A tech industry insider's guide to where to stay, eat, work, and play while visiting the tech scene's mothership, San Francisco. All products featured on WIRED are independently selected by our editors. However, we may receive compensation from retailers and/or from purchases of products through these links. You've probably read plenty of recent news stories about how San Francisco is a failed city. Our infrastructure is crumbling, our streets are scary, our social fabric is torn and frayed. Most of that stuff is false. Yes, San Francisco has issues, but they're the same problems nearly all US cities are facing as they struggle to reorient themselves to our new, post-pandemic economic reality. The "doom loop" narrative that's often repeated in the national press is a gross exaggeration. The truth is that San Francisco is thriving.

Conversational agents are designed to understand dialogs and generate meaningful responses to communicate with humans. After the popularity of ChatGPT, with its surprising performance and powerful conversational ability, commercial Large Language Models (LLMs) for general NLP tasks such as GPT-4 [1], etc., sprung up and brought the generative AI as a solution to the public view. These LLMs work quite well in content generation tasks, but their deficiency in fact-and-knowledge-oriented tasks is wellestablished by now [13]. These models themselves cannot tell whether the text they generate is based on facts or made-up stories, and they cannot always follow the given data and rules strictly and sometimes even modify the data at will, also called hallucination. The reasoning that these LLMs appear to perform is also at a very shallow level.

March 3 (Reuters) - Shares of artificial intelligence-based (AI) product makers zoomed on Friday, as a strong forecast from retail darling C3.ai Inc (AI.N) amplified an ongoing euphoria in the segment driven by the launch of OpenAI's ChatGPT. C3.a1 forecast better-than-expected revenue and profit for both the fourth quarter and fiscal year 2023, after its third-quarter results topped Wall Street estimates. Shares of the AI software provider were up 16% at $24.80, and were one of the top five trending stocks on StockTwits. If the gains hold, the stock is set to notch its strongest one-day gain in a month. "The company is starting to gain momentum in building significant enterprise opportunities in its pipeline with its suite of innovative enterprise AI solutions," said Wedbush analyst Daniel Ives.

During an extended weekend in Santa Monica, I came across a series of Coco delivery robots. These cooler-sized robots on wheels are parked in a few spots between Ocean Avenue and 3rd Street, waiting for their orders. This Coco delivery robot was spotted in Santa Monica on 2nd Street crossing Santa Monica Boulevard at about 10:30am. This article was also published in German.

Forecasting and decision-making are generally modeled as two sequential steps with no feedback, following an open-loop approach. In this paper, we present application-driven learning, a new closed-loop framework in which the processes of forecasting and decision-making are merged and co-optimized through a bilevel optimization problem. We present our methodology in a general format and prove that the solution converges to the best estimator in terms of the expected cost of the selected application. Then, we propose two solution methods: an exact method based on the KKT conditions of the second-level problem and a scalable heuristic approach suitable for decomposition methods. The proposed methodology is applied to the relevant problem of defining dynamic reserve requirements and conditional load forecasts, offering an alternative approach to current \emph{ad hoc} procedures implemented in industry practices. We benchmark our methodology with the standard sequential least-squares forecast and dispatch planning process. We apply the proposed methodology to an illustrative system and to a wide range of instances, from dozens of buses to large-scale realistic systems with thousands of buses. Our results show that the proposed methodology is scalable and yields consistently better performance than the standard open-loop approach.

Caroline Grace has always enjoyed vintage technology. An IT tech in the Mid-Ohio Valley, they collect retro games, laser discs and cassette tapes, but mostly, vinyl records. Their collection is in the thousands, and hundreds of those are video game soundtracks. "I've been a big fan of games all my life," says Grace. "Some of my earliest memories are playing games like Wonder Boy III: The Dragon's Trap and Goof Troop with my dad and brother. I get positive feelings from listening to the Wonder Boy III music now. I have a lot of pleasant memories of playing it with my family back in the day."

I had a couple questions I had wanted to ask Tom Siebel ever since his company, C3.ai (ticker "AI") went public in December of 2020. When I finally get my chance, I have to wait a moment, because Siebel is irrepressible. He often has the advantage on a reporter, leading with a kind of gusto that eclipses the standard question and answer protocol. "I had an epiphany this week," Siebel tells me, over a Greek coffee at a Greek joint in midtown Manhattan, Friday afternoon. Siebel, in the course of meeting with investors in New York and Boston this week, including Fidelity, has had reason to crystallize his thoughts about what artificial intelligence means, at least for his company and his customers. It is a good time to go on the road. At a recent stock price of $24.24, C3.ai shares are down sixty-five percent in the past year, twenty-two percent this year, and seventy-four percent since the IPO. And the entire software market is somewhat in the toilet as valuations are revised across the board.

JD: It doesn't really make sense to me, [but] I'm pleased that we have our 45 issued patents, our 40 pending patents, all in the US. And our 30 European patents are unaffected by the ruling. And honestly, look, I'm carrying on with my research. AR: I always thought the origin of the patent fight was not about money. My own reading of why it was so strongly fought was that it was not over commercial control but over credit--who did the science--and the truth.

The enduring appeal of both Kong and Godzilla has to do with their simplicity. "King Kong," made in Hollywood, débuted in 1933; "Godzilla," produced in Japan, came out in 1954. Both films relied on a stark and clarified premise: fantastic monsters let loose in ordinary human reality, which, in the light of their presence, is revealed to be even more hideous than the monsters themselves. That symbolic power, rather than their physical power, is the source of their enduring appeal, and it's the fundamental element that "Godzilla vs. Kong," the new mashup, directed by Adam Wingard, stomps into oblivion. The film is garishly overloaded with splices and grafts from other movies, other genres, and other premises, including a mythical setting and an evil corporation.

"Streets of Rage 4" is developed by three indie studios, Dotemu, Lizardcube and Guard Crush Games. The Streets series was Sega's answer to the popular arcade brawler, "Final Fight," made by Capcom. In the '90s, the popular series was exclusive to the Super Nintendo. "Streets of Rage" would be its dirty competition: grittier in style and crunchier in feeling. It would eventually become the game that could turn any Nintendo fan from grateful to envious.