WSJ : Steve Ballmer, the Non-Investing Guru of Investing

Steve Ballmer, the Non-Investing Guru of Investing
The former CEO’s unconventional and hugely profitable strategy: own a lot of Microsoft

The ultrawealthy tend to follow a time-honored model of investing: diversifying.

Much like universities and pensions, the richest people on earth spread their money across stocks, bonds and a cadre of alternatives, such as private equity, hedge funds and real estate.

Or at least most of them do. There is one very glaring exception: Steve Ballmer.

The former Microsoft MSFT -0.10%decrease; red down pointing triangle chief executive has put nearly all his eggs in one basket and kept them there for years. Ballmer has virtually no exposure to alternative investments. Instead, he keeps more than 80% of his portfolio in Microsoft stock and the rest in stock index funds.

As the stock market has climbed to new heights and Microsoft has transformed into one of the best-performing stocks of the past decade, Ballmer’s investment strategy has proven incredibly effective.

“Microsoft’s outperformed just about every other asset I could have owned,” said Ballmer, Microsoft’s top individual shareholder, in a recent interview with The Wall Street Journal. “It’s a little hard to say that it hasn’t worked out.”

Ballmer, 68 years old, served as CEO from 2000 until stepping down in 2014, after coming under pressure for not moving fast enough to compete with Apple and Google in key consumer markets. During his tenure at Microsoft’s helm, the company’s stock price fell by about a third. When he left the company, Microsoft’s market capitalization was around $300 billion.

These days, powered by its cloud business and the company’s prescient bets on artificial intelligence, it is more than $3 trillion.

To put it another way, Microsoft has gained an average 28.8% annually, including dividends, over that rough time span. In that same time, the S&P 500 returned an average 12.9% a year, including dividends. Endowments of more than $1 billion gained an average 7.8% a year, according to Wilshire Trust Universe Comparison Service.

Few have benefited more than Ballmer, who calls himself “a non-investor investor.” He now ranks ninth on the Bloomberg Billionaires Index list of the world’s richest people, above even Warren Buffett, with an estimated net worth of more than $150 billion. This summer, he briefly surpassed his former boss Bill Gates, who gives regularly to his foundation in a mix of cash, Microsoft stock and other securities.

As simple as his investment philosophy is, it took Ballmer decades to implement. In the 1980s, he bet on individual stocks such as Colgate-Palmolive on the advice of CNBC’s Jim Cramer, a college buddy who was Ballmer’s financial adviser at Goldman Sachs. He then tried diversifying but, in part, found it difficult to identify money managers who consistently outperform. So he dumped most of his alternative investments.

His fortune now is parked almost exclusively in stock index funds—and Microsoft. (He declined to say how big his stake is.)

Connecting via Microsoft Teams from his office in Bellevue, Wash., Ballmer spoke with the Journal about how he manages his money.

WSJ: Can you walk us through your investment strategy?

Ballmer: I own an index fund. We moved it back to be just the U.S. and Europe. Keep it simple. Maybe we own Japan too. We’re mostly dialing out of private equity, but you can’t get out of that overnight. The only stock I really study still is Microsoft, because that’s still overwhelmingly, overwhelmingly, overwhelmingly the No. 1 thing that I own.

I like it. It’s simple. We’ve been very blessed financially. What I seek in this instance is not to have to spend a lot of time, anxiety, brain power in an area where we’re blessed enough if we make 7% because that’s the standard return on the S&P over the long run.

[My approach] partly was shaped by Warren Buffett, who argues most people should just buy an S&P index fund. At least for us [Ballmer and his wife, Connie], making an extra 0.7% compounded isn’t going to change our lives. At the end of the day, we’ll be able to give our kids all that we want to give them, whatever that winds up being. I make no comment on that. Everything else we do is just a balance between how much goes to philanthropy and how much goes to the government.

WSJ: I feel like I would be tempted to check Microsoft’s stock price every day. Do you do that?

Ballmer: Oh, I check it, you know, not infrequently. But not because it matters financially. The stock market is two things to people. One is—how am I doing financially? And the other is a little bit like studying the box score—did my team win yesterday?

WSJ: You mentioned the S&P 500. Is that literally the index you own? Or do you have a more bespoke bucket of products of different indexes?

Ballmer: If it’s not the S&P, it’s the Russell 2000 or some broad mirror of the market. We do a little bit of tax harvesting. You could read about that, if you like, on ProPublica. They have my tax return and wrote it up.

[ProPublica reported last year, based on a trove of leaked Internal Revenue Service data, that Ballmer and other billionaires use tax-loss harvesting strategies to reduce the taxes they pay. It estimated Ballmer benefited from at least $138 million in tax savings between 2014 and 2018 via trades Goldman Sachs made for him that violated the spirit, if not the letter, of a law prohibiting so-called wash sales.

The inquiries from ProPublica prompted Goldman to halt such trades, and the bank said at the time it had conducted a review and found a small percentage of tax-efficient trades it makes for clients were “inadvertently made in a manner inconsistent with our strategy.” Goldman also said it tries to provide investment advice “consistent with both the letter and the spirit” of relevant laws and regulations.

A Ballmer spokeswoman said in a statement, “Steve takes his responsibility to pay taxes very seriously and has always paid what he owes.”]

We did decide to not own oil and gas stocks anymore. That’s a decision we made in the last year or so as we’ve gotten more involved in climate. It’s not going to change the world of valuations for those stocks, but it feels more value consistent. So those would be the perturbations in what we do.

WSJ: Do you ever discuss your investment philosophy and strategy with other very wealthy people? Do your friends think you’re crazy?

Ballmer: Well, no, not at all, because it’s a little hard to say that it hasn’t worked out. I don’t know how anybody else has done, because everybody’s got their strategy, and it’s hard to peek inside their curtains.

In terms of a non-Microsoft piece, the indexes have done pretty well. And people say I’ve done well on my Clippers investment. (Ballmer bought the NBA’s Los Angeles Clippers in 2014, after he stepped down from Microsoft.)

WSJ: You paid $2 billion for the Clippers. There are rankings that put the value of the team now around $4 billion or $5 billion.

Ballmer: Look, if it’s a double in 10 years, it’s basically an index fund. And oh, by the way, it’s really underperformed Microsoft. I’m not saying it’s a bad investment—don’t get me wrong. But again, compared to Microsoft—I mean, honestly, almost any investment compared to Microsoft just pales.

WSJ: Do you think there’s anything applicable in your investment strategy to an everyday investor?

Ballmer: I would say, ‘Keep it simple’—unless you’re really going to become an expert.

I always like to talk about [how] in business, you have what I’ll call zero-trick ponies, one-trick ponies and two-trick ponies. A zero-trick pony is a business that never gets to incredible scale. A one-trick pony is what you do when you get to scale one big idea—you milk it, you milk it, you milk it. Most tech companies are one-trick ponies. Cisco, great company, one-trick pony. Even Google, I might argue, is still a one-trick pony centered around advertising. Amazon, two-trick pony. Apple, two-trick pony. Microsoft, two-trick pony.

If you put that back in the investment context, you can’t necessarily believe that an investor who did one trick super well will necessarily do the next investment trick super well, at least by my mental model. I haven’t studied every investor in the world, and maybe there’s people who consistently are doing great. But as in operating businesses, finding multiple-trick ponies is really hard. Why should somebody who’s a casual investor believe they’re necessarily going to find the people who can do these amazing tricks?

WSJ: Do you think there’s any luck in how well you’ve done owning Microsoft stock and holding it?

Ballmer: If there’s anybody that will tell you all their success is because they were a genius, I’d say they should go look for the luck.

