Connect with us

Published

2 years ago

on

inflection point — The generative AI revolution has begunhow did we get here? A new class of incredibly powerful AI models has made recent breakthroughs possible.

HaomiaoHuang – Jan 30, 2023 12:00 pm UTC Enlarge / This image was partially AI-generated with the prompt “a pair of robot hands holding pencils drawing a pair of human hands, oil painting, colorful,” inspired by the classic M.C. Escher drawing. Watching AI mangle drawing hands helps us feel superior to the machines… for now. AurichAurich Lawson | Stable Diffusion reader comments 38 with 0 posters participating Share this story Share on Facebook Share on Twitter Share on Reddit

Progress in AI systems often feels cyclical. Every few years, computers can suddenly do something theyve never been able to do before. Behold! the AI true believers proclaim, the age of artificial general intelligence is at hand! Nonsense! the skeptics say. Remember self-driving cars?

The truth usually lies somewhere in between.

Were in another cycle, this time with generative AI. Media headlines are dominated by news about AI art, but theres also unprecedented progress in many widely disparate fields. Everything from videos to biology, programming, writing, translation, and more is seeing AI progress at the same incredible pace. Why is all this happening now?

Further ReadingThe basics of modern AIhow does it work and will it destroy society this year?You may be familiar with the latest happenings in the world of AI. Youve seen the prize-winning artwork, heard the interviews between dead people, and read about the protein-folding breakthroughs. But these new AI systems arent just producing cool demos in research labs. Theyre quickly being turned into practical tools and real commercial products that anyone can use.

Theres a reason all of this has come at once. The breakthroughs are all underpinned by a new class of AI models that are more flexible and powerful than anything that has come before. Because they were first used for language tasks like answering questions and writing essays, theyre often known as large language models (LLMs). OpenAIs GPT3, Googles BERT, and so on are all LLMs. Advertisement

But these models are extremely flexible and adaptable. The same mathematical structures have been so useful in computer vision, biology, and more that some researchers have taken to calling them “foundation models” to better articulate their role in modern AI.

Where did these foundation models came from, and how have they broken out beyond language to drive so much of what we see in AI today? The foundation of foundation models

Theres a holy trinity in machine learning: models, data, and compute. Models are algorithms that take inputs and produce outputs. Data refers to the examples the algorithms are trained on. To learn something, there must be enough data with enough richness that the algorithms can produce useful output. Models must be flexible enough to capture the complexity in the data. And finally, there has to be enough computing power to run the algorithms.

The first modern AI revolution took place with deep learning in 2012, when solving computer vision problems with convolutional neural networks (CNNs) took off. CNNs are similar in structure to the brain’s visual cortex. Theyve been around since the 1990s but werent yet practical due to their intense computing power requirements.

In 2006, though, Nvidia released CUDA, a programming language that allowed for the use of GPUs as general-purpose supercomputers. In 2009, Stanford AI researchers introduced Imagenet, a collection of labeled images used to train computer vision algorithms. In 2012, AlexNet combined CNNs trained on GPUs with Imagenet data to create the best visual classifier the world had ever seen. Deep learning and AI exploded from there.

CNNs, the ImageNet data set, and GPUs were a magic combination that unlocked tremendous progress in computer vision. 2012 set off a boom of excitement around deep learning and spawned whole industries, like those involved in autonomous driving. But we quickly learned there were limits to that generation of deep learning. CNNs were great for vision, but other areas didnt have their model breakthrough. One huge gap was in natural language processing (NLP)i.e., getting computers to understand and work with normal human language rather than code. Advertisement

The problem of understanding and working with language is fundamentally different from that of working with images. Processing language requires working with sequences of words, where order matters. A cat is a cat no matter where it is in an image, but theres a big difference between this reader is learning about AI and AI is learning about this reader.

Until recently, researchers relied on models like recurrent neural networks (RNNs) and long short-term memory (LSTM) to process and analyze data in time. These models were effective at recognizing short sequences, like spoken words from short phrases, but they struggled to handle longer sentences and paragraphs. The memory of these models was just not sophisticated enough to capture the complexity and richness of ideas and concepts that arise when sentences are combined into paragraphs and essays. They were great for simple Siri- and Alexa-style voice assistants but not for much else.

