, perhaps hinting at plans to make a pitch for a public debut. In March, the company bought Remix, which created software for cities to use for transportation planning and street design, for $100 million. It also acquired Fleetonomy.

Via won’t be the only mobility startup racing toward a 2022 public debut. This month, Voi Scooters raised a $115 million Series D in preparation for an IPO and Kakao Mobility picked up a $4.2 billion valuation ahead of its rumored listing.

TechCrunch has a series of questions for the filing, once we get our hands on it. For example, we’re curious whether the company is able to generate traditional SaaS gross margins while selling to governments. And how long are its sales cycles? Even more, how deeply penetrated is its core market? And where did all that capital raised get spent? Will I see it in investing cash flow or operating losses?

All told, the 2022 IPO cycle is shaping up to start off with a number of very, very interesting debuts.

#Bluesky and #Mastodon users are having a fight that could shape the next generation of social media
▻https://techcrunch.com/2024/02/14/bluesky-and-mastodon-users-are-having-a-fight-that-could-shape-the-next-

Software developer Ryan Barrett found this out the hard way when he set out to connect the AT Protocol and #ActivityPub with a bridge called Bridgy Fed.

The conflict harks back to blogging culture in the early 2000s, when people worried about their innermost thoughts and feelings being indexed on Google. These bloggers wanted their posts to be public, so that they could try to form communities with like-minded people on platforms like LiveJournal, but they didn’t want their intimate musings to accidentally fall into the wrong hands.

New MIT CSAIL study suggests that AI won’t steal as many jobs as expected | TechCrunch
▻https://techcrunch.com/2024/01/22/new-mit-csail-study-suggests-that-ai-wont-steal-as-many-jobs-expected

https://techcrunch.com/wp-content/uploads/2023/11/GettyImages-496822526.jpg?resize=1200,800

Will AI automate human jobs, and — if so — which jobs and when?

That’s the trio of questions a new research study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), out this morning, tries to answer.

There’s been many attempts to extrapolate out and project how the AI technologies of today, like large language models, might impact people’s’ livelihoods — and whole economies — in the future.

Goldman Sachs estimates that AI could automate 25% of the entire labor market in the next few years. According to McKinsey, nearly half of all work will be AI-driven by 2055. A survey from the University of Pennsylvania, NYU and Princeton finds that ChatGPT alone could impact around 80% of jobs. And a report from the outplacement firm Challenger, Gray & Christmas suggests that AI is already replacing thousands of workers.

But in their study, the MIT researchers sought to move beyond what they characterize as “task-based” comparisons and assess how feasible it is that AI will perform certain roles — and how likely businesses are to actually replace workers with AI tech.

Contrary to what one (including this reporter) might expect, the MIT researchers found that the majority of jobs previously identified as being at risk of AI displacement aren’t, in fact, “economically beneficial” to automate — at least at present.

The key takeaway, says Neil Thompson, a research scientist at MIT CSAIL and a co-author on the study, is that the coming AI disruption might happen slower — and less dramatically — than some commentators are suggesting.

Early in the study, the researchers give the example of a baker.

A baker spends about 6% of their time checking food quality, according to the U.S. Bureau of Labor Statistics — a task that could be (and is being) automated by AI. A bakery employing five bakers making $48,000 per year could save $14,000 were it to automate food quality checks. But by the study’s estimates, a bare-bones, from-scratch AI system up to the task would cost $165,000 to deploy and $122,840 per year to maintain . . . and that’s on the low end.

“We find that only 23% of the wages being paid to humans for doing vision tasks would be economically attractive to automate with AI,” Thompson said. “Humans are still the better economic choice for doing these parts of jobs.”

Now, the study does account for self-hosted, self-service AI systems sold through vendors like OpenAI that only need to be fine-tuned to particular tasks — not trained from the ground up. But according to the researchers, even with a system costing as little as $1,000, there’s lots of jobs — albeit low-wage and multitasking-dependent — that wouldn’t make economic sense for a business to automate.

