Category: Uncategorized

Elon Musk said he’s backing away from DOGE in May, but that doesn’t mean the federal worker firings are over.

In fact, they’re only heating up.

While the first era of DOGE firings continues to face legal issues, the next set could be on stronger footing. That’s because agencies have the chance to craft more methodical plans. In particular, many are offering buyout-like deferred resignation plans for workers to voluntarily quit in exchange for months of paid administrative leave.

These methods could prove to be on a more solid legal footing than the first round of firings, which focused on new or newly promoted workers, cited low performance ratings, and did not provide notice. It all means that the DOGE ethos is alive and well in the federal government, with or without Musk.

How round 2 of firings is different

Musk and DOGE spent much of the last three months overseeing a wave of probationary federal workers. Those workers were at the start of their tenure in their new roles, including internal promotions. They have fewer protections than longer-serving federal employees.

That round of cuts saw several snafus because, rather than targeting specific programs, DOGE simply removed many people who had been working for less than a year or two. That led to scenarios like workers who handle the nuclear stockpile or study bird flu getting fired and then ultimately rehired. It also faced legal issues, with a judge ordering in March that those workers needed to be reinstated.

Going back to the drawing board with terminations means a chance to execute reduction-in-force plans, or RIFs, that follow proper procedures.

“I think probably the courts have done a huge favor to DOGE by putting people back in their office until they can do a more well-calculated RIF,” Michele Evermore, a senior fellow at the left-leaning think tank the Century Foundation, said.

The White House and DOGE office did not immediately respond to a request for comment from BI.

Over the past month, new RIFs have gone out. The Department of Education announced in March that it was terminating over 1,300 workers, slashing its workforce by 50%. The Small Business Administration said it would reduce its workforce by 43%.

The Department of Health and Human Services had already started terminating employees as part of its plan to slash 10,000 from its workforce.

While the next round of reductions appears to be more targeted, it doesn’t mean they’re immune to litigation.

Federal worker unions have vowed to fight back. Doreen Greenwald, president of the National Treasury Employees Union, said in a statement that the union would “pursue every legal avenue to stop this unprecedented attack on the very foundation of our national government.”

It’s unclear if they’ll have a case. David Super, an administrative law professor at Georgetown Law, said that the plans may vary by department, and some could be more vulnerable to legal challenges than others.

He added that it’s “entirely possible that those agencies work really, really carefully and produce correct RIF plans.”

Federal workers are getting the option to quit before they’re fired

In lieu of RIFs, some agencies are choosing another acronym: DRPs. Deferred resignation programs are the next round of the “Fork In The Road” that offered employees early resignation earlier in the year, and might be an easier solution for those aiming to trim head count because workers voluntarily opt into them. Bloomberg reported, for instance, that nearly 20% of Department of Labor workers have opted for voluntary separation. Some Social Security Administration employees have also received a DRP.

Two internal emails at the Internal Revenue Service, viewed by BI, said that over 10,000 agency employees applied for the second round of the deferred resignation program and that further reductions in force are coming.

While DRPs might help the government shed workers, the payout could end up being less than the severance provided in a RIF, said Alan Lescht, an employment attorney specializing in federal workers.

“They’re trotting this out to people who are being targeted for RIFs or could potentially be in a RIF. And for federal employees that are like that, the RIFs may actually provide a larger payout,” Lescht said. And, opting into a DRP means workers are giving up rights to their positions — essentially, forfeiting the ability to be reinstated. He suggests workers take a “deep dive” into their job prospects before opting out.

“For many people, it may turn out that the best option is to stay and fight because the government has to satisfy a lot of requirements in order to justify a firing under the RIF rules,” Lescht said. “The RIF rules are very extensive and very complicated, and it’s very likely that with the speed with which the government is moving, that they will make mistakes.”

Have a tip? Contact these reporters via Signal at asheffey.97 and julianakaplan.33 or via email at asheffey@businessinsider.com and jkaplan@businessinsider.com. Use a personal email address and a nonwork device; here’s our guide to sharing information securely.

Stay informed with free updates

Simply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.

Advertising groups and tech start-ups have been racing to find ways to help brands boost their likelihood of surfacing in results from artificial intelligence chatbots, marking a new era of “search engine optimisation”.