Look, we had luck in Microsoft’s success. Forget the stock price. I had luck, essentially, in getting to listen to the right people. But I also had luck in terms of my loyalty to the company and not wanting to be a seller as a leader of the business. It turned out to be a great investment thing, too.

FT : Palantir and Anduril join forces with tech groups to bid for Pentagon contr

Palantir and Anduril join forces with tech groups to bid for Pentagon contracts
Consortium likely to include Elon Musk’s SpaceX in move to grab a bigger slice of $850bn US defence budget

Palantir and Anduril, two of the largest US defence technology companies, are in talks with about a dozen competitors to form a consortium that will jointly bid for US government work in an effort to disrupt the country’s oligopoly of “prime” contractors.

The consortium is planning to announce as early as January that it has reached agreements with a number of tech groups. Companies in talks to join include Elon Musk’s SpaceX, ChatGPT maker OpenAI, autonomous shipbuilder Saronic, and artificial intelligence data group Scale AI, according to several people with knowledge of the matter.

“We are working together to provide a new generation of defence contractors,” said one person involved in developing the group.

The move comes as tech companies seek to grab a bigger slice of the US government’s huge $850bn defence budget from traditional prime contractors such as Lockheed Martin, Raytheon and Boeing.

The consortium will bring together the heft of some of Silicon Valley’s most valuable companies and will leverage their products to provide a more efficient way of supplying the US government with cutting-edge defence and weapons capabilities, according to a second person involved.

It comes as defence tech start-ups have attracted record amounts of funding this year, as investors bet they will be among the winners of higher federal spending on national security, immigration and space exploration under Donald Trump’s incoming government.

Wars in Ukraine and the Middle East and geopolitical tensions between the US and China have heightened the government’s reliance on tech companies developing advanced AI products that can be used for military purposes, and encouraged investors to the sector.

Palantir’s share price has skyrocketed by 300 per cent in the past year, giving the company a market capitalisation of $169bn — larger than Lockheed Martin. The data intelligence group was co-founded by tech investor Peter Thiel, who also provided the initial backing for Anduril, which launched in 2017 and was this year valued at $14bn.

Meanwhile, SpaceX was valued at $350bn this month, making it the world’s largest private start-up, and OpenAI has soared to a valuation of $157bn since it was founded in 2015.

Each of the companies has attempted to grab a slice of the government’s defence budget. While SpaceX and Palantir have won large public contracts going back two decades, some are newer to government procurement. OpenAI updated its terms of service this year to no longer explicitly prohibit the use of its AI tools for military purposes.

US defence procurement has long been criticised as slow and anti-competitive, favouring a small number of decades-old primes, such as Lockheed Martin, Raytheon and Boeing. These vast conglomerates typically produce ships, tanks and aircraft that are costly and take years to design and manufacture.

Silicon Valley’s burgeoning defence industry has prioritised producing smaller, cheaper, autonomous weapons that they claim will better protect the US and its allies in a modern conflict.

One person involved in developing the consortium described it as “aligning industry” in order to “execute the technical priorities of the Department of Defense” and “solve critical software capability problems”.

Some tie-ups between the tech groups expected to be in the consortium have already been agreed and integration work will begin immediately.

Palantir’s “AI Platform”, which delivers cloud-based data processing, was this month integrated with Anduril’s autonomous software, “Lattice”, to deliver AI for national security purposes.

Similarly, Anduril combined its counter-drone defence systems with OpenAI’s advanced AI models to jointly work on US government contracts related to “aerial threats”.

A joint statement from Anduril and OpenAI about that partnership said it “aims to ensure that the US Department of Defense and intelligence community have access to the most advanced, effective, and safe AI-driven technologies available in the world”.

Anduril, OpenAI and Scale AI declined to comment on the development of the consortium. Palantir, SpaceX and Saronic did not respond to requests for comment.

WSJ : Paramount’s Media Heiress Will Leave the Stage After Last Act in a Chaotic

Paramount’s Media Heiress Will Leave the Stage After Last Act in a Chaotic Drama
An $8 billion deal Shari Redstone struck with Skydance Media must go through the FCC next year; new owners seek over $2 billion in cost cuts

Paramount Global PARA 0.76%increase; green up pointing triangle boss Shari Redstone took a winding path to sell her family’s media empire. Along the way, a CEO was dethroned, multiple board members resigned and investors lashed out.

Now, the 70-year-old media heiress is preparing to hand the reins to a new owner in 2025—but she promises not to go away entirely, and neither will Paramount’s problems.

Redstone earlier this year struck an $8 billion deal with Hollywood production company Skydance Media that will reshape the entertainment landscape. She is selling National Amusements, the family company that owns movie theaters and through which she controls Paramount.

As part of the transaction, Skydance will merge with Paramount, gaining control of a collection of assets that include the iconic studio behind “The Godfather” and “Titanic,” CBS and cable networks like MTV.

Skydance Chief Executive Officer David Ellison will face a formidable task: Paramount shares are down about 26% this year as a growing Paramount+ streaming business hasn’t been enough to offset the accelerating decline of the company’s cable business, which took a $6 billion write-down in August. The landscape is shifting as more people abandon traditional TV; Comcast is spinning off nearly all its cable channels in 2025.

Ellison and his top lieutenant, former NBCUniversal CEO Jeff Shell, have explored how to catapult Paramount’s position in streaming. They are looking at integrating the company’s free, ad-supported Pluto service into Paramount+, people familiar with the situation said.

Paramount’s current management had explored streaming partnerships with a number of companies including Amazon, YouTube, Warner Bros. Discovery and Netflix, though nothing has come to fruition. One main sticking point with Warner has been which company would have control of any joint venture, according to people involved in the discussions. Ellison and Shell are seeking over $2 billion in cost cuts.

Redstone, daughter of the late media titan Sumner Redstone, was a reluctant seller, and shortly after the deal was signed, she was already tearfully telling associates that she had regrets about relinquishing her family’s business. She managed to secure some trappings of mogul life. Skydance and its investors agreed to take on National Amusements’ financial obligations and pay for the remainder of Redstone’s lease for her private jet, and will cover the expenses for her Central Park-area apartment in New York City for the next few years, according to people familiar with the situation.

Redstone’s relationship with Donald Trump may also come in handy as the next administration evaluates the Paramount-Skydance deal. The president-elect’s incoming nominee for Federal Communications Commission chairman, Brendan Carr, has said the agency could scrutinize whether CBS’s handling of a “60 Minutes” interview with Kamala Harris—which critics said was edited in misleading ways—violated standards that require broadcasters to act in the public interest. Trump filed a $10 billion lawsuit against CBS, which has denied wrongdoing.

Redstone has told people close to her that she understands Trump’s frustration with CBS, even though she doesn’t support the lawsuit. She has gotten along well with Trump for years. He was supportive in the court battles that helped her gain control of Paramount several years ago, people familiar with the matter said, and the two still talk on occasion.

Larry Ellison, the billionaire co-founder of Oracle and father of David Ellison, also has a longstanding relationship with Trump.

Dance partners
The final deal came after a will-they, won’t-they corporate drama for the ages. The negotiations, in which Skydance was code-named “Sparrow” and Paramount “Pluto,” stopped and restarted multiple times.

Exclusive talks with Skydance, whose credits include TV shows such as Amazon’s “Tom Clancy’s Jack Ryan” and movies like “Top Gun: Maverick,” started on April 3. Some Paramount investors balked at initial proposals that they said gave Redstone a sweeter deal than ordinary shareholders.