#Intelligence_artificielle #travail

Law firm that handles data breaches was hit by data breach | TechCrunch
▻https://techcrunch.com/2024/01/04/orrick-law-firm-data-breach

An international law firm that works with companies affected by security incidents has experienced its own cyberattack that exposed the sensitive health information of hundreds of thousands of data breach victims.

#mise_en_abime

El gobierno no puede apoderarse de sus datos, pero puede comprarlos.
▻https://techcrunch.com/2023/05/21/the-government-cant-seize-your-data-but-it-can-buy-it

https://techcrunch.com/wp-content/uploads/2023/05/GettyImages-182162036.jpg?resize=1200,800

EU lawmakers bag late night deal on ‘global first’ AI rules | TechCrunch
▻https://techcrunch.com/2023/12/08/eu-ai-act-political-deal

https://techcrunch.com/wp-content/uploads/2023/12/eu-ai-act-trilogue.png?resize=1200,838

Tout l’article est très intéressant.

Full details of what’s been agreed won’t be entirely confirmed until a final text is compiled and made public, which may take some weeks. But a press release put out by the European Parliament confirms the deal reached with the Council includes a total prohibition on the use of AI for:

biometric categorisation systems that use sensitive characteristics (e.g. political, religious, philosophical beliefs, sexual orientation, race);
untargeted scraping of facial images from the internet or CCTV footage to create facial recognition databases;
emotion recognition in the workplace and educational institutions;
social scoring based on social behaviour or personal characteristics;
AI systems that manipulate human behaviour to circumvent their free will;
AI used to exploit the vulnerabilities of people (due to their age, disability, social or economic situation).

The use of remote biometric identification technology in public places by law enforcement has not been completely banned — but the parliament said negotiators had agreed on a series of safeguards and narrow exceptions to limit use of technologies such as facial recognition. This includes a requirement for prior judicial authorisation — and with uses limited to a “strictly defined” lists of crime.

Civil society groups have reacted sceptically — raising concerns the agreed limitations on state agencies’ use of biometric identification technologies will not go far enough to safeguard human rights. Digital rights group EDRi, which was among those pushing for a full ban on remote biometrics, said that whilst the deal contains “some limited gains for human rights”, it looks like “a shell of the AI law Europe really needs”.

There was also agreement on a “two-tier” system of guardrails to be applied to “general” AI systems, such as the so-called foundational models underpinning the viral boom in generative AI applications like ChatGPT.

As we reported earlier, the deal reached on foundational models/general purpose AIs (GPAIs) includes some transparency requirements for what co-legislators referred to as “low tier” AIs — meaning model makers must draw up technical documentation and produce (and publish) detailed summaries about the content used for training in order to support compliance with EU copyright law. For “high-impact” GPAIs (defined as the cumulative amount of compute used for their training measured in floating point operations is greater than 10^25) with so-called “systemic risk” there are more stringent obligations.

“If these models meet certain criteria they will have to conduct model evaluations, assess and mitigate systemic risks, conduct adversarial testing, report to the Commission on serious incidents, ensure cybersecurity and report on their energy efficiency,” the parliament wrote. “MEPs also insisted that, until harmonised EU standards are published, GPAIs with systemic risk may rely on codes of practice to comply with the regulation.”

The Commission has been working with industry on a stop-gap AI Pact for some months — and it confirmed today this is intended to plug the practice gap until the AI Act comes into force.

While foundational models/GPAIs that have been commercialized face regulation under the Act, R&D is not intended to be in scope of the law — and fully open sourced models will have lighter regulatory requirements than closed source, per today’s pronouncements.

The package agreed also promotes regulatory sandboxes and real-world-testing being established by national authorities to support startups and SMEs to develop and train AIs before placement on the market.

