new The article critiques Eliezer Yudkowsky's pessimistic views on AI alignment and the scalability of current AI capabilities. The author argues that AI progress will be smoother and integrate well with current alignment techniques, rather than rendering them useless. They also believe that humans are more general learners than Yudkowsky suggests, and the space of possible mind designs is smaller and more compact. The author challenges Yudkowsky's use of the security mindset, arguing that AI alignment should not be approached as an adversarial problem. — lesswrong.com, 23h ago

new Whether either side is correct remains to be seen, technology of any description, no matter how potentially sentient, will in all likelihood be ethically determined by the uses or misuses it suffers at the hands of humans. — AIBC, 1d ago

new ...“The AI itself is neither good nor evil; it is simply blind and will maximize any goal you give it.”... — Wonderful Engineering, 1d ago

new However, when we turn to the future of artificial intelligence (AI) and psychiatry, we have little history to fall back on. With the advance and emergence of ChatGPT, now version 4, certain intellectual capabilities are beginning to rival—if not at times exceed—that of humans. The upgrades now are coming fast and furious. — Psychiatric Times, 23h ago

new This seems clearly false in the case of deep learning, where progress on instilling any particular behavioral tendencies in models roughly follows the amount of available data that demonstrate said behavioral tendency. It's thus vastly easier to align models to goals where we have many examples of people executing said goals. As it so happens, we have roughly zero examples of people performing the "duplicate this strawberry" task, but many more examples of e.g., humans acting in accordance with human values, ML / alignment research papers, chatbots acting as helpful, honest and harmless assistants, people providing oversight to AI models, etc. See also:... — alignmentforum.org, 23h ago

new All of these unfounded fears “have been part of the Silicon Valley culture for decades, often having futuristic sci-fi visions of technology. But they are not true. What real AI programmers really do is refine data sets and tweak features and algorithms to get the right results.”... — PostX News, 19h ago

new As noted here earlier, this new and disruptive ability of AI tools to move through a myriad of data — image, speech, text, etc. — is what makes modern artificial intelligence solutions worthy of being called “intelligent.”... — pymnts.com, 1d ago

new AI, AI, AI! The Tech Is Everywhere — But Why Is It Evolving So Fast? [Thoughts After Dark]... — thomasnet.com, 23h ago

new Hi Jennifer, and team. Good morning. Thanks for giving some information about the other BioStrand deal that you guys are working on. I was wondering if you could also comment on kind of what the pipeline looks like beyond that for future AI drug discovery deals? And also, how you expect the economics of those deals to look over time?... — Insider Monkey, 1d ago, Event

new It will be interesting to monitor progress. The current collaboration is for research not clinical purposes; the latter requires various FDA certifications. One earlier measure of success will be the generation of scientific papers, agreed Zhou. The hope is that many of the learnings will eventually make their way into healthcare. — HPCwire, 9h ago

new AI - might play these sort of games well, but I challenge AI to beat me in a contest of splitting wood. — Sott.net, 2d ago

new I’m sure most of you know about ChatGPT by now. It’s that buzzword floating around conversations whenever the topic changes to AI, whether it’s Microsoft’s... — TECHTELEGRAPH, 1d ago

new ...in which the computer is said to have passed if the human asking the question cannot discern between the replies of a human and the replies of a computer. The newer AI models are, from the perspective of content and semantics, impressively human-like. Yet something seems to be missing. — 7wData, 2d ago

new ...“The AI itself is not good or evil, it’s just blind, it will just optimise whatever goal you give it.”... — inews.co.uk, 2d ago

new The impact in the marketplace has, naturally, extended to academia. Digital health and e-medicine may be buzzwords now yet, according to Hong, academic institutions have varying definitions of the term in the curriculum programmes they offer: “Some gear it towards biomedical information automatics, and on the other spectrum, it’s geared towards AI computing, big data and statistics. But soon, digital health won’t be a separate discipline; it will be simply healthcare.”... — WAN-IFRA, 1d ago

new Microsoft:"We are going to put AI into your coffee and donut, and Outlook, and Edge, and Word, and...."Google:"Microsoft is putting AI into everything, we better do that too. Announce something."Amazon:"How can we dupe people into buying more stuff? I know! Make a catalog, but call it a "browser with AI !!"... — slashdot.org, 1d ago

new This being said, while AI is not expected to fully mimic humans’ abilities, it is becoming good at performing repetitive basic, or robotic tasks. — #NOWTESDEFI, 2d ago