Meanwhile, Redstone and the Paramount board were losing faith in CEO Bob Bakish, who had voiced concerns about the Skydance deal and sought out alternatives. Some directors felt the long-range plans Bakish had presented to the board recently didn’t include enough cost-cutting and had overly optimistic growth projections, according to people familiar with the situation.

Certain board members approached George Cheeks, then the head of CBS, and two other top executives, studio chief Brian Robbins and cable networks chief Chris McCarthy, to get a sense of what the executives would do if they were running the company.

Later in April, Cheeks was getting off a plane in Las Vegas to see comedian Cedric the Entertainer perform when Redstone called to say Bakish was out. Meanwhile, merger talks with Skydance Media were stalling.

The Wall Street Journal first reported on Bakish’s fate on April 26. The company announced his exit days later, and named Cheeks, Robbins and McCarthy as co-CEOs.

‘The three amigos’
At that point, it seemed as if Redstone intended for Paramount to go it alone. Cheeks, Robbins and McCarthy’s initial plans called for the possibility of a streaming joint venture, cost cuts and a potential sale of Pluto, people close to the company said.

In the early going, staffers found the new arrangement inefficient and were confused about how to address the new leadership team, with the nickname “the three amigos” catching on. Officially, the trio was given the moniker of “Office of the CEO,” which was then changed to “Office of the co-CEOs.”

Skydance remained interested in Paramount, but it had competition. Apollo Global had sent a letter to Paramount expressing interest in buying the company for $26 billion. That offer looked compelling to many investors, compared with the math for Skydance, whose financials were revealed in a Journal report.

Eventually, Skydance sweetened its original offer. Under the new proposed terms, all nonvoting and voting shareholders would have an option to cash out at a premium. The talks accelerated, and the deal seemed to be nearing completion.

As the Paramount special board committee gathered on June 11 to consider the deal, Redstone was signaling to one board member that something was wrong, people familiar with the matter said. Minutes before the call, the special committee’s lawyer, Faiza Saeed, a partner with Cravath, Swaine & Moore, received an email from Redstone’s deal lawyer explaining that the parties had failed to find common ground on fundamental issues.

“As such we do not have an agreement on a deal with Skydance nor do we anticipate that we will find a path forward for this transaction,” the email said.

Redstone was still concerned about the legal risks from the deal as well as the financial haul for her family’s company, which she concluded was being watered down.

Almost immediately after the deal collapsed, Ellison wrote to Redstone and her son Tyler Korff, who had been working on the deal, acknowledging there were missteps on his side, according to people familiar with the exchange. The two sides quietly got back into negotiations, while the rest of the business world—including the Paramount board—thought the deal was dead.

Saeed, the attorney for Paramount’s special committee, received yet another unexpected email from Redstone’s side on July 2, this time to say that Redstone’s National Amusements had reached a deal with Skydance and it was now up to the board to finalize a merger of Paramount and Skydance.

‘I will never disappear’
To help pacify Paramount’s nonvoting investors, Skydance agreed to provide $4.5 billion that Paramount can use for an offer to buy out about 50% of nonvoting shares. Redstone got $1.75 billion for National Amusements.

Skydance’s Shell and Ellison are building out Paramount’s management team as they prepare to take control of the company. Former Netflix executive Cindy Holland is expected to oversee streaming, reporting to Ellison. Cheeks is expected to be head of TV, according to people familiar with the situation. Dana Goldberg, chief creative officer at Skydance, is expected to oversee the combined studio, and Robbins, one of the co-CEOs, is expected to depart. Bloomberg earlier reported some of the expected management changes.

Under the deal’s terms, Redstone and Korff could join the board of the combined company, but both have decided against it, according to people familiar with the situation.

Despite that, Redstone made it clear that even though she was selling the company, she wasn’t planning to go away completely.

“One thing I promised everybody in the company is I will never disappear in terms of being your advocate and in terms of helping you be who you need to be, who you want to be in this company,” Redstone said at Advertising Week New York in October, when addressing a controversy at CBS News.

TechCrunch : Hollywood angels: Here are the celebrities who are also star VCs

Hollywood angels: Here are the celebrities who are also star VCs

Becoming a venture capitalist has become the latest status symbol in Hollywood.

Everyone these days, from Olivia Wilde to Emma Watson, is either launching a venture firm or is in the process of backing the next hottest company. Robinhood, Uber, and Klarna all have some glitzy names on their cap tables, paving the way for the rise of “Hollywood angels.”

For founders, having celebs on the cap table can bring a lot of buzz to the startup — and bonus points if the celeb knows what they are doing, too. Here are some of the top celebs investing and what they typically look to invest in. (This list will be updated periodically).

Snoop Dogg
The famous rapper and personality is a prolific angel investor and is also a founder of Casa Verde Capital, a firm focused on cannabis investments. As an angel investor, according to PitchBook, he’s backed companies such as Cameo, Reddit, Klarna, Robinhood, and the fintech MoonPay, alongside other celebs on this list. He’s made at least 26 investments, according to PitchBook. The managing partner of Casa Verde Capital is Karan Wadhera and the firm closed a $94 million Fund II in 2020.

Sara and Erin Foster
The Foster sisters are two of Hollywood’s most famous siblings, best known these days as the producers of the Kristen Bell Netflix show “Nobody Wants This.” They launched a venture firm in 2022 looking to raise $20 million to back consumer companies, and the fund has officially closed, their reps confirmed to us. They know the consumer area well: They already have a podcast, a clothing line, and a hit television show.

PitchBook shows the firm, which they run alongside business partner Phil Schwarz, has written at least six checks, including into companies like the diaper brand Kudos (alongside Gwyneth Paltrow) and the luxury boxed wine company Juliet.

As angel investors, the sisters have also invested in the alkaline water company ZenWTR. Sara is also an angel investor in the swimwear company Summersalt and the ear-piercing startup Rowan.

Kevin Hart
A few years ago, comedian Kevin Hart launched a venture firm, HartBeat Ventures, alongside Robert Roman, an investor with a background in financial management for entertainment and sports celebrities.

The firm was founded around 2021 with the mission to support minority and underrepresented founders. It received its first outside investment from J.P. Morgan as part of one of the bank’s initiatives to fund more firms led by women and people of color.

The firm has invested in at least 14 companies, according to PitchBook, including the art marketplace Masterworks, the 3D printing company Ready Player Me, and Rihanna’s Savage x Fenty lingerie line. Tashi Nakanishi and Walid Samaha are partners at the firm, according to LinkedIn, while Ron Everline serves as vice president.

HartBeat Ventures began raising its Fund I in 2022 with no target amount. As of January 2024, it has raised at least $28 million, according to an SEC filing from January. An October SEC filing also shows that HartBeat is trying to raise capital for what seems to be an SPV looking to invest in Aston Martin’s Formula One car racing team.

Paris Hilton
The famed socialite is also an angel investor, according to PitchBook. She has made investments through her company 11:11 Media, of which she is the CEO and co-founder. Her investments have included the crypto company MoonPay, the biotech company Colossal Biosciences, the community event platform AfterParty, and the podcast platform Fireside Chats.

The theme for her seems to be social platforms, biotechnology, and SaaS products that make consumer life easier. She’s made at least 22 investments, according to PitchBook, and happens to be married to Carter Reum, founder of well-known Los Angeles firm M13 Ventures, which has invested in companies such as ClassPass, Canvas, and Pinterest, according to its website.

Ashton Kutcher
Perhaps one of the most famed and notable celebrity VCs is Ashton Kutcher. He is the co-founder of two VC firms, A-Grade and Sound Ventures, both of which he started with Guy Oseary. Shortly after launching Sound, Effie Epstein joined as managing partner.

