The telescope’s studies could help end a long-standing disagreement over the rate of cosmic expansion. But scientists say more measurements are needed.
A veteran NASA scientist says his company has tested a propellantless propulsion drive technology that produced one Earth gravity of thrust.
Bostrom’s Deep Utopia
Posted in futurism, robotics/AI
Robin Hanson comments on Nick Bostrom’s new tome … has a great cover with a number of interesting questions and a subtitle that hints that it might address the meaning of life in a future where AI and robots can do everything. But alas, after much build up and anticipation, he leaves that question unanswered, with an abrupt oops, out of time on page 427. … He tries to address meaty topics like, what keeps life interesting? What is our purpose and meaning when the struggle is gone? Can fulfillment get full? But in each case, the pedagogy is more of a survey of all possible answers versus the much more difficult task of making specific predictions. (More)
👉 Researchers have developed a method called Selective Language Modeling (SLM), which trains language models more efficiently by focusing on the most relevant tokens.
Researchers introduce a new method called “Selective Language Modeling” that trains language models more efficiently by focusing on the most relevant tokens.
The method leads to significant performance improvements in mathematical tasks, according to a new paper from researchers at Microsoft, Xiamen University, and Tsinghua University. Instead of considering all tokens in a text corpus equally during training as before, Selective Language Modeling (SLM) focuses specifically on the most relevant tokens.
The researchers first analyzed the training dynamics at the token level. They found that the loss for different token types develops very differently during training. Some tokens are learned quickly, others hardly at all.
Engineers and developers at Intel are always working to push the boundaries of what’s possible, leaning on Moore’s Law — the idea that the number of transistors on a single chip will double every two years with a minimal increase in cost.
But over the last five years, Intel has had its ups and downs, demonstrated by the wavering value of its stock. It went from a high of $68 per share to more recently trading at $36 per share.
By investing $100 billion in American factories and innovation, the company hopes to turn that trend around. In late March, the company learned that it had secured $8.5 billion from the Biden administration, paired with another $11 billion in loans, with the goal of bringing chip manufacturing back to the U.S.
Approximately 22,500 exposed Palo Alto GlobalProtect firewall devices are likely vulnerable to the CVE-2024–3400 flaw, a critical command injection vulnerability that has been actively exploited in attacks since at least March 26, 2024.
CVE-2024–3400 is a critical vulnerability impacting specific Palo Alto Networks’ PAN-OS versions in the GlobalProtect feature that allows unauthenticated attackers to execute commands with root privileges using command injection triggered by arbitrary file creation.
The flaw was disclosed by Palo Alto Networks on April 12, with the security advisory urging system administrators to apply provided mitigations immediately until a patch was made available.
Indeed, the costs of building fabs in Germany, Japan, and the U.S. are higher than the costs of building fabs in Taiwan and TSMC has complained about it a number of times in the past. The company even had to delay production start at its Fab 21 near Phoenix, Arizona, due to problems with tools installation and negotiations with trade unions.
Therefore, if a TSMC customer wants to produce its chips at a specific location, then the foundry will charge a premium. How high is that premium will be remains to be seen, but last year a media report indicated that chips made in Arizona on TSMC’s N5 and N4 production nodes could be from 20% to 30% more expensive than the same chips produced in Taiwan.
Due to higher construction and operational expenses of fabs in Japan, Germany, and the U.S., TSMC plans to transfer these additional costs to its customers to sustain its target gross margin of 53%. Although American chip designers may not welcome the increased production costs in the U.S., they will probably manufacture chips intended for government and other markets less sensitive to price increases at the Arizona facility. Consequently, they should manage to pass on these higher costs to at least some of their end customers without jeopardizing their market competitiveness.
MRI scans of long COVID patients with brain fog suggest that the blood brain barrier may be leaky.