Getting the right training data was another challenge. ImageNet was a collection of one hundred thousand labeled images that required significant human effort to generate, mostly by grad students and Amazon Mechanical Turk workers. And ImageNet was actually inspired by and modeled on an older project called WordNet, which tried to create a labeled data set for English vocabulary. While there is no shortage of text on the Internet, creating a meaningful data set to teach a computer to work with human language beyond individual words is incredibly time-consuming. And the labels you create for one application on the same data might not apply to another task. Page: 1 2 3 4 5 6 7 8 Next → reader comments 38 with 0 posters participating Share this story Share on Facebook Share on Twitter Share on Reddit Advertisement Channel Ars Technica ← Previous story Next story → Related Stories Today on Ars

Related Topics:
Continue Reading

Sports

NHL Power Rankings: A new No. 1, and the game(s) of the month for all 32 teams

Published

22 mins ago

on

November 1, 2024

By

NHL Power Rankings: A new No. 1, and the game(s) of the month for all 32 teams

The second calendar month of the 2024-25 NHL season is here. But which games are the most tantalizing in the next 30 days?

For this edition of the Power Rankings, we’ve identified the game (or games) of the month for all 32 teams, whether it’s a rivalry matchup, a reunion with an old teammate or a key early season clash between top Stanley Cup contenders.

How we rank: A panel of ESPN hockey commentators, analysts, reporters and editors sends in a 1-32 poll based on the games through Wednesday, which generates our master list here.

Note: Previous ranking for each team refers to the previous edition, published Oct. 25. Points percentages are through Thursday’s games.

Previous ranking: 2
Points percentage: 90.00%

Nov. 7 vs. the Avalanche. The Jets’ promising 2024 playoff entry ended pretty quickly at the hands of the Avalanche this past spring. A single game in November will not be sufficient for vengeance, but as the Jets continue to prove themselves as a top contender, beating other top-tier teams is a requirement.

Next seven days: @ CBJ (Nov. 1), vs. TB (Nov. 3), vs. UTA (Nov. 5), vs. COL (Nov. 7)

Previous ranking: 3
Points percentage: 77.78%

Nov. 1, 2 vs. the Panthers. It’s another round of the NHL Global Series. Though the Devils-Sabres series in Prague proved one-sided, this pairing of Finn-loaded contenders will be a delight to the thousands in the building in Tampere.

Next seven days: vs. FLA (Nov. 1), vs. FLA (Nov. 2), vs. CHI (Nov. 7)

Previous ranking: 1
Points percentage: 72.22%

Nov. 23 at the Oilers. Not only is this the “Mark Messier matchup” — and one that would’ve made a great Stanley Cup Final this past spring — but these two teams have the shortest Cup-winner odds in their respective conferences, per ESPN BET. Perhaps we’ll get these two iconic franchises in the Cup Final this season?

Next seven days: vs. OTT (Nov. 1), vs. NYI (Nov. 3), vs. BUF (Nov. 7)

Previous ranking: 4
Points percentage: 77.78%

Nov. 19 at the Blues. The Wild and Blues fell just shy of the playoffs last season, and given the strength of the Central Division, one imagines there won’t be room for both of them in the 2025 postseason. Can the Wild get revenge for the Blues’ 4-1 win back on Oct. 15?

Next seven days: vs. TB (Nov. 1), vs. TOR (Nov. 3), vs. LA (Nov. 5), @ SJ (Nov. 7)

Previous ranking: 8
Points percentage: 68.18%

Nov. 17 vs. the Capitals. It’s not often that a player gets traded on the same day he’s scheduled to do a meet-and-greet with fans. But such was the case for goaltender Logan Thompson. He’ll make his return to Vegas with his new teammates for this one, after stopping 24 of 26 shots to beat his old team 4-2 on Oct. 15 in D.C.