Sound Ventures, founded in 2015, looks to back software companies and has more than $1 billion in assets under management, according to PitchBook. It also invests out of thematic funds, and last year launched an AI Growth Thematic fund, which invested in companies such as Hugging Face, OpenAI, Anthropic, and World Labs. He told the audience at TechCrunch’s 2024 Disrupt that he’s backing competitors in AI because he thinks there’s room for multiple massive AI model companies. His other notable investments between A-Grade and Sound include Affirm, Airbnb, Airtable, Duolingo, and Uber.

Nas
The famed rapper Nas founded his firm QueensBridge Venture Partners back in 2014. It’s made more than 130 investments across various industries and has around $140 million in assets under management, according to PitchBook. Anand Murthy is a partner at the firm. The firm has invested in companies such as the luggage brand Away (alongside Jay-Z) and the bedding company Parachute, according to PitchBook.

As an angel investor, Nas has made more than 50 investments, according to PitchBook. He’s backed SeatGeek, Robinhood, Ring, Coinbase, and the fintech Mercury.

Gwyneth Paltrow
The Goop founder has made at least 24 investments as an angel investor, according to PitchBook. She, alongside Moj Mahdara, is also the co-founder of Kinship Ventures, which invests in consumer and wellness companies. The firm was looking to raise a $75 million VC fund, Axios reported last year, and has backed the crypto payment company MoonPay. According to other SEC filings, it seems Kinship Ventures also raised what seems to be an SPV to back OpenAI and the bitcoin mining company TeraWulf.

As an angel investor, Paltrow has invested in the diaper company Kudos, the hair company Crown Affair, the beauty e-commerce marketplace Thirteen Lune, and the beverage brand Olipop, alongside actress Priyanka Chopra, according to PitchBook. Actress Mindy Kaling, rapper Logic, and the Jonas Brothers, also backed Olipop.

Emma Watson
The actress has cut at least two angel investor checks, according to PitchBook. Investments include at least two companies, including the biomanufacturing company FabricNano, and Hertility Health. Watson is a well-regarded activist, especially when it comes to the environment and women’s rights. It’s no wonder, then, that her two angel investments are in companies that seek to make product manufacturing more sustainable and a women’s health platform.

Olivia Wilde
Actress and director Olivia Wilde teamed up with Neil Sirni, who led Roc Nation’s venture division Arrive, to launch the venture firm Proximity, Bloomberg reported in October. Wilde does not have a PitchBook profile just yet but Bloomberg reports that her new firm will focus on consumer and enterprise technology and has already cut a few checks, including into Pendulum Therapeutics.

Investor Jason Mack and artist Santigold, known for her song “L.E.S. Artistes,” are also partners at the firm, according to Bloomberg. SEC filings show that Proximity is raising two new SPV funds, including a $5 million health fund and another $10 million fund.

Jay-Z
The music mogul is an angel investor and co-founder of RocNation and of what is now MarcyPen Capital Partners, an investment firm stemming from the merger of Jay-Z’s Marcy Venture Partners with Pendulum Holding’s investment arm Pendulum Opportunities.

Roc Nation has a venture arm called Arrive Opportunities Management, which was once managed by Neil Sirni, the investor now working with Olivia Wilde at Proximity. That venture arm has made at least 40 investments, according to PitchBook, backing companies such as Sweetgreen and Epic Games.

Before the merger, Marcy Venture Partners invested in several companies, such as Rihanna’s Fenty x Savage (making Rihanna one of the few Black women in the world to have a unicorn company) and the web3 company Spatial Labs. Charlie Hanna and Mason O’Hanlon are listed as investors at MarcyPen Capital Partners on PitchBook and LinkedIn.

Furthermore, as an angel investor, Jay-Z has made some notable investments — at least 27, according to PitchBook — into companies like Oatly, Flowhub, Impossible Foods, and the luggage company Away.

The Information : Why Netflix Will Lose in Sports

Why Netflix Will Lose in Sports

Netflix shareholders may be a happy bunch right now, after the bonanza year the video-streaming stock has enjoyed. Though its revenue is growing at a decent but unspectacular clip—15% in the third quarter—its shares have so far risen 87% this year to above $900 a share. Since their low point in mid-2022, during a freakout about Netflix’s subscriber stall, the stock has soared more than 440%. Given how badly every other streamer is doing, investors seem to regard Netflix as the only streaming stock worth buying.

Shareholders shouldn’t get too comfortable. Today’s news that Netflix has won the rights to exclusively stream the FIFA Women’s World Cup in 2027 and 2031 is a bearish sign for the stock. It’s further evidence—on top of next week’s NFL games streaming on Netflix—that the company has abandoned the admirable discipline that kept it out of live sports and is charging into that morass. That should worry anyone who owns shares in the company.

Buying the TV rights to live sports is a ticket to lose a boatload of money. Traditional TV networks feel they have little choice because it’s the only way left for them to draw lots of viewers and the ad dollars that follow. It doesn’t seem to matter to them that, given how much sports rights have risen in cost, the networks (aside from Disney’s ESPN) may not make money from sports programs.

Sure, the TV executives may not be wrong. Witness the impact of Warner Bros. Discovery’s loss of the NBA rights to companies willing to pay more: The future of its cable channels is in doubt.

But Netflix isn’t in that unenviable situation. It has a global lead in subscribers, giving it the resources to spend enough on programming to maintain its lead. It’s making a bundle of money—$5.5 billion in free cash flow in the first nine months of 2024, nearly three times as much as Warner Bros. Discovery, which has slightly more revenue.

True, Netflix isn’t growing fast anymore. To remedy that, it’s introduced an advertising tier, which in turn is why it is turning to sports. The flaw with this strategy is that all sports will do is pit Netflix against every other TV network and streaming service for both ad dollars and streaming rights. The only winners are the sports leagues, who rake in the money.

Bulls will argue that Netflix’s sports deals so far haven’t been exorbitant. That’s true—but its sports investment won’t end there. If it wants to really use sports to draw a significant amount of advertising dollars, it will have to spend a lot more money buying more rights. It’s a no-win situation that Netflix would be better off avoiding.

(ZeroHedge) Santa Claus Rally Or Did The Fed Steal Christmas?

Santa Claus Rally Or Did The Fed Steal Christmas?