Companies such as Profound and Brandtech have developed software for monitoring how frequently brands were surfaced by AI-powered services such as OpenAI’s ChatGPT, Anthropic’s Claude and Google’s Overviews feature.

Brands such as fintech company Ramp, jobs search site Indeed and Pernod Ricard-owned Scottish whisky maker Chivas Brothers have adopted the software. They are hoping to reach millions of users who regularly use generative AI products as a new method to search for information online — a shift that poses a long-term threat for Google’s main business.

“This is about much more than just getting your website indexed in their results. This is about recognising large language models as the ultimate influencer,” said Jack Smyth, partner at marketing technology group Brandtech, which has created its own interface for brands. 

These new tools are able to predict an AI model’s sentiment towards companies by feeding a slew of text prompts to chatbots and analysing the results. The technology is then used to create a ranking of brands, allowing agencies to advise on how best to ensure they are mentioned by the models.

The moves come as advertisers face pressure from the rising use of AI to create and target their marketing. Meta and Google have been developing self-serve tools for running ad campaigns directly to brands, in a potential threat to the work of agencies and media buyers. 

Some agencies are spotting the opportunity to offer new services to brands as AI becomes more prevalent and so-called search engine optimisation becomes less relevant.

Research from consultancy Bain found that 80 per cent of consumers now rely on AI-written results for at least 40 per cent of their searches, reducing organic web traffic by up to 25 per cent. About 60 per cent of searches now end without the users clicking through to another website, its research found.

However, on Thursday, Google’s parent company Alphabet announced its core search and advertising business grew almost 10 per cent to $50.7bn in the first quarter of the year.

The strong results provided reassurance to investors concerned about the growing popularity rival AI chatbots such as Elon Musk’s Grok, while also being on alert for evidence that answers from Google’s own Gemini chatbot and AI summaries are cannibalising its search business by reducing the number of user clicks on ads.

Still, agencies are racing to help corporate clients trying to appear within the results generated by AI services.

Brandtech has created a ‘Share of Model’ product that charges brands to see similar analysis and offers guidance on adjusting website text and image assets to better serve AI search.

Profound, which raised $3.5mn in seed funding in August led by Khosla Ventures, offers a data analytics platform which allows brands to track common queries related to their industry and understand their performance in AI searches.

“Traditional search has been one of the biggest monopolies in the history of the internet,” said James Cadwallader, co-founder of Profound. “And for the first time, it feels like the castle walls are cracking. This is a CDs to streaming moment”.

The software requires an understanding of how the individual models surface brands. ChatGPT, for example, uses a traditional web search and then evaluates the different sources for what information is most relevant for the user, including assessing the credibility and authority of the website.

Adam Fry, OpenAI’s ChatGPT search lead, said users are being more nuanced and precise in the questions they are asking, such as “can you find a quiet restaurant for a family of five in New York”, instead of “restaurants in New York”.

“The really new thing here is you have a layer of ChatGPT’s model, a layer of intelligence above traditional search,” said Fry.

Meanwhile, Perplexity, an AI-driven search engine, is currently piloting sponsored “questions” as a suggested follow-up after a user query.

“LLMs understand more content and can be more nuanced. They can find contradictions or find if information is misleading . . . so it’s a much more thorough process than reviewing links,” said Denis Yarats, co-founder of Perplexity. “It is much harder to be a target of SEO because the only sort of true strategy is to be as relevant as possible and provide good content.”

Data visualisation by Janina Conboye

David Sacks, the White House crypto czar and close friend of Elon Musk, said on the latest All-In podcast that Musk’s decision to spend less time at the White House DOGE office is indicative of his management style at his companies.

On the Friday episode, Sacks said that Musk is simply running DOGE like he typically runs his other companies, recalling the Twitter acquisition. Sacks was invited to help Musk purchase the social media platform and was a trusted confidant during the transition.

“I saw this before when I was part of the Twitter transition — is that for the first three months or so he was basically full time at Twitter HQ, learning the business down to the database level. I mean, every nook and cranny of that business, he learned about,” Sacks said. “Once he felt like he had a mental model and he had the people in place that he trusted, he could move to more of a maintenance mode.”

Musk said on Tuesday’s Tesla earnings call that he’ll scale back his time at DOGE and focus more on his EV company. The CEO said he’ll spend a day or two a week on his government work “as long as it is useful.”