#Intelligence_artificielle #AIAct #Europe #Régulation

À propos de la restriction d’accès actuelle à Twitter :

https://user-images.githubusercontent.com/71434691/250187597-4e190dba-3c5a-4acb-99ba-2e897fa69cbc.jpeg

▻https://platform.twitter.com/embed/Tweet.html?id=1674942336583757825

Les détails sur le scraping par des boites d’IA qui serait à l’origine de la restriction :

▻https://techcrunch.com/2023/06/30/twitter-now-requires-an-account-to-view-tweets

Comme le fait remarquer quelqu’un sur le github nitter, ça pourrait être une mesure temporaire du genre opération militaire spéciale qui devait durer trois jours.

[edit] Les instances nitter étant donc en rade, alternative Fediverse / ActivityPub via bird.makeup pour accéder à twitter sans être logué.

[edit 2] D’autres astuces de contournement dans ce ticket :
▻https://github.com/zedeus/nitter/issues/919
(ne permettent pas toutes un accès anonyme)

Telegram shares users’ data in copyright violation lawsuit | TechCrunch
▻https://techcrunch.com/2022/11/29/telegram-shares-data-of-users-accused-of-copyright-violation-following-c

Telegram has disclosed names of administrators, their phone numbers and IP addresses of channels accused of copyright infringement in compliance with a court order in India in a remarkable illustration of the data the instant messaging platform stores on its users and can be made to disclose by authorities.

Tumblr to add support for ActivityPub, the social protocol powering Mastodon and other apps | TechCrunch
▻https://techcrunch.com/2022/11/21/tumblr-to-add-support-for-activitypub-the-social-protocol-powering-masto

Tumblr will add support for ActivityPub, the open, decentralized social networking protocol that today is powering social networking software like Twitter alternative Mastodon, the Instagram-like Pixelfed, video streaming service PeerTube, and others. The news was revealed in response to a Twitter user’s complaint about Mastodon’s complexities. Automattic CEO Matt Mullenweg — whose company acquired Tumblr from Verizon in 2019 — suggested the user “come to Tumblr” as the site would soon “add activitypub for interconnect.”

It’s official : Broadcom to acquire VMware in massive $61B deal | TechCrunch
▻https://techcrunch.com/2022/05/26/broadcom-to-acquire-vmware-in-massive-61b-deal

It turns out that fire was burning hot, and today, Broadcom announced it is acquiring VMware in a massive $61 billion deal.

The deal is a combination of cash and stock, with Broadcom assuming $8 billion in VMware debt.

J’ai honte, j’étais un petit peu passé à côté de cette nouvelle, qui date de mai dernier.

Activision Blizzard illegally withheld raises from unionizing workers, labor board finds | TechCrunch
▻https://techcrunch.com/2022/10/03/activision-blizzard-illegally-withheld-raises-from-unionizing-workers-la

https://techcrunch.com/wp-content/uploads/2022/07/GettyImages-1234271856.jpg?resize=1200,938

Gaming giant Activision Blizzard unlawfully retaliated against workers at Raven Software who formed a union, the National Labor Relations Board found.

The quality assurance (QA) department at subsidiary Raven Software, who mostly work on “Call of Duty,” announced that they would form a union in January. Activision Blizzard sought to block the union, reasoning that the union only comprises the 28-employee QA department, while as a whole, Raven Software has around 230 employees. Regardless, the Raven Software QA testers, who operate under the name Game Workers Alliance (GWA), made history in May when their union vote passed 19-3. Now, the GWA is the first officially recognized union at a major U.S. gaming company.

While the GWA was in the process of unionizing, Activision Blizzard converted about 1,100 QA contractors to full-time staffers and increased the minimum wage to $20 per hour. But workers at Raven Software, who are among the lowest paid in the studio, were denied these wage increases. Activision Blizzard claimed that, due to laws under the National Labor Relations Act (NLRA), the company wasn’t allowed to change the pay rate of its employees in the midst of a union effort. The Communication Workers of America, which represents the union, said that this was a disingenuous attempt at union busting.