Powell & A Government Shutdown Hits Stocks
Last week, we noted the ongoing market churn that could last into this week’s Fed meeting. To wit:
“That certainly seemed the case this past week, with the market trading being fairly sloppy. Attempts to push the market higher were repeatedly met with sellers, and we saw a rotation from over-owned to under-owned assets. Notably, that selling pressure arrived as expected, and while such could persist until early next week, we should be getting close to the end of the distribution and rebalancing process. The good news is that the recent consolidation paves the way for ‘Santa Claus to visit Broad and Wall.”
That process continued as expected this past week but became violent on Wednesday following the Federal Reserve meeting. While the Fed cut rates as expected, the market shock came from the lift in its outlook for interest rates in 2025 by a half percentage point. The market is assuming that the Fed is giving up on the idea that inflation will return to the 2% target next year, an idea that they had confidence in as recently as September. That more hawkish outlook undermined the view that elevated valuations were justified by easier monetary conditions, which now seems to be reversing. We suspect that this view is rather short-sighted, and given the economic dynamics both abroad and in the U.S., slower economic growth will lead to a “dovish” pivot by the Fed in the first half of 2025.
The markets also struggled with concerns about a Government shutdown. As we discussed in October 2023, shutdowns are NOT a threat to the market in the long term. To wit:
“What is critical to understand about Government shutdowns is that mandatory spending (social security, welfare, interest on the debt) continues as needed. Shutdowns are primarily about discretionary spending. Such is why it mainly involves Government employment and the shuttering of national parks and monuments. According to Goldman Sachs, the shutdown would have only impacted about 2% of Federal spending overall. Notice that the vast majority of Government spending is directly a function of the social welfare system and interest on the debt.”
Please note that during a Government shutdown, all MANDATORY spending continues. In other words, the government WILL NOT default on its debt, and social security payments will continue, despite rhetoric to the contrary.
Furthermore, market reactions to government shutdowns have become increasingly muted. The reason is that the markets have learned that funding typically arrives at the 11th hour via a ‘continuing resolution’ to provide temporary funding through the next political event, such as midterm elections, inauguration, etc. While these short-term spending bills eventually translate into longer-term spending bills, the real problem is that continuing resolutions (CRs) increase spending by 8% annually. Such is why debt has exploded since Congress stopped passing budgets in 2009 under President Obama and opted for CRs. The debt surge is the direct result of automatically compounding 8% annual spending increases plus additional spending.
However, as shown, government shutdowns, if they occur, can temporarily impact markets, but the event tends to be mild and short-lived.
Nonetheless, the market has triggered a short-term MACD sell signal, which warned investors that some “event” could exert downward pressure on stocks. As noted, the Fed and “Government Shutdown” drama sufficiently triggered sellers as portfolio rebalancing and distributions concluded. With relative strength oversold on Friday, the setup for a reflexive rally into year-end has become a much higher-probability event. However, the ongoing sell signal is deep enough to limit whatever reflexive rally does arrive. Such is particularly true as money flows have deteriorated over the last few weeks.
While we still expect a rally into year-end, as we will discuss, there is a not-so-insignificant possibility of further turmoil. We suggest continuing to manage risk, and with significant gains already booked for this year, there is little need to stretch for further returns at this juncture.
Will Santa Claus Visit Broad And Wall?
Will “Santa still visit Broad and Wall?” That is the question on everyone’s mind. As we will discuss, there are certainly reasons to be concerned, but let’s start with the market statistics and reasons behind the fabled year-end rally.
The actual Wall Street saying is, “If Santa Claus should fail to call, bears may come to Broad & Wall.” The Santa Claus Rally, also known as the December effect, is a term for more frequent than average stock market gains as the year winds down. However, as is always the case with data, average returns sometimes differ from reality.
Stock Trader’s Almanac explored why end-of-year trading has a directional tendency. The Santa Claus indicator is pretty simple. It looks at market performance over a seven-day trading period – the last five trading days of the current trading year and the first two trading days of the New Year. The stats are compelling.
“The stock market has risen 1.48% on average during the 7 trading days in question since both 1950 and 1969. Over the 7 trading days in question, stock prices have historically risen 76% of the time, which is far more than the average performance over a 7-day period.“
The end of the year tends to be strong for a couple of reasons. First, professional managers tend to “window dress” portfolios for year-end reporting purposes. Secondly, given that many professional funds make year-end distributions, there tends to be a need to rebalance portfolios. The following graph in orange shows aggregate cumulative returns by day count for the December months we analyzed. In the graph, we plotted returns alongside daily aggregated average returns by day. Unsurprisingly, the recent sloppy trading and correction this past week all coincide with the historical norms of December.
Visually, one notices the “sweet spot” in the two graphs between the 10th and 14th trading days. The 14th trading day, in most cases, falls within a few days of Christmas.
However, there is always a risk.
Did The Fed Steal Christmas?
While there is a decently high probability that stock prices will climb heading into year-end, there is a not-so-insignificant 24% chance they won’t. With the substantial November advance and new highs into early December, the question is whether anyone is “left to buy?” As noted, not every December has a “Santa Claus Rally.” 2018, as shown, is a good reminder that once in a while, investors receive a lump of coal in their stockings. At that time, the Federal Reserve was on a rate hiking campaign and insisted that it was “nowhere near the neutral rate” on monetary policy. Furthermore, since the market had declined steeply since early September, sentiment and investor positioning were very negative.
Interestingly, December 2024 has some of the same backdrops as September 2018.
First, the S&P 500 rallied strongly this year, approaching our year-end target of 6000. That rally has led to a sharp increase in bullish sentiment between retail and professional investors. As shown, U.S. equity allocations are at record highs among professional investors.
Furthermore, like in 2018, when retail equity allocations and valuations were elevated, investor allocations are at the highest on record, coinciding with the second-highest valuation levels.
There is also an abundance of optimism about future stock prices, just like in 2018.
What is important to remember about 2018 is that investor optimism was fine until the Fed said it “was nowhere near the neutral rate.” Of course, following a 20% decline and two months later, the Fed was magically at that neutral rate.
Today, investor exuberance is tied to a further accommodative easing in 2025. However, like in 2018, the Fed suggested it isn’t near its “neutral rate,” as shown in its latest projections. While the “long-run” projections are still for economic growth of 1.8% (down from 2.0% and 1.9% previously) and inflation of 2%, the short-term outlooks for 2025 were adjusted modestly higher. That uptick disappointed investors even though the end goals remain the same, which will require Fed funds to adjust lower.(Side note: The Fed’s projections are almost always too optimistic, which suggests the recent bout of hawkishness will give way to a dovish reversal next year.)
The adjustment to the Fed’s view was minimal from an investing perspective. However, the market reacted violently because the combination of exuberance and overbought factors created the perfect environment for a reversal.
Technically Speaking
First, while the market rallied into year-end on many optimistic assumptions, breadth has been deteriorating noticeably. From the NYSE Advance-Decline line to the percentage of stocks trading above their respective 50 and 200-DMA, overall participation has declined rapidly. While such does not mean a market crash is imminent, such previous deterioration has eventually coincided with short-term corrections and consolidations. Unsurprisingly, that is exactly what happened as the collision of the Fed and a looming shutdown gave sellers the push they needed.
Secondly, the market was, and is, technically extended on many levels after the past two years of excess returns. The monthly market analysis shows the S&P 500 is significantly overbought on a relative strength basis, deviated from the long-term mean, and pushing well into the top of its bullish trend from the 2009 lows. While we discussed the same factors in the middle of 2021, it took several months before the market gave way and corrected the excesses in 2022. Given the market’s current momentum, we suspect the bullish run will likely last into the first half of next year but could be sooner if earnings expectations decline.
What is crucial to understand is that these technical extremes are just the “kindling” for a correction. To “ignite” the correction, some event must provide the catalyst. In this case, it was the more hawkish pivot by the Fed and the threat of a shutdown. As is always the case, the event that causes a sharp unwinding of the market, like we saw on Wednesday, is always unexpected. The “surprise factor” causes the sudden shift in market expectations for earnings growth and outlooks. The risk going forward is “if” the Fed is correct in its outlook, the more optimistic outlook for earnings expectations will need to be reassessed. If that is the case, the market will decline to reduce valuations for a new reality.
Given that current valuations are at the second-highest level on record, such an event would seem more likely. Notably, short-term valuations are solely a function of sentiment. Investors are paying well above the earnings growth that is occurring. Historically, earnings have disappointed those expectations.
Does any of this mean that “Santa Won’t Visit Broad And Wall?” Of course not. However, I would not completely dismiss the risk of “getting a lump of coal” this year.
Given the uncertainty, both into year-end and 2025, how should we approach it?
Calculating The Madness
Let me repeat something that seems apropos:
Sir Isaac Newton once said:
“I can calculate the motions of the heavenly bodies, but not the madness of the people..”
As we head into year-end, we will navigate the risk of overly extended and bullish markets against the seasonally strong end-of-year period.
We believe that capital preservation and risk management lead to better outcomes over the long term. However, managing risk can be frustrating in the short run as the “Fear Of Missing Out” overrides common sense and logic.
If you disagree, that is okay.
When the opportunity presents itself and the “madness has subsided,” these are the questions we will ask ourselves before we add exposure to portfolios:
  1. What is the expected return from current valuation levels? (___%)
  2. If I am wrong, what is my potential downside, given my current risk exposure? (___%)
  3. What actions should I take now if #2 exceeds #1? (#2 – #1 = ___%)
How you answer those questions is entirely up to you.
What you do with the answers is also up to you.
We are all trying to answer the question, “How much of the ‘narrative’ already got priced into the market?”
By looking at the data, it would be easy to assume the answer is “much.”
While bullishness prevails, this is a great time to set aside the narratives and return our focus to the basic portfolio management rules.
How We Are Trading It
Since we have our “stockings hung by the chimney with care,” we can stuff them with a few essential investment guidelines to follow as we approach year-end.
  • Investing is not a competition. There are no prizes for winning but severe penalties for losing.
  • Emotions have no place in investing. You are generally better off doing the opposite of what you “feel” you should be doing.
  • The ONLY investments you can “buy and hold” provide an income stream with a return of principal function.
  • Market valuations (except at extremes) are very poor market timing devices.
  • Fundamentals and Economics drive long-term investment decisions – “Greed and Fear” drive short-term trading. Knowing what type of investor you are determines the basis of your strategy.
  • “Market timing” is impossible– managing risk exposure is logical and possible.
  • Investment is about discipline and patience. Lacking either one can be destructive to your investment goals.
  • There is no value in daily media commentary– turn off the television and save yourself the mental capital.
  • Investing is no different from gambling—both are “guesses” about future outcomes based on probabilities. The winner is the one who knows when to “fold” and when to go “all in.”
  • No investment strategy works all the time. The trick is knowing the difference between a bad investment strategy and one temporarily out of favor.
While anxiously anticipating the arrival of the “Santa Claus Rally,” we must also remember the lesson of 2018.
Nothing is guaranteed.