The decision comes after investors and Wall Street analysts pleaded for the chief executive to re-align his priorities back to his company, as Tesla experienced sluggish sales and backlash from some of its core customer base.

Sacks said Musk’s method is how he can simultaneously manage multiple companies, and the CEO can now step back from DOGE.

Musk “has these intense bursts where he focuses on something, gets the right people and structure in place, feels like he understands it, and then he can delegate more,” he said. “And I think that he has reached that point with DOGE.”

The venture capitalist said Musk isn’t completely stepping down from his government role but is instead rationing his time in the White House. Musk is limited to 130 days per year as a designated special government employee.

“My sense is that DOGE is going to continue, it’s just that Elon is shifting to a mode where he can manage it one day a week or two days a week as opposed to being there five days a week,” he said.

While Sacks is not known to be a member of the DOGE team, the venture capitalist has had a yearslong personal and business relationship with Musk that can be traced back to their executive roles at PayPal. The two are part of what later became known as the PayPal Mafia, which also includes Peter Thiel and Reid Hoffman.

Musk and Sacks did not respond to requests for comment.

AI agents have been a hot topic in Web3 circles for some time now, sparking one of the most ambitious narratives in crypto: the dream of autonomous, intelligent entities managing capital, risk, and strategy across decentralised protocols. These systems, it was said, would not only outperform humans at execution, but also liberate users from the constant monitoring and micromanagement of their digital assets.

At the height of this excitement, bold predictions began shaping this narrative; “within 1 year, the majority of all DeFi TVL will be managed by AI Agents”.

But, as time passes, the buzz of this topic is giving way to reality, especially if the current most popular AI agents are X profiles with a token. That’s not the grand vision for these agents. With their infrastructure still in its nascent or development stage, the concept of DeFi-native agents will remain abstract for only so long. The AI economy now sits in a holding pattern—waiting for the build to catch up to the narrative.

The Reality Of AI Agents

While the term “AI agent” has taken many forms, embedding these agents into blockchain environments seems to carry a particular charge—amplifying expectations. Over the past six months, it has become clear that the combination of AI Agents and Web3 has astronomical potential, but only if the sector can move past early speculation and build for long-term value for end-users, a vision shared by many.

Interestingly, early optimism was expressed by projects like Fetch in late 2023, who wrote:

“While their potential in various application domains is still being explored, agent-based systems truly represent an immense opportunity for both companies and people”.

It sparked the initial wave, but real use cases remained limited at that time. However, it demonstrates that this is far from a passing trend. It’s a subject of ongoing extensive research, showing that AI agents can reconstruct how value is created and distributed across decentralized systems.

AI Agent Variety

A handful of foundational projects—Giza, Axal, and Theoriq to name a few—are architecting the primitives for agent-dedicated infrastructure in DeFi, each with a distinct approach.

Giza is advancing verifiable on-chain inference through zero-knowledge machine learning, enabling agents to act with cryptographic accountability. Axal prioritises execution integrity, developing systems for runtime verification and constraint enforcement. Theoriq, by contrast, explores decentralized intelligence through AI swarms—simulated collectives of agents coordinating within shared environments. This goes to show how multidimensional this space has become.

Interestingly, this also addresses a growing issue in DeFi: the fragmentation of AI agents. Handling token swaps, yield strategies, or cross-chain bridging often operates in isolation, with little to no coordination between them. The result is a disjointed user experience that’s difficult to navigate and scale.

This fragmented environment creates inefficiencies and friction, particularly for users who juggle multiple platforms and blockchains. The proposed solution—Agentic DeFi—calls for intelligent agent swarms that can collaborate across tasks, chains, and user intents to deliver a unified experience.

Theoriq’s model gestures toward this future. By exploring AI swarms, which are simulated collectives of agents that share data and goals, we can establish a core architecture for agent ecosystems that don’t just act independently, but operate as synchronized systems.

Although ambitious, these initiatives are still in their early stages. Very few are operating at high thresholds, but we can already spot a product-market fit, with accomplishments, such as Giza, being one example.

Notably, each agent framework is solving a different layer of the same problem. This reflects a maturing space, where builders are no longer racing to replicate, but instead developing complementary solutions. All of these pieces must ultimately fit together to form a cohesive future.

Is Intelligence The Bottleneck?