Next seven days: vs. UTA (Nov. 2), @ EDM (Nov. 6)

Previous ranking: 10
Points percentage: 77.78%

Nov. 8 vs. the Penguins. Although Alex Ovechkin and Sidney Crosby are on the proverbial back-nine of their careers, both are still performing at a high level; that makes any Capitals-Penguins game a must-see (even if Sid’s Penguins are seriously lagging behind Ovi’s Capitals in the standings).

Next seven days: vs. CBJ (Nov. 2), @ CAR (Nov. 3), vs. NSH (Nov. 6)

Previous ranking: 6
Points percentage: 70.00%

Nov. 27 vs. the Capitals. Long live the Southeast Division. The Lightning and Capitals are both in playoff position as of this writing — and might well be at the end of the month when this game takes place, too. An important game for each club as they hope to remain in the postseason mix by season’s end.

Next seven days: @ MIN (Nov. 1), @ WPG (Nov. 3), @ STL (Nov. 5), vs. PHI (Nov. 7)

Previous ranking: 12
Points percentage: 77.78%

Nov. 27 vs. the Rangers. The Hurricanes’ playoff run this past spring was terminated by the Blueshirts, and this will be their first showdown with their divisional foes.

Next seven days: vs. WSH (Nov. 3), vs. PHI (Nov. 5), vs. PIT (Nov. 7)

Previous ranking: 11
Points percentage: 68.18%

Nov. 16, 19 against the Jets. November is a quirky month in the Panthers’ schedule: They have four occurrences of consecutive games against the same opponent. We’ll highlight occurrence No. 3 for this spot, as the Jets look every bit as awesome as their NFL namesakes do not. A potential Stanley Cup Final matchup.

Next seven days: vs. DAL (Nov. 1), vs. DAL (Nov. 2), vs. NSH (Nov. 7)

Previous ranking: 9
Points percentage: 61.54%

Nov. 12, 14 at the Panthers. The Panthers represent the standard to which all top Eastern Conference contenders are comparing themselves. So this pair of games in Sunrise will be critical for the Devils, a team that wants to count itself in that cohort.

Next seven days: @ CGY (Nov. 1), @ EDM (Nov. 4), vs. MTL (Nov. 7)

Previous ranking: 7
Points percentage: 59.09%

Nov. 5 vs. the Bruins. No matter which team is higher in the standings — or favored in the playoff series — the Maple Leafs have had major trouble beating the Bruins in recent seasons. As the Bruins’ season is off to a tepid start, it could be time for a statement victory for the Leafs.

Next seven days: @ STL (Nov. 2), @ MIN (Nov. 3), vs. BOS (Nov. 5)

Previous ranking: 16
Points percentage: 61.11%

Nov. 9 vs. the Oilers. There’s nothing like a heated playoff series to throw more fuel on a rivalry’s fire. So after the players (and fans) got after one another in the second round this past spring, what can we expect in the opening clash between these Pacific contenders?

Next seven days: @ SJ (Nov. 2), @ ANA (Nov. 5), @ LA (Nov. 7)

Previous ranking: 14
Points percentage: 63.64%

Nov. 29 at the Ducks. This SoCal rivalry isn’t as heated as it was when both teams were contenders, but there is always a little extra juice when they match up. And, with two of this season’s best uniform redesigns, it will be aesthetically pleasing, too.

Next seven days: vs. CHI (Nov. 2), @ NSH (Nov. 4), @ MIN (Nov. 5), vs. VAN (Nov. 7)

Previous ranking: 5
Points percentage: 55.00%

Nov. 3 vs. the Oilers. Few rivalries in professional sports are as consistently nasty as the Battle of Alberta. The Flames had to watch their cross-province antagonists get to within a game of a Stanley Cup this past spring. Will they continue their pursuit of derailing another Edmonton playoff run by taking another W from them, as they did back on Oct. 13?