Mashable : Tesla recalls nearly 700,000 vehicles for a warning light issue

Tesla recalls nearly 700,000 vehicles for a warning light issue
That's a lot of cars.

Tesla recalled nearly 700,000 vehicles this week due to an issue with the tire pressure warning light. The problem could cause users to unknowingly drive on low-pressure tires.

"The tire pressure monitoring system (TPMS) warning light may not remain illuminated between drive cycles, failing to warn the driver of low tire pressure," wrote the National Highway Traffic Safety Administration (NHTSA) in a statement.

The recall affects certain 2024 Cybertrucks, 2017-2025 Model 3s, and 2020-2025 Model Ys. In good news for Tesla owners, the company said it released a free over-the-air software update that would remedy the problem. Letters notifying affected owners are scheduled to be mailed on Feb. 15.

A Reuters report, citing recall management firm BizzyCar, stated that Tesla accounted for 21 percent of all U.S. recalls in the first three quarters of 2024, though most issues were resolved with software updates.

The Cybertruck, especially, has been the subject of several high-profile recalls throughout the year.

WWD : Nike’s New CEO Admits the Brand Has Been Competing With Its Partners, Pled

Nike’s New CEO Admits the Brand Has Been Competing With Its Partners, Pledges to Fix Relationships With Foot Locker, JD Sports and More

In his first earnings call as Nike‘s chief executive officer, Elliott Hill pledged his “unwavering commitment” to the company’s retail partners.

Hill, who succeeded John Donahoe as CEO in October, said in a call with analysts on Thursday that re-earning the trust of the company’s wholesale partners is a top priority to help realign the business as a whole. Nike’s decision to exit several wholesale doors in 2021 has kept the company behind competitors that have achieved controlled distribution in crucial channels like run specialty. Now Hill must build on recent efforts to bolster wholesale sales and reengage key partners.

Hill acknowledged that the strong focus on Nike’s digital sales over the last few years has led the brand to compete with its own retail partners for consumer demand.

“Some partners and channels feel we’ve turned our back on them and we’ve stopped engaging consistently,” Hill said, adding that he has met with many of these key retailers directly in recent weeks. He called executives by name from Dick’s Sporting Goods, Foot Locker, JD Sports, Sports Direct and more chains, many of whom have offered positive commentary about Nike’s turnaround in recent weeks.

“They’re all encouraged by our commitment to delivering new, innovative product, telling emotional and inspiring stories and elevating Nike Direct,” Hill said. “We know our sales teams will have to earn every open-to-buy dollar, but we’re investing to make sure our partners feel supported. We’ll give them access to our best products and the breadth and depth they need, educate their teams on the latest Nike innovation and provide them with the marketing support both in store and out of home.”

“We will win when our partners win,” Hill added.

In line with this goal, Nike recently rehired and elevated Tom Peddie to the role of vice president, general manager of North America to oversee wholesale.

In addition to fixing wholesale, Hill’s turnaround strategy hinges on a renewed “obsession with sport,” which includes a focus on product innovation, demand creation and less centralization of teams. While he noted some changes might impact financial results in the short term, Hill said he is “making the decisions that are best for the health of our brand and business.”

In the second quarter, Nike reported top- and bottom-line results that beat the expectations of analysts.

WSJ : The Next Great Leap in AI Is Behind Schedule and Crazy Expensive

The Next Great Leap in AI Is Behind Schedule and Crazy Expensive
OpenAI has run into problem after problem on its new artificial-intelligence project, code-named Orion

OpenAI’s new artificial-intelligence project is behind schedule and running up huge bills. It isn’t clear when—or if—it’ll work. There may not be enough data in the world to make it smart enough.

The project, officially called GPT-5 and code-named Orion, has been in the works for more than 18 months and is intended to be a major advancement in the technology that powers ChatGPT. OpenAI’s closest partner and largest investor, Microsoft, had expected to see the new model around mid-2024, say people with knowledge of the matter.

OpenAI has conducted at least two large training runs, each of which entails months of crunching huge amounts of data, with the goal of making Orion smarter. Each time, new problems arose and the software fell short of the results researchers were hoping for, people close to the project say.

At best, they say, Orion performs better than OpenAI’s current offerings, but hasn’t advanced enough to justify the enormous cost of keeping the new model running. A six-month training run can cost around half a billion dollars in computing costs alone, based on public and private estimates of various aspects of the training.

OpenAI and its brash chief executive, Sam Altman, sent shock waves through Silicon Valley with ChatGPT’s launch two years ago. AI promised to continually exhibit dramatic improvements and permeate nearly all aspects of our lives. Tech giants could spend $1 trillion on AI projects in the coming years, analysts predict.

The weight of those expectations falls mostly on OpenAI, the company at ground zero of the AI boom.

The $157 billion valuation investors gave OpenAI in October is premised in large part on Altman’s prediction that GPT-5 will represent a “significant leap forward” in all kinds of subjects and tasks.

GPT-5 is supposed to unlock new scientific discoveries as well as accomplish routine human tasks like booking appointments or flights. Researchers hope it will make fewer mistakes than today’s AI, or at least acknowledge doubt—something of a challenge for the current models, which can produce errors with apparent confidence, known as hallucinations.

AI chatbots run on underlying technology known as a large language model, or LLM. Consumers, businesses and governments already rely on them for everything from writing computer code to spiffing up marketing copy and planning parties. OpenAI’s is called GPT-4, the fourth LLM the company has developed since its 2015 founding.