There’s a growing consensus that the bottleneck isn’t intelligence—it’s oftentimes efficient infrastructure. For agents to operate resourcefully within DeFi, they must plug into modular environments that allow them to execute safely, adapt intelligently, and remain accountable to human-defined constraints. But plugging into the “money legos” is very high-level. How to do it to mitigate any risk and be cost-efficient–that’s the dilemma.

As one recent analysis noted, ”Without blockchain’s inherent transparency and security, there is no trusted foundation on which AI agents can build reliable interactions”.

What’s needed is a robust foundation of vault frameworks, risk engines, and liquidity systems—each enabling the agent to take actions with safeguards in place. Modules can define what agents are permitted to do with capital, just like risk modules help them assess uncertainty, and liquidity modules allow them to monitor the available liquidity and trigger redemptions if necessary.

DeFi-Ready AI Agents

The vision of agents running vaults, rebalancing portfolios, or participating in governance is achievable. And we’re getting there. But it won’t be reached through surface-level integrations or overpromised retail bots or memecoins. What you can take away from this is that agents don’t just need intelligence; they need infrastructure. Without DeFi frameworks for agents, dynamic risk controls, and composable liquidity tooling, the road might be rocky. They need interoperability, coordination, and modular environments designed to support dynamic, cross-functional behaviour.

That’s why differentiated approaches within AI agents matter. Giza’s verifiability layer, Axal’s runtime enforcement, and Theoriq’s coordinated swarms are not competing with each other. They’re complementary.

A Tesla Cybertruck owner believed Elon Musk’s claims that the Cybertruck would be able to “act as a boat” and “cross rivers”, and he got his $100,000 stuck because of it.

Elon Musk has often made claims about how Tesla vehicles could float and briefly serve as a boat in the past.

We have never been taken too seriously because Tesla’s warranty states something different about taking the vehicle into water.

However, the CEO doubled down on the claim specifically for the Cybertruck.

Advertisement – scroll for more content

Ahead of launching the production version of the Cybertruck, Musk claimed the vehicle would be “waterproof enough” to serve as a boat and cross rivers:

Cybertruck will be waterproof enough to serve briefly as a boat, so it can cross rivers, lakes and even seas that aren’t too choppy.

The CEO added that the goal is for a Cybertruck to be able to cross the water between SpaceX’s Starbase and South Padre Island in Texas, which is about 360 meters (1,100 feet).

We have been taking the Cybertruck more seriously with water because we learned that Tesla built a ‘wade mode’ for the truck to be able to go into the water. Tesla says the mode increases the ride height to the max and temporarily “pressurizes the battery pack.”

The problem is that it is activated through the off-roading mode, which is not covered under Tesla’s warranty – so we are taking everything with a grain of salt.

Whenever Tesla’s warranty contradicts what Musk says, it is better to follow to the warranty.

A Tesla Cybertruck owner in Truckee, California, appears not to have received this sage advice since they activated the wade mode and attempted to get into the water.

The Cybertruck owner quickly got stuck. The local California Highway Patrol (CHP) shared some pictures of the aftermath (via Facebook):

CHP Truckee helped with the recovery and commented on the incident:

Cybertruck activated “Wade Mode”… and waded a bit too far… We’re all for testing boundaries… but maybe not the waterline. Remember folks, “Wade Mode” isn’t “Submarine Mode.” If your plans include exploring the great outdoors, make sure to know your limits and the terrain.

There’s no detail on the damage to the Cybertruck, if any.

As we recently reported, repair costs for the stainless steel electric pickup truck can increase rapidly.

This Cybertruck owner is also not the first one to get stuck in water. We previously reported on a Tesla Cybertruck sinking into the water when launching a jet ski.

Electrek’s Take

At the risk of stating the obvious, this is clearly more of a user error than a Cybertruck problem.

I think the verdict is clear: Cybertruck is far from the best electric pickup truck for off-roading.

However, in general, you shouldn’t expect a truck to get out of water on a muddy bank.

I think a lot of Cybertruck owners are new to trucking and off-roading, and they are making the truck look worse than it is at off-roading.

If you want to take your Cybertruck off-road, I recommend to first go with an off-roading guide that can help avoid some simple mistakes like this.

Also, in general, don’t take Elon Musk’s claims at face value when he says that Tesla vehicles can do something that sounds like an exaggeration. It probably is an exaggeration.

