Meta chief AI scientist Yann LeCun says Elon Musk risks 'killing breakthrough innovation' at xAI. Here's why.
Musk sparked a conversation about the roles of researchers and engineers at tech companies on Tuesday in an X post. Musk said his AI startup, xAI, would ditch the "researcher" job title in favor of "engineer."
"This false nomenclature of 'researcher' and 'engineer', which is a thinly-masked way of describing a two-tier engineering system, is being deleted from @xAI today," Musk said. "There are only engineers. Researcher is a relic term from academia."
This false nomenclature of 'researcher' and 'engineer', which is a thinly-masked way of describing a two-tier engineering system, is being deleted from @xAI today.
There are only engineers.
Researcher is a relic term from academia. https://t.co/yNZSdVXGhY
— Elon Musk (@elonmusk) July 29, 2025
Two days later, LeCun shared a screenshot of Musk's X post on LinkedIn with a multi-paragraph response.
"If you make no distinction between the two activities, if you don't evaluate researchers and engineers with different criteria, you run the risk of killing breakthrough innovation," LeCun said. "True breakthroughs require teams with a long horizon and minimal constraints from product development and management."
Musk isn't the first person to question the distinction between AI researchers and engineers. Other leading AI companies have, too.
In a 2023 X post, OpenAI President Greg Brockman said that the company didn't want to put its workers into such defined buckets. Instead, the ChatGPT-maker settled on the phrase "Member of Technical Staff."
Anthropic, which makes Claude, also uses "Member of Technical Staff" as a job title.
"While there's historically been a division between engineering and research in machine learning, we think that boundary has dissolved with the advent of large models," Anthropic says on its careers page.
LeCun, however, says the research labs that shaped what the science and tech industries have become were all separate from engineering divisions.
"The industry research labs of yore that have left an indelible mark on scientific and technological progress (Bell Labs Area 11, IBM Research, Xerox PARC, etc) were all research divisions that were clearly separate from engineering divisions," LeCun said.

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
6 minutes ago
- Yahoo
I Asked ChatGPT How Long It Would Take for Jeff Bezos To Spend All His Money
Jeff Bezos has an estimated net worth of $214 billion, according to the Bloomberg Billionaires Index. It's such an outrageous number that it doesn't feel real. Read More: Find Out: So naturally, I asked ChatGPT: How long would it take for Jeff Bezos to spend all his money? The answers were both enlightening and absurd. Wait, Isn't Most of His Wealth Tied Up in Stock? Yes, exactly. The majority of Bezos's fortune comes from Amazon stock and other assets, and not from piles of cash in a bank account. If he wanted to spend it all, he'd have to liquidate assets, deal with taxes, and probably move markets in the process. But for the sake of this thought experiment, I asked ChatGPT to treat Bezos' $250 billion as fully liquid. No taxes. No asset sales. Just pure, spendable wealth. Scenario 1: Spending $1,000 a Day 'If Bezos spent $1,000 every day, it would take him roughly 685,000 years to go broke,' ChatGPT calculated. That's longer than the entire existence of Homosapiens. He could buy a new iPhone every morning for the next 600,000 years and still have money left over for dinner. Discover More: Scenario 2: Spending $1 Million a Day Let's say he took it up a notch and started spending $1 million a day, luxury jets, gold-plated everything, villas on every continent. 'At that rate, it would take 685 years to run out,' ChatGPT said. That means he'd still be spending in the year 2710. Long after we've colonized Mars or uploaded our consciousness to the cloud. Scenario 3: Giving Away $1 Billion Per Year Now we're in philanthropic territory. If Bezos committed to giving away $1 billion every year, how long would that last? 'About 250 years,' ChatGPT said. To put that in perspective: The Bill & Melinda Gates Foundation spent about $7 billion in 2022. Bezos could do that every year for two and a half centuries. Scenario 4: Giving It All Away Tomorrow Let's say Bezos liquidated everything and gave away $250 billion in one giant act of generosity. Here's what that could cover, according to ChatGPT: End world hunger for 7.5 years, based on UN World Food Programme estimates of $33 billion a year. Wipe out all U.S. out-of-pocket medical expenses for 2023, which totaled around $505 billion, according to Centers for Medicare & Medicaid Services. Give every household in the U.S. a $1,750 check (based on ~143 million households). Pay for one year of tuition for 22 million public college students, based on national averages. That's real-world, tangible impact. And yet, he'd still likely make some of it back. So … Could Bezos Ever Go Broke? 'Not unless he really tried,' ChatGPT said. Even a billionaire-level spending spree encounters a bigger problem: Jeff Bezos' wealth continues to grow. In 2023, Amazon's stock surged about 80%, according to Investor's Business Daily, driven by renewed momentum in cloud, advertising, and AI. With most of Bezos's fortune tied to Amazon stock, that massive return added tens of billions to his net worth without him doing a thing. At that scale, wealth becomes self-sustaining, and it compounds. 'Let's say he spends $1 billion a year,' ChatGPT explained. 'Even that would be hard to keep up with. If his investments earned a modest 7% return annually, a pretty average stock market gain, he'd still come out ahead. On a $200 billion fortune, that's $14 billion in growth a year. So even after spending $1 billion, he'd be $13 billion richer.' In other words, he would be earning faster than he could spend it. So asking, 'Could Bezos ever spend it all?' overlooks the central fact: it's not just the size of his fortune, it's how fast it regenerates. More From GOBankingRates Mark Cuban Warns of 'Red Rural Recession' -- 4 States That Could Get Hit Hard Clever Ways To Save Money That Actually Work in 2025 7 Tax Loopholes the Rich Use To Pay Less and Build More Wealth This article originally appeared on I Asked ChatGPT How Long It Would Take for Jeff Bezos To Spend All His Money Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data