While GPT-4 acted like a smart high-schooler, the eventual GPT-5 would effectively have a Ph.D. in some tasks, a former OpenAI executive said. Earlier this year, Altman told students in a talk at Stanford University that OpenAI could say with “a high degree of scientific certainty” that GPT-5 would be much smarter than the current model.

There are no set criteria for determining when a model has become smart enough to be designated GPT-5. OpenAI can test its LLMs in areas like math and coding. It’s up to company executives to decide whether the model is smart enough to be called GPT-5 based in large part on gut feelings or, as many technologists say, “vibes.”

So far, the vibes are off.

OpenAI and Microsoft declined to comment for this article. In November, Altman said the startup wouldn’t release anything called GPT-5 in 2024.

Training day
From the moment GPT-4 came out in March 2023, OpenAI has been working on GPT-5.

Longtime AI researchers say developing systems like LLMs is as much art as science. The most respected AI scientists in the world are celebrated for their intuition about how to get better results.

Models are tested during training runs, a sustained period when the model can be fed trillions of word fragments known as tokens. A large training run can take several months in a data center with tens of thousands of expensive and coveted computer chips, typically from Nvidia.

During a training run, researchers hunch over their computers for several weeks or even months, and try to feed much of the world’s knowledge into an AI system using some of the most expensive hardware in far-flung data centers.

Altman has said training GPT-4 cost more than $100 million. Future AI models are expected to push past $1 billion. A failed training run is like a space rocket exploding in the sky shortly after launch.

Researchers try to minimize the odds of such a failure by conducting their experiments on a smaller scale—doing a trial run before the real thing.

From the start, there were problems with plans for GPT-5.

In mid-2023, OpenAI started a training run that doubled as a test for a proposed new design for Orion. But the process was sluggish, signaling that a larger training run would likely take an incredibly long time, which would in turn make it outrageously expensive. And the results of the project, dubbed Arrakis, indicated that creating GPT-5 wouldn’t go as smoothly as hoped.

OpenAI researchers decided to make some technical tweaks to strengthen Orion. They also concluded they needed more diverse, high-quality data. The public internet didn’t have enough, they felt.

Generally, AI models become more capable the more data they gobble up. For LLMs, that data is primarily from books, academic publications and other well-respected sources. This material helps LLMs express themselves more clearly and handle a wide range of tasks.

For its prior models, OpenAI used data scraped from the internet: news articles, social-media posts and scientific papers.

To make Orion smarter, OpenAI needs to make it larger. That means it needs even more data, but there isn’t enough.

“It gets really expensive and it becomes hard to find more equivalently high-quality data,” said Ari Morcos, CEO of DatologyAI, a startup that builds tools to improve data selection. Morcos is building models with less—but much better—data, an approach he argues will make today’s AI systems more capable than the strategy embraced by all top AI firms like OpenAI.

OpenAI’s solution was to create data from scratch.

It is hiring people to write fresh software code or solve math problems for Orion to learn from. The workers, some of whom are software engineers and mathematicians, also share explanations for their work with Orion.

Many researchers think code, the language of software, can help LLMs work through problems they haven’t already seen.

Having people explain their thinking deepens the value of the newly created data. It’s more language for the LLM to absorb; it’s also a map for how the model might solve similar problems in the future.

“We’re transferring human intelligence from human minds into machine minds,” said Jonathan Siddharth, CEO and co-founder of Turing, an AI-infrastructure company that works with OpenAI, Meta and others.

In AI training, Turing executives said, a software engineer might be prompted to write a program that efficiently solves a complex logic problem. A mathematician might have to calculate the maximum height of a pyramid constructed out of one million basketballs. The answers—and, more important, how to reach them—are then incorporated into the AI training materials.

OpenAI has worked with experts in subjects like theoretical physics, to explain how they would approach some of the toughest problems in their field. This can also help Orion get smarter.

The process is painfully slow. GPT-4 was trained on an estimated 13 trillion tokens. A thousand people writing 5,000 words a day would take months to produce a billion tokens.

OpenAI also started developing what is called synthetic data, or data created by AI, to help train Orion. The feedback loop of AI creating data for AI can often cause malfunctions or result in nonsensical answers, research has shown.

Scientists at OpenAI think they can avoid those problems by using data generated by another of its AI models, called o1, people familiar with the matter said.

OpenAI’s already-difficult task has been complicated by internal turmoil and near-constant attempts by rivals to poach its top researchers, sometimes by offering them millions of dollars.

Last year, Altman was abruptly fired by OpenAI’s board of directors, and some researchers wondered if the company would continue. Altman was quickly reinstated as CEO and set out to overhaul OpenAI’s governance structure.

More than two dozen key executives, researchers and longtime employees have left OpenAI this year, including co-founder and Chief Scientist Ilya Sutskever and Chief Technology Officer Mira Murati. This past Thursday, Alec Radford, a widely admired researcher who served as lead author on several of OpenAI’s scientific papers, announced his departure after about eight years at the company.

Reboot
By early 2024, executives were starting to feel the pressure. GPT-4 was already a year old and rivals were starting to catch up. A new LLM from Anthropic was rated by many in the industry as better than GPT-4. Several months later, Google launched the most viral new AI application of the year, called NotebookLM.

As Orion stalled, OpenAI started developing other projects and applications. They included slimmed-down versions of GPT-4 and Sora, a product that can produce AI-generated video.

That led to fighting over limited computing resources between teams working on new products and Orion researchers, according to people familiar with the matter.

Competition among AI labs has grown so fierce that major tech companies publish fewer papers about recent findings or breakthroughs than is typical in science. As money flooded the market two years ago, tech companies started viewing the results of this research as trade secrets that needed guarding. Some researchers take this so seriously they won’t work on planes, coffee shops or anyplace where someone could peer over their shoulder and catch a glimpse of their work.

That secretive attitude has frustrated many longtime AI researchers, including Yann LeCun, chief AI scientist at Meta. LeCun said work from OpenAI and Anthropic should no longer be viewed as research, but as “advanced product development.”

“If you’re doing it on a commercial clock, it’s not called research,” said LeCun on the sidelines of a recent AI conference, where OpenAI had a minimal presence. “If you’re doing it in secret, it’s not called research.”

In early 2024, OpenAI prepared to give Orion another try, this time armed with better data. Researchers launched a couple of smaller-scale training runs over the first few months of the year to build up confidence.

By May, OpenAI’s researchers decided they were ready to attempt another large-scale training run for Orion, which they expected to last through November.

Once the training began, researchers discovered a problem in the data: It wasn’t as diversified as they had thought, potentially limiting how much Orion would learn.

The problem hadn’t been visible in smaller-scale efforts and only became apparent after the large training run had already started. OpenAI had spent too much time and money to start over.

Instead, researchers scrambled to find a wider range of data to feed the model during the training process. It isn’t clear if this strategy proved fruitful.

Orion’s problems signaled to some at OpenAI that the more-is-more strategy, which had driven much of its earlier success, was running out of steam.

OpenAI isn’t the only company worrying that progress has hit a wall. Across the industry, a debate is raging over whether improvement in AIs is starting to plateau.

Sutskever, who recently co-founded a new AI firm called Safe Superintelligence or SSI, declared at a recent AI conference that the age of maximum data is over. “Data is not growing because we have but one internet,” he told a crowd of researchers, policy experts and scientists. “You can even go as far as to say that data is the fossil fuel of AI.”

And that fuel was starting to run out.