Next seven days: vs. NJ (Nov. 1), vs. EDM (Nov. 3), @ MTL (Nov. 5), @ BOS (Nov. 7)

Previous ranking: 15
Points percentage: 55.56%

Nov. 12 at the Maple Leafs. The Battle of Ontario is another excellent NHL rivalry — the most recent flare-up occurring in February when Leafs defenseman Morgan Rielly took umbrage to Ridly Greig‘s enthusiastic empty-net goal. Will this next bout feature similar fireworks?

Next seven days: @ NYR (Nov. 1), vs. SEA (Nov. 2), @ BUF (Nov. 5), vs. NYI (Nov. 7)

Previous ranking: 18
Points percentage: 50.00%

Nov. 20 vs. the Predators. It was quite a low-scoring affair on Oct. 15 as the Seahawks beat the Titans 7-3 … oh wait, it was a hockey game. As it turns out, the total of 10 goals in that game has been matched twice since in Kraken games, both of which Seattle has won. Will it be another offensive onslaught in this return match against the Preds?

Next seven days: @ OTT (Nov. 2), @ BOS (Nov. 3), @ COL (Nov. 5)

Previous ranking: 13
Points percentage: 54.55%

Nov. 2, 15, 30 against the Golden Knights. The Golden Knights set the benchmark for early success as an expansion team, getting all the way to the Stanley Cup Final. The Utah Hockey Club isn’t technically an expansion team, but as the NHL’s newest club, they hope to chart a similar path of early success. A win (or three) against Vegas in November would signify they’re well on their way.

Next seven days: @ VGK (Nov. 2), @ WPG (Nov. 5), @ STL (Nov. 7)

Previous ranking: 27
Points percentage: 61.11%

Nov. 5-12, against the Sharks, Kings, Ducks, Kraken. To the surprise of some, the Blue Jackets remain on the (early) playoff bubble in the Eastern Conference. An extended West Coast trip can help bring a team together, and grabbing some Ws doesn’t hurt either.

Next seven days: vs. WPG (Nov. 1), @ WSH (Nov. 2), @ SJ (Nov. 5)

Previous ranking: 19
Points percentage: 45.45%

Nov. 29 at the Stars. Enjoy watching elite puck-moving defensemen? Circle this game on your calendar, as Colorado’s Cale Makar will visit Dallas’ Miro Heiskanen.

Next seven days: @ NSH (Nov. 2), vs. SEA (Nov. 5), @ WPG (Nov. 7)

Previous ranking: 20
Points percentage: 45.45%

Nov. 12, 16 against the Bruins. Goalie goals are a thing of beauty. Goalie fights are a thing of beauty … in another way. Jordan Binnington has been on the verge of one for years — and nearly got in one against the Senators this week — while Bruins netminder Jeremy Swayman also has the fighting spirit, as one might say.

Next seven days: vs. TOR (Nov. 2), vs. TB (Nov. 5), vs. UTA (Nov. 7)

Previous ranking: 17
Points percentage: 40.91%

Nov. 9 vs. the Senators. It’s Linus Ullmark-Jeremy Swayman reunion night! Though Ullmark didn’t play for the B’s as long as Steven Stamkos did for the Lightning, he did win the Vezina Trophy for a team that set an NHL record for regular-season dominance. What will the reception be like at TD Garden?

Next seven days: @ PHI (Nov. 2), vs. SEA (Nov. 3), @ TOR (Nov. 5), vs. CGY (Nov. 7)

Previous ranking: 24
Points percentage: 45.00%

Nov. 2, 5 against Red Wings, Senators. For the past few seasons, we’ve been waiting for one of these three rising Atlantic Division teams to take the next step and qualify for the playoffs. Is this the season? These two games for the Sabres will be a good barometer of where they stand in particular.

Next seven days: vs. NYI (Nov. 1), @ DET (Nov. 2), vs. OTT (Nov. 5), @ NYR (Nov. 7)

Previous ranking: 21
Points percentage: 50.00%

Nov. 30 at the Avalanche. The Oilers and Avalanche are not exactly where many predicted they’d be in the standings. And the loss of Connor McDavid for the next two to three weeks doesn’t help matters for Edmonton. But with McDavid back in time (hopefully) for this matchup, this should be a good measuring stick for both clubs.