Digital Trends
7 minutes ago
- Digital Trends
Apple is eyeing a ChatGPT-like search, but it must focus beyond Siri
It's no secret that Apple is currently struggling to deliver a smash-hit AI product, the way Google has served with Gemini, or Microsoft has achieved with Copilot. The company has been trying a similar overhaul with Siri, but those plans have been beset by delays, and it is only expected to see the light of day in late 2026. The delay spooked Apple to such an extent that the company inked a stopgap deal with OpenAI, which helped integrate ChatGPT with Siri, and broadly, with the Apple Intelligence stack. But it seems Apple is working on a radical in-house solution, one that would essentially be a watered-down approach to ChatGPT, but with internet search capabilities. Siri, but flavored like ChatGPT Lite? According to Bloomberg, a newly formed Answers, Knowledge and Information (AKI) team at Apple is working on a ChatGPT-inspired search framework for Siri. 'While still in early stages, the team is building what it calls an 'answer engine' — a system capable of crawling the web to respond to general-knowledge questions,' says the report. In addition to Siri, Apple reportedly plans to integrate the search functionality within Spotlight and Safari, as well. Spotlight has already received a massive functional upgrade in macOS Tahoe, so it won't be surprising to see it evolve into a universal answering hotspot, one that covers local data and information sourced from the internet. Recommended Videos It may sound chaotic at first, but it's not entirely alien. How does Siri, Spotlight, or Safari know when I want an AI to answer my query, or launch a web Search? Well, look no further than Dia. The universal search box in the AI-focused browser dynamically switches between 'chat' and 'Google' mode as you type your search keywords. When you type 'Birkin bag' in the text field, it defaults to web search mode. But as you type 'where to buy a Birkin bag,' the search field automatically switches to chat mode and offers the answer, just the way ChatGPT or any other AI answering engine like Perplexity would handle your questions. Right now, when you summon Siri on your iPhone and ask it a question that requires searching the internet or just pulling knowledge from an information bank, it opens a prompt box asking whether the question can be offloaded to ChatGPT. Once you agree, ChatGPT kicks into action and offers the required information. Of course, it's not seamless. With Siri gaining web search capabilities and enhanced natural language comprehension (akin to a ChatGPT or Gemini), it would be much easier for users to simply ask anything they want and get it answered. In its current state, Siri feels like a relic of the past, especially when compared to products such as Google's Gemini Live or ChatGPT's voice mode. In fact, Gemini works better on iPhones than Siri. As far as Apple's plans go, building something as advanced as ChatGPT or Gemini seems like a far-fetched goal. As per Bloomberg, plans for 'LLM Siri' have kept running into delays, and the recent exodus of top AI talent casts more doubts over Apple's ambitions of reimagining Siri for the AI era. It's not just about a phone assistant Building a next-gen virtual assistant – just the way Google Assistant has evolved into Gemini, or Copilot at Microsoft – is not the only area where Apple is currently lagging far behind the competition. In fact, Big Tech is now as focused as much on chatbots as it is on web browsers. Agentic workflows are now being seen as the next big thing in the field of AI. In a recent interview, co-founder and chief of Perplexity, Aravind Srinivas, explained why browsers are more suitable for AI than AI chatbots and apps: 'You get full transparency and visibility, and you can just stop the agent when you feel like it's going off the rails and just complete the task yourself, and you can also have the agent ask for your permission to do anything. So that level of control, transparency, trust in an environment that we are used to for multiple decades, which is the browser.' Unfortunately, Apple is severely lagging behind in the browser wars. With the introduction of AI Mode in Search and deeply integrating Gemini across its Workspace ecosystem, Google has changed how deeply AI can change web browsing and web-based workflows. Safari desperately needs an AI overhaul Upstart browsers such as Dia and Perplexity's Comet have proved that the era of legacy tools such as extensions is coming to an end. Soon, skills and custom agents will take over. Less than a week ago, Microsoft introduced Copilot Mode in Edge. I have spent a few days with the new AI-powered tools in Edge, and I believe it's a bold (and dramatically more practical) new direction for web browsers. In comparison, Safari misses out on any such AI-driven experiences. From a context-aware sidebar to multi-tab contextual actions, Apple's browser is sorely missing out on the conveniences that AI is bringing to modern age web browsers. Assuming Apple succeeds at building its own ChatGPT-like answer engine, it would take a massive undertaking to build meaningful features around it in Safari. Right now, what Apple needs to do is not just build an answering engine, but pay close attention to the competition. I am sure Apple is monitoring the shifting landscape of AI agents and browsers. It simply has to pick up pace, or as CEO Tim Cook hinted at in a recent all-hands meeting, the company 'will make the investment to do it.' Will Apple acquire a hot AI lab like Perplexity or Anthropic? Only time will tell, but the company certainly has to take a more holistic approach with AI than just focus on building the next great AI chatbot.
Yahoo
26 minutes ago
- Yahoo
Veteran fund manager turns heads with new Meta Platforms stock price target
Veteran fund manager turns heads with new Meta Platforms stock price target originally appeared on TheStreet. On any given day, nearly half the world is on Meta Platforms () . The parent of Facebook, Instagram, Messenger, Threads and WhatsApp, Meta has dramatically changed the way we live — for better or worse. 💵💰Don't miss the move: Subscribe to TheStreet's free daily newsletter 💰💵 The Menlo Park, Calif., social media giant is moving aggressively with artificial intelligence into areas that some people find just a bit spooky. "Over the last few months, we've begun to see glimpses of our AI systems improving themselves, and the improvement is slow for now but undeniable," Chief Executive Mark Zuckerberg said during the second-quarter earnings call. "And developing superintelligence, which we define as AI that surpasses human intelligence in every way, we think is now in sight." To build this future, Zuckerberg said the company had set up Meta Superintelligence Labs, "which includes our foundations, product, and fair teams as well as a new lab that is focused on developing the next generation of our models." Meta CEO says AI creating greater efficiency The company seems to be on the right track. Meta Platforms handily beat Wall Street's quarterly forecasts and the stock is up nearly 52% this year. Zuckerberg said the quarter's strong performance stemmed largely from "AI unlocking greater efficiency and gains across our ad system." More AI Stocks: Google plans major AI shift after Meta's surprising $14 billion move Meta delivers eye-popping AI announcement Veteran trader surprises with Palantir price target and comments Chris Versace, TheStreet Pro portfolio's lead manager, liked what he heard from Zuck & Co. and boosted his price target for Meta shares to $850 from $725. "The hike in our target reflects a combination of a few factors, including the company's still expanding reach, advertising gains, and improving monetization efforts," he said. With trade tensions appearing to cool as the Trump administration announces deals and other conversations are extended, Versace said, "the reduced uncertainty tied to an economy that is still expanding should foster more normalized advertising spend." "In that environment, we continue to see Meta benefiting, especially as advertisers continue to shift increasingly toward digital platforms," he said. "They also want to tap the 3.48 billion in daily active people across all of Meta's platforms vs. 3.27 billion at the end of June 2024." While many were in awe of the company's impressive top- and bottom-line beats, Versace said the year-over-year five-percentage-point improvement in its consolidated operating margin caught his eye. "We should see further improvement in the second half of 2025, but higher expenses will be a headwind as Meta invests further in AI capacity and talent," he noted. "However, we see this as another instance of the company investing today to drive its business and profits higher in the coming quarters." Fund manager: Capital spending at ludicrous levels The future ain't gonna come cheap. Meta, Amazon () , Alphabet () and Microsoft () are set to spend as much as a cumulative $364 billion in their respective 2025 fiscal years, up from their prior estimates of around $325 billion. Meta narrowed its 2025 capital-spending outlook to a range of $66 billion to $72 billion, up $30 billion year over year at the midpoint, Versace investment firms also noted Meta's capital expenditures, including Scotiabank, which raised its price target on the company's shares to $685 from $675 and affirmed a sector-perform rating on the shares. Meta posted its largest beat in several years, and the revenue comparison in fiscal Q1 2026 against the year-earlier period looks "much easier," the firm said, according to The Fly. However, the firm said it remains on the sidelines until it sees the company progress on profiting from its higher capital expenditures while also outpacing the hit to earnings. Doug Kass, a longtime hedge-fund manager and TheStreet Pro contributor, picked up on the capex issue in a lengthy comment on X. "META has burned through $30 billion in the last two quarters," he wrote. "Cash is now down to $12 billion" and "the capital and operating expense are now becoming a burden - nearly all of depreciation schedules [are] an absolute joke." Kass called the amount of capital spending "ludicrous ... in part because of the accounting gimmicks that hide its true cost. "At some point this will start getting into the numbers, even with the gimmicks," he said. "And I still do not know where all the space, power, water, and even ancillary equipment will be found for all of this. There is only so much stuff the air conditioning manufacturers can make!" Kass cited Amazon's latest earnings report, noting that "the capital expense is starting to bite and hit the numbers." "Margins at [Amazon Web Services] are finally starting to feel it, and they disappointed," he added. "Growth was nothing to write home about either. That is why the stock is selling off."Veteran fund manager turns heads with new Meta Platforms stock price target first appeared on TheStreet on Aug 3, 2025 This story was originally reported by TheStreet on Aug 3, 2025, where it first appeared. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data