Reasoning
Their struggles on Orion led OpenAI researchers to a new approach to making an LLM smarter: reasoning. Spending a long time “thinking” could allow LLMs to solve difficult problems they haven’t been trained on, researchers say.

Behind the scenes, OpenAI’s o1 offers several responses to each question and analyzes them to find the best one. It can perform more complex tasks, like writing a business plan or creating a crossword puzzle, while explaining its reasoning—which helps the model learn a little bit from each answer.

Researchers at Apple recently released a paper that argues reasoning models, including versions of o1, were most likely mimicking the data they saw in training rather than actually solving new problems.

The Apple researchers said they found “catastrophic performance drops” if questions were changed to include irrelevant details—like tweaking a math problem about kiwis to note that some of the fruits were smaller than others.

In September, OpenAI launched a preview of its o1 reasoning model and released the full version of o1 earlier this month.

All that added brainpower is expensive. OpenAI is now paying to generate multiple answers to a single query, instead of just one.

In a recent TED talk, one of OpenAI’s senior research scientists played up the advantages of reasoning.

“It turned out that having the bot think for just 20 seconds in a hand of poker got the same boost in performance as scaling up the model by 100,000x and training for 100,000 times longer,” said Noam Brown, the OpenAI scientist.

A more advanced and efficient reasoning model could form the underpinnings of Orion. OpenAI researchers are pursuing that approach and hoping to combine it with the old method of more data, some of which could come from OpenAI’s other AI models. OpenAI could then refine the results with material generated by people.

On Friday, Altman announced plans for a new reasoning model smarter than anything the company has released before. He didn’t say anything about when, or whether, a model worthy of being called GPT-5 is coming.

WSJ : Nissan Needs a Honda Rescue. What Went So Wrong?

Nissan Needs a Honda Rescue. What Went So Wrong?
Japanese carmaker, still reeling from Ghosn affair, looks to merger to help it with aging model lineup and restless U.S. dealers

Nissan 7201 -0.40%decrease; red down pointing triangle, once a symbol of Japan’s carmaking prowess, spent the past five years trying to regain its footing after the arrest of longtime leader Carlos Ghosn, only to find itself in a precarious spot again.

The carmaker has been slashing jobs, cutting vehicle production and reporting lower profit. Nissan has been slower than rivals in refreshing its lineup, and it has fallen behind in the electric-vehicle race it once led.

Now, it is looking to Honda 7267 0.78%increase; green up pointing triangle for a lifeline.

The two carmakers said they were in talks over a merger that would be poised to create the world’s third-largest automaker by sales. The idea of two of Japan’s biggest rival brands cohabiting under the same roof would once have been considered absurd, but the companies have grown closer as Nissan weakens, announcing this year plans to share costs and work together on electric vehicles.

Nissan shares rose more than 30% in the two trading days after the announcement about merger talks, while Honda shares declined almost 5% in the same period.

The combination could help fortify the carmakers against global threats including a wave of new competitors coming from China that have taken a lead in the electric-vehicle race.

For Honda, a merger with Nissan offers the promise of sharing the high cost of developing new technologies. The Japanese government, which is worried about the auto industry’s competitiveness versus China and is subsidizing technology research, has suggested it would welcome the combination.

Nissan is also the largest shareholder in Mitsubishi Motors and collaborates with it on technology, a relationship that would likely carry over into a merged Honda-Nissan company. Together, Honda, Nissan and Mitsubishi sell more than eight million vehicles annually.

But Honda and Nissan have hurdles to overcome before they can make their deal final. The two companies have different cultures. They sell the same types of sport-utility vehicles and sedans for the mass market in the U.S. and elsewhere.

Nissan, which has a 25-year-old alliance with French carmaker Renault, has yet to fully recover from the arrest of Ghosn, its then-chairman, in late 2018. Stripped of his position and charged with financial crimes that he denied, Ghosn fled Japan a year later hidden inside a box on a private jet.

After his abrupt dethroning, Nissan reversed Ghosn’s growth plans, announcing mass layoffs and factory closures in 2019 and cycling through top executives and managers. The company’s global sales tumbled to 3.4 million vehicles last year from more than 5.6 million in 2018.

Renault and Nissan resolved their long-running tensions last year by agreeing that Nissan would regain its independence. Renault is gradually reducing its stake in Nissan to 15% from around 43%.

Freedom wasn’t as liberating as Nissan hoped. The company has struggled to fund its current business of selling mostly gasoline-powered vehicles such as the Rogue SUV and the Altima sedan while also investing in next-generation technologies. Its latest cost-cutting plan, released in November by CEO Makoto Uchida, calls for billions of dollars in savings and 9,000 job reductions.

Pooling resources could help Honda and Nissan, No. 2 and No. 3 in global sales among Japan automakers, approach the scale of No. 1 Toyota—which spends around $7.5 billion a year on research and development, and sells similar vehicles in many of the same markets.

Nissan said it was maintaining strong investments in R&D despite cost-cutting efforts. “We will also continue to leverage smart partnerships” to get the benefits of scale, said a Nissan spokesman.

Any partnership may need the blessing of Renault, but analysts say the deal is largely positive for the French carmaker because a struggling Nissan has been a weight on earnings.

The biggest fix-it job for a combined company would be in the U.S., where Nissan’s sales plummeted by one-third between 2019 and 2023.

Dealers say Nissan historically has been slow to refresh its lineup, and they have had to give steep discounts to unload dated-looking vehicles. That reputation has eroded the appeal of Nissan, and today buyers of the company’s vehicles tend to be people with poorer credit histories or those looking for a deal, dealers say. As a result, Nissan’s customer base has been harder hit by rising interest rates that have pushed up monthly payments.

Nissan’s replacement rate for its models—an indicator of how new its lineup is—has dropped to among the worst in the U.S. auto industry while Honda is above average, according to Bank of America’s annual “Car Wars” report.

Nissan says its sales decline over the past five years was the result of eliminating several models, such as the Maxima sedan, as well as selling fewer vehicles to rental-car companies. A company spokesman said Nissan wanted to roll out models more quickly and refresh them more often.

“Honda is a better-run company than Nissan, and I’m keeping my fingers crossed that it will rub off on Nissan,” said Adam Lee, chairman of Lee Auto Malls in Maine. His dealership group has two Nissan dealerships and one Honda store.

A survey of 600 auto retailers by Kerrigan Advisors, a buy-sell firm for dealerships, found that more than half of the respondents had no trust in Nissan. Founder Erin Kerrigan said many Nissan dealers were losing money and, if they wanted to sell their dealerships, couldn’t easily find a buyer.

On EVs, Nissan has gone from early mover to laggard. Its Nissan Leaf was one of the first fully battery-powered cars to hit the market in 2010, but Leaf sales have never taken off. Executives, wary of high development costs, have moved more slowly than rivals to expand Nissan’s EV lineup and production.

At a time when hybrid sales are taking off, Nissan doesn’t have one to sell in the U.S. It plans to sell a plug-in hybrid version of its Rogue in 2026. Honda, meanwhile, offers several hybrid models and a new all-electric SUV called the Prologue.

A merged Honda-Nissan would have to meld two distinctive cultures. Honda CEOs, starting with founder Soichiro Honda, have usually been engineers. Nissan historically favored graduates of the prestigious University of Tokyo, and its top jobs often went to sales leaders.

The history of the auto industry is littered with mergers and partnerships that didn’t work out.

Stephanie Brinley, an analyst at S&P Global Mobility, said it takes a long time for merged companies to cut costs and get more productive by combining manufacturing and product development—but Nissan needs help right now for its flailing U.S. business.

“A merger does not address the competitive issues holding Nissan’s sales back,” she said.