Next seven days: @ CGY (Nov. 3), vs. NJ (Nov. 4), vs. VGK (Nov. 6)

Previous ranking: 26
Points percentage: 45.00%

Nov. 9 vs. the Rangers. At some point, the Red Wings have to turn a corner back toward contention. Will this Original Six matchup be that turning point, following two losses to the Blueshirts in October?

Next seven days: vs. BUF (Nov. 2), @ CHI (Nov. 6)

Previous ranking: 25
Points percentage: 50.00%

Nov. 3, 19 against the Blackhawks. There was no uncertainty about who should be the No. 1 pick in the 2023 NHL draft; and sure enough, Connor Bedard was selected by the Blackhawks. However, the Ducks’ selection of Leo Carlsson at No. 2 raised some eyebrows at the time. Not anymore, as the Swede is second on his team in scoring this season. This pair of matchups will be a showcase for two of the league’s brightest young stars.

Next seven days: vs. CHI (Nov. 3), vs. VAN (Nov. 5)

Previous ranking: 22
Points percentage: 40.00%

Nov. 3 at the Rangers. Perhaps this Sunday matinee against their archrivals will stir something in the Isles, who haven’t gotten off to the best start in the first full season with Patrick Roy behind the bench.

Next seven days: @ BUF (Nov. 1), @ NYR (Nov. 3), vs. PIT (Nov. 5), @ OTT (Nov. 7)

Previous ranking: 28
Points percentage: 35.00%

Nov. 2, 11 against the Avalanche. The truth is, we could list all 15 games in November here; high preseason expectations have led to underwhelming early results, making every game count from here. But we’ll pick this pair against the Avalanche, as Nathan MacKinnon & Co. have been in that upper tier of contention (to which the Preds aspire) for the past handful of seasons.

Next seven days: vs. COL (Nov. 2), vs. LA (Nov. 4), @ WSH (Nov. 6), @ FLA (Nov. 7)

Previous ranking: 30
Points percentage: 40.91%

Nov. 26 vs. the UHC. At the 2022 NHL draft in Montreal, the Canadiens shocked many of their fans in attendance by selecting Juraj Slafkovsky first overall. Two picks later, the Coyotes took Logan Cooley. Much has happened since then — including the Coyotes transforming into the Utah Hockey Club — but this game will another showcase for two rising stars from that class.

Next seven days: @ PIT (Nov. 2), vs. CGY (Nov. 5), @ NJ (Nov. 7)

Previous ranking: 29
Points percentage: 31.82%

Nov. 6 vs. the Red Wings. Two iconic franchises. One of the most aesthetically pleasing uniform matchups in pro sports. No, they’re not exactly on top of the standings as they once were, but there’s something special about this old-school Norris Division showdown.

Next seven days: @ LA (Nov. 2), @ ANA (Nov. 3), vs. DET (Nov. 6), @ DAL (Nov. 7)

Previous ranking: 31
Points percentage: 40.91%

Nov. 7 at the Lightning. Matvei Michkov, the Flyers’ first-round pick in 2023, has taken the NHL by storm, with nine points through 10 games. In this matchup, he’ll be sharing the ice with two of the best Russian players in recent history, winger Nikita Kucherov and goaltender Andrei Vasilevskiy.

Next seven days: vs. BOS (Nov. 2), @ CAR (Nov. 5), @ TB (Nov. 7)

Previous ranking: 23
Points percentage: 37.50%

Nov. 19 vs. the Lightning. Jake Guentzel lost his first game as a non-Penguin in Pittsburgh (with the Hurricanes, after last season’s trade deadline). Will his second trip back to the visitors locker room go any better — and how will fans greet him?

Next seven days: vs. MTL (Nov. 2), @ NYI (Nov. 5), @ CAR (Nov. 7)

Previous ranking: 32
Points percentage: 33.33%

Nov. 16 at the Penguins. In 2016, these two teams met in the Stanley Cup Final. This season, they’re meeting in the basement of the standings — with one of their appearances more shocking than the other’s. Can the Sharks punctuate this mini East Coast trip with a win against Sidney Crosby & Co.?