Getting solar at your home is one of the best financial moves you can make while helping the economy move away from fossil fuels. EnergySage enables you to find a trusted, reliable solar installer near you that offers competitive pricing It is a free service that makes it easy for you to go solar – whether you’re a homeowner or renter. They have hundreds of pre-vetted solar installers competing for your business, ensuring you get high-quality solutions and save 20 to 30% compared to going it alone. Some installers even offer Tesla products. Plus, it’s free to use and you won’t get sales calls until you select an installer and you share your phone number with them.

As of market close on April 22, each “Magnificent Seven” stock has a negative price return in 2025. Among this cohort of megacap technology stocks, Microsoft (MSFT 1.21%) and Meta Platforms (META 2.65%) have dropped by the least amounts — falling by 13% and 14.5%, respectively.

Both companies are set to report earnings for the first calendar quarter of 2025 on April 30. Let’s explore why Microsoft and Meta could be good buys right now, despite ongoing turbulence in the stock market.

Image source: Getty Images.

What road bumps could Microsoft and Meta face in the short term?

I can’t think of a bigger potential headwind for technology businesses right now outside of the new tariff policies. Both Microsoft and Meta are investing billions into AI infrastructure — from Nvidia chips to custom silicon engineering, data center buildouts, and more.

The details surrounding which items and raw materials are subject to tariffs are complex. I think it’s reasonable that both Microsoft and Meta could be looking at higher costs related to their AI infrastructure plans. In addition, it’s not entirely clear how corporations are planning for how tariffs could impact their business operations.

As a result, companies could be preparing to scale back spending in areas such as cloud computing, cybersecurity, or advertising — all of which would lead to decelerating sales for Microsoft and Meta. A slowing sales base coupled with rising prices would take a toll on profitability for each business.

One way to mitigate shrinking profits is for Microsoft and Meta to scale back their own AI capital expenditure plans. However, investors may not be encouraged by that choice since AI is the foundation of each company’s growth narrative right now. Slowing that down for the sake of near-term profitability may not sit well with investors.

Why I still like Microsoft for the long run

I see the ongoing sell-off across the tech sector as an opportunity to buy the dip in high-quality names. Right now, Microsoft’s forward price-to-earnings (P/E) ratio of 28 is slightly below the company’s three-year average.

MSFT PE Ratio (Forward) data by YCharts

Even though IT budgets could be operating under tighter controls for the time being, I tend to think that businesses are going to identify cost savings in areas outside of mission-critical infrastructure such as cloud computing and cybersecurity software.

Although I’m not expecting a monster quarter from Microsoft next week, I remain cautiously optimistic that cloud growth from Windows Azure will show some signs of resilience. When you complement this with Microsoft’s diversified ecosystem that includes personal computing, social media (LinkedIn), gaming, and more, I see Microsoft as a business that is relatively insulated from a possible economic slowdown caused by the tariff environment.

Why I still like Meta for the long run

On the surface, you might think that Meta is facing outsized pressure compared to its peers given the company really only has two sources of growth: advertising and the metaverse. Candidly, the company’s metaverse ambitions are far from reaching widespread scale or profitability, and the digital advertising landscape is packed with competition from the likes of Alphabet, TikTok, and Snap, just to name a few. With that said, I think these are surface-level arguments.

Meta’s relative price resilience compared to its Magnificent Seven peers could suggest that investors are less worried about the company’s growth prospects. I think this makes sense, too. I don’t see tariffs having much of an impact on Meta’s business overall. Similar to Microsoft, the company could witness a brief slowdown in revenue growth, but I don’t think it will be detrimental.

With leading platforms including Facebook, WhatsApp, and Instagram in its ecosystem, Meta is in a lucrative position to continue monetizing its billions of users — especially as AI tailwinds unlock new opportunities in the consumer market.

META PE Ratio (Forward) data by YCharts

As of this writing, Meta is trading right in line with its three-year average forward P/E. Considering the company has made huge strides in the world of AI to help diversify the business over the last three years, it would appear that investors aren’t applying much value to this potential growth right now.

Remember to think long term

The big thing investors should keep in mind is that these tariff policies could change at any time. Moreover, even if trade negotiations with other countries linger to the point of an economic slowdown, such a cycle won’t last forever.