new All in all, the paper strays from making any declarative statements about job impacts. It instead analyzes jobs that are more likely to have some “exposure” to AI generation, meaning it will take 50% less time to complete a job’s common task. All in all, most high-paying white collar work will find AI pushing into their fields. Those in science or “critical thinking” fields, as the paper calls it, will have less exposure, pointing to the modern AI’s utter limitations in creating novel content. Programmers and writers, on the other hand, are likely to see quite a lot of exposure. — Gizmodo, 1d ago

new ...for profiles related to generative AI or LLMs specifically. These jobs are getting created because of the emerging hype in the field. But what about the existing jobs that might get... — Analytics India Magazine, 1d ago

new He is uniquely suited for the task. He checks off all the academic boxes. He’s a known known in the world of AI. He’s also libertarian and thinks, deeply, about why and how it’s OK for people to want different things and be unconstrained in their individual choices. And he applies this thinking to AI. — nationalpost, 2d ago

new I think the dominant feature of the AI age is that life in itself—if not the world in itself—... — McKinsey & Company, 2d ago

new ..."In the specific case of the more mechanistic end of journalism—sports reports, financial results—I do think that AI tools are replacing, and likely increasingly to replace, human delivery," he said. — techxplore.com, 1d ago

new Among other criticisms, the paper argued that much of the text mined to build GPT-3 — which was initially... — VentureBeat, 1d ago

new The speakers noted that friction emerges in three areas: Technical, organizational, and financial. Technical friction is often well understood. AI and machine learning is hard. Building models is hard. Deploying those models on the right infrastructures is hard. — RTInsights, 1d ago

new Because it is a serious critique of economics, the book has to go back into the foundations of the discipline. A useful starting point is the distinction between ‘risk’ and ‘uncertainty’ which faces anybody thinking about, or preparing for, the future. Briefly, risk is about where there are probabilities attached to future events; uncertainty is where there are not. The latter events include the ‘unknown unknowns’, or as Kay and King call them ‘radical uncertainties’. — interest.co.nz, 1d ago

new Sure, math is limited, but the limitations of it are our limitations. There are no limitations inherent to math. — lesswrong.com, 2d ago

new The path of: [low level OpenAI employees think better about x-risk -> improved general OpenAI reasoning around x-risk -> improved decisions] seems high EV to me. — lesswrong.com, 1d ago

new Incentives often push people toward acceleration and away from voicing concerns. When people at AI labs, I think it’s worth commending. — lesswrong.com, 1d ago

new ...is becoming real. Artificial general intelligence is the next step.) It can explain various scientific concepts:And it can write basic academic essays. (Such systems are going to cause... — Inferse.com, 1d ago

new Artificial intelligence chatbots such as OpenAI’s ChatGPT and Microsoft’s BingGPT have generated lots of buzz lately, much of it centered around how helpful or harmful these tools might be for media companies. While the helpful camp welcomes the benefits of assisting reporters with initial research, aggregating data and... — Digital Content Next, 1d ago

new So our society tells all of these stories about how good, true, and beautiful it is to do whatever Pharma wants. — substack.com, 2d ago

new .... i.e humans failed to give the AI the correct goal. Yudkowsky has since stated the originally intended lesson was of... — lesswrong.com, 2d ago

new How could this technology be used in Apple’s products? Well, as it happens, I can think of a few ways that it might be deployed, not all of which are simply about just creating a chatbot. — Macworld, 1d ago

new The next step is to add phrases that your user is most likely to ask and how the bot responds to them. The bot builder offers suggestions, but you can create your own as well. The best part is that since the bots are NLP-powered, they are capable of recognizing intent for similar phrases as well. The more phrases you add, the more amount of data for your bot to learn from and the higher the accuracy. — web3newshubb.com, 2d ago

new ...“Modern robotics might then feel like a particular kind of cultural paradox, where the best kind of religion is the one that eventually involves no humans at all,” the anthropologist concludes. “But in this circularity of humans creating robots, robots becoming gods, and gods becoming human, we’ve only managed to, once again, re-imagine ourselves.”... — Wonderful Engineering, 2d ago

new Nevertheless, there are some challenges to beat first. For example, builders should guarantee AR windscreens and MR experiences aren’t too distracting for drivers behind the wheel. Furthermore, apps and instruments for constructing in VR will should be intuitive sufficient to make sure workforce adoption. If we are able to overcome these challenges, a future in XR automotive experiences might not be too far-off. — Crypto News Bay, 2d ago

new Despite Apple's best efforts, Siri has been idling in learning mode. Perhaps this is because Siri is still primarily built on Machine Learning and not Generative AI. It's the difference between learning and creating. — TechRadar, 2d ago, Event