#jeu_vidéo #jeux_vidéo #activision_blizzard #ressources_humaines #syndicalisme #game_workers_alliance #gwa #assurance_qualité #nlra #national_labor_relations_act #raven_software #daniel_alegre #bobby_kotick #harcèlement_sexuel #viol #sec #microsoft

Microsoft brings DALL-E 2 to the masses with Designer and Image Creator
Microsoft is making a major investment in DALL-E 2, OpenAI’s AI-powered system that generates images from text, by bringing it to first-party apps and services. During its Ignite conference this week, Microsoft announced that it’s integrating DALL-E 2 with the newly announced Microsoft Designer app and Image Creator tool in Bing and Microsoft Edge.
▻https://techcrunch.com/2022/10/12/microsoft-brings-dall-e-2-to-the-masses-with-designer-and-image-creator

Seeking to bring OpenAI’s tech to an even wider audience, Microsoft is launching Designer, a Canva-like web app that can generate designs for presentations, posters, digital postcards, invitations, graphics and more to share on social media and other channels. Designer — whose announcement leaked repeatedly this spring and summer — leverages user-created content and DALL-E 2 to ideate designs, with drop-downs and text boxes for further customization and personalization.

Another new Microsoft-developed app underpinned by DALL-E 2 is Image Creator, heading to Bing and Edge in the coming weeks. As the name implies, Image Creator — accessed via the Bing Images tab or bing.com/create, or through the Image Creator icon in the sidebar within Edge — generates art given a text prompt by funneling requests to DALL-E 2, acting like a frontend client for OpenAI’s still-in-beta DALL-E 2 service.

L’article évoque aussi la question des #droit liés aux images générés, le #deepfake violent ou pornographique et la question des #dataset utilisés pour entraîner ce #programme appelé #ia ou #reseau_de_neurones génératif d’images.

Meta announces legs
▻https://techcrunch.com/2022/10/11/meta-announces-legs

The announcement that the avatars, which were previously floating torsos with arms and heads, now have evolved to walk was something Zuckerberg was very excited about with his avatar jumping for joy during the keynote.

Alongside announcements around the appearance and movements of the new full body avatars, Meta also announced that there will soon be an avatar store where people will be able to spend real money to buy accessories for their Meta avatar . There was notably no mention of NFTs.

#metawatch #meta #metaverse

Google revives RSS | TechCrunch
▻https://techcrunch.com/2021/05/19/undead-again-google-brings-back-rss

https://techcrunch.com/wp-content/uploads/2021/05/GettyImages-172873092.jpg?w=594

Facebook’s Kustomer buy could face EU probe after merger referral
▻https://techcrunch.com/2021/04/06/facebooks-kustomer-buy-could-face-eu-probe-after-merger-referral/?guccounter=1&guce_referrer=aHR0cHM6Ly90LmNvLw&guce_referrer_sig=AQAAALQ

The European Union may investigate Facebook’s $1 billion acquisition of customer service platform Kustomer after concerns were referred to it under EU merger rules. A spokeswoman for the Commission confirmed it received a request to refer the proposed acquisition from Austria under Article 22 of the EU’s Merger Regulation — a mechanism which allows Member States to flag a proposed transaction that’s not notifiable under national filing thresholds (e.g. because the turnover of one of the (...)

#Facebook #domination #données #santé #Kustomer

##santé

https://techcrunch.com/wp-content/uploads/2020/12/facebook-oversight-scrutiny.jpg

Memes for sale | TechCrunch
▻https://techcrunch.com/2021/03/09/memes-for-sale/?guccounter=1

https://techcrunch.com/wp-content/uploads/2021/03/poptart1redrainbowfix-1.gif?w=400

The creator of the Nyan Cat, Chris Torres, has organized an informal collection of meme originators — the creators or original popularizers of meme images — into a two-week-long auction of their works. Under the hashtag #memeconomy the creators of memes like Bad Luck Brian, Coughing Cat, Kitty Cat Dance, Scumbag Steve, Twerky Pepe and some others are finally finding a way to monetize the creation of genuine cultural phenomena that have been used freely for decades.