In the meantime, investors are continuing to sell off growth stocks given all of the uncertainty in the market right now. In my eyes, Microsoft and Meta are trading for reasonable valuations and I think investors should take advantage, buy the dip while it lasts, and prepare to hold on for the long term.

As of market close on April 22, each “Magnificent Seven” stock has a negative price return in 2025. Among this cohort of megacap technology stocks, Microsoft (MSFT 1.21%) and Meta Platforms (META 2.65%) have dropped by the least amounts — falling by 13% and 14.5%, respectively.

Both companies are set to report earnings for the first calendar quarter of 2025 on April 30. Let’s explore why Microsoft and Meta could be good buys right now, despite ongoing turbulence in the stock market.

Image source: Getty Images.

What road bumps could Microsoft and Meta face in the short term?

I can’t think of a bigger potential headwind for technology businesses right now outside of the new tariff policies. Both Microsoft and Meta are investing billions into AI infrastructure — from Nvidia chips to custom silicon engineering, data center buildouts, and more.

The details surrounding which items and raw materials are subject to tariffs are complex. I think it’s reasonable that both Microsoft and Meta could be looking at higher costs related to their AI infrastructure plans. In addition, it’s not entirely clear how corporations are planning for how tariffs could impact their business operations.

As a result, companies could be preparing to scale back spending in areas such as cloud computing, cybersecurity, or advertising — all of which would lead to decelerating sales for Microsoft and Meta. A slowing sales base coupled with rising prices would take a toll on profitability for each business.

One way to mitigate shrinking profits is for Microsoft and Meta to scale back their own AI capital expenditure plans. However, investors may not be encouraged by that choice since AI is the foundation of each company’s growth narrative right now. Slowing that down for the sake of near-term profitability may not sit well with investors.

Why I still like Microsoft for the long run

I see the ongoing sell-off across the tech sector as an opportunity to buy the dip in high-quality names. Right now, Microsoft’s forward price-to-earnings (P/E) ratio of 28 is slightly below the company’s three-year average.

MSFT PE Ratio (Forward) data by YCharts

Even though IT budgets could be operating under tighter controls for the time being, I tend to think that businesses are going to identify cost savings in areas outside of mission-critical infrastructure such as cloud computing and cybersecurity software.

Although I’m not expecting a monster quarter from Microsoft next week, I remain cautiously optimistic that cloud growth from Windows Azure will show some signs of resilience. When you complement this with Microsoft’s diversified ecosystem that includes personal computing, social media (LinkedIn), gaming, and more, I see Microsoft as a business that is relatively insulated from a possible economic slowdown caused by the tariff environment.

Why I still like Meta for the long run

On the surface, you might think that Meta is facing outsized pressure compared to its peers given the company really only has two sources of growth: advertising and the metaverse. Candidly, the company’s metaverse ambitions are far from reaching widespread scale or profitability, and the digital advertising landscape is packed with competition from the likes of Alphabet, TikTok, and Snap, just to name a few. With that said, I think these are surface-level arguments.

Meta’s relative price resilience compared to its Magnificent Seven peers could suggest that investors are less worried about the company’s growth prospects. I think this makes sense, too. I don’t see tariffs having much of an impact on Meta’s business overall. Similar to Microsoft, the company could witness a brief slowdown in revenue growth, but I don’t think it will be detrimental.

With leading platforms including Facebook, WhatsApp, and Instagram in its ecosystem, Meta is in a lucrative position to continue monetizing its billions of users — especially as AI tailwinds unlock new opportunities in the consumer market.

META PE Ratio (Forward) data by YCharts

As of this writing, Meta is trading right in line with its three-year average forward P/E. Considering the company has made huge strides in the world of AI to help diversify the business over the last three years, it would appear that investors aren’t applying much value to this potential growth right now.

Remember to think long term

The big thing investors should keep in mind is that these tariff policies could change at any time. Moreover, even if trade negotiations with other countries linger to the point of an economic slowdown, such a cycle won’t last forever.

In the meantime, investors are continuing to sell off growth stocks given all of the uncertainty in the market right now. In my eyes, Microsoft and Meta are trading for reasonable valuations and I think investors should take advantage, buy the dip while it lasts, and prepare to hold on for the long term.

We recently published a list of Why These 10 Firms Recorded