Next seven days: vs. VAN (Nov. 2), vs. CBJ (Nov. 5), vs. MIN (Nov. 7)

Continue Reading

Environment

Exxon earnings beat, increases fourth-quarter dividend

Published

2 hours ago

on

November 1, 2024

By

Exxon earnings beat, increases fourth-quarter dividend

An Exxon gas station is seen in the Brooklyn borough of New York City on Oct. 6, 2023.

Michael M. Santiago | Getty Images

Exxon Mobil beat third-quarter earnings expectations, as the oil major reached its highest liquids production level in more than four decades.

Here is what Exxon reported for the third quarter compared with what Wall Street was expecting, based on a survey of analysts by LSEG: 

  • Earnings per share: $1.92 adjusted, vs. $1.88 per share expected.
  • Revenues: $90 billion, vs. $93.94 billion expected

The oil major booked net income of $8.61 billion in the quarter, or $1.92 per share, down about 5% compared to $9.1 billion, or $2.25 per share, in the year-ago period. Exxon’s profits have declined as refining margins and natural gas prices have pulled back from from historically high levels in 2023.

The company returned $9.8 billion to shareholders in the quarter and increased its fourth-quarter dividend to $0.99 per share.

Exxon said it has reached its high production level in more than 40 years at 3.2 million barrels per day.

The oil major’s stock rose about 1% in pre-market trading. Exxon shares have gained 16.8% this year.

This is a developing story. Please check back for updates.

Continue Reading

Environment

Chevron beats earnings expectations, returns more than $7 billion to shareholders

Published

2 hours ago

on

November 1, 2024

By

Chevron beats earnings expectations, returns more than billion to shareholders

Chevron beats earnings expectations, returns more than $7 billion to shareholders

Chevron beat third-quarter earnings and revenue expectations, returning a record amount of cash to shareholders.

Shares were up 2.6% in the premarket following the report’s release.

The oil major’s quarterly profit, however, declined substantially compared to the year-ago period due to lower margins on refined product sales, lower prices and the absence of favorable tax times.

Chevron is aiming to streamline its portfolio, with asset sales in Canada, Congo and Alaska expected to close in the fourth quarter of 2024. The company is also target $2 billion to $3 billion in cost reductions from 2024 through the end of 2026.

Here is what Chevron reported for the third quarter compared with what Wall Street was expecting, based on a survey of analysts by LSEG: 

  • Earnings per share: $2.51 adjusted, vs. $2.43 expected
  • Revenue: $50.67 billion, vs. $48.99 billion expected

Chevron’s net income came in at $4.49 billion, or $2.48 per share, down 31% from $6.53 billion, or $3.48 per share, in the third quarter of 2023. When adjusted for foreign currency impacts, the company reported earnings of $2.51 per share, solidly topping Wall Street’s expectations for the quarter.

Chevron booked revenues of $50.67 billion, also beating Street expectations but declining 6% from the $54.1 billion reported in the third quarter last year.

The oil major returned a record $7.7 billion to shareholders in the quarter, including $4.7 billion in share buybacks and $2.9 billion in dividends.

Chevron produced 3.36 million oil-equivalent barrels per day in the quarter, a 7% increase over the third quarter of 2023, driven by record output in the Permian Basin.

Chevron’s stock is largely flat for the year, underperforming the S&P 500 energy sector which has gained more than 6%. Shares have struggled to gain ground as uncertainty looms over the company’s pending $53 billion acquisition of Hess.

The Federal Trade Commission has cleared the deal, though it prohibited John Hess from joining Chevron’s board.

Chevron remains locked in a dispute with Exxon Mobil, which is claiming a right of first refusal over Hess Corp.’s lucrative oil assets in Guyana. If an arbitration court rules in Exxon’s favor, Chevron’s acquisition of Hess would fail to close.

Continue Reading

Trending