They’re mostly being hosted on booming new crypto art and collectibles platform Foundation, which launched in February and has already hosted $6 million in sales of over 1,000 NFTs. I have a lot to say about NFTs and can’t say them all here, but I found this project fascinating and wanted to note it. The fact is that memes are internet art (sorry). They are unique creations that took elements of participatory and performance art and injected them into the veins of the internet. In many ways, they have millions of creators, as the original editions may have planted the seed but every use and permutation gave them additional strands of DNA, crafting their cultural importance upload by upload. They have let us express ourselves — our desire, disgust, joy and lust — when words just wouldn’t suffice.

These “originals” are made original by the act of them being minted on the blockchain by the original artists. I know, it’s a distinction that may seem slim when the same images can be had anywhere at any time, but that’s the beauty of the re-organization that is happening within all of DeFi and crypto at the moment. We are stripping out layers of commerce and communication that benefited only platforms and participants that took part in the origination and sale of art from the perspective of frameworks like the DMCA and DRM. Those relationships are being rethought. The recapture of value for works that have already been broadly distributed has been historically relegated to “licensing them for t-shirts.” And extremely rarely elevated to the level of fine art sale.

#NFT #Idéologie_propriétaire #Culture_numérique #Spéculation #Blockchain

Affectiva raises $14 million to bring apps, robots emotional intelligence
▻https://techcrunch.com/2016/05/25/affectiva-raises-14-million-to-bring-apps-robots-emotional-intelligence/?guccounter=1

Affectiva, a startup developing “emotion recognition technology” that can read people’s moods from their facial expressions captured in digital videos, raised $14 million in a Series D round of funding led by Fenox Venture Capital. According to co-founder Rana el Kaliouby, the Waltham, Mass.-based company wants its technology to become the de facto means of adding emotional intelligence and empathy to any interactive product, and the best way for organizations to attain unvarnished insights (...)

#Affectiva #algorithme #biométrie #facial #reconnaissance #bénéfices

Google fires top AI ethics researcher Margaret Mitchell
▻https://techcrunch.com/2021/02/19/google-fires-top-ai-ethics-researcher-margaret-mitchell/?guccounter=1

Google has fired Margaret Mitchell, the founder and former co-lead of the company’s ethical AI team. Mitchell announced the news via a tweet. Google confirmed Mitchell’s firing in a statement to TechCrunch ; Google said : After conducting a review of this manager’s conduct, we confirmed that there were multiple violations of our code of conduct, as well as of our security policies, which included the exfiltration of confidential business-sensitive documents and private data of other (...)

#Google #éthique #discrimination #licenciement #surveillance #travail

Minneapolis police tapped Google to identify George Floyd protesters
▻https://techcrunch.com/2021/02/06/minneapolis-protests-geofence-warrant

The warrant ordered the search giant to turn over user account data. Police in Minneapolis obtained a search warrant ordering Google to turn over sets of account data on vandals accused of sparking violence in the wake of the police killing of George Floyd last year, TechCrunch has learned. The death of Floyd, a Black man killed by a white police officer in May 2020, prompted thousands to peacefully protest across the city. But violence soon erupted, which police say began with a masked (...)

#Google #activisme #géolocalisation #police #violence #ACLU

Google pledges grants and facilities for COVID-19 vaccine programs | TechCrunch
▻https://techcrunch.com/2021/01/25/google-pledges-grants-and-facilities-for-covid-19-vaccine-programs

https://techcrunch.com/wp-content/uploads/2021/01/GettyImages-1229247383.jpg?w=600

Encore de la philanthropie pour Google, mais on préfèrerait qu’ils payent leurs impôts et que les représentants élus dépensent allouent l’argent...

#GAFAM #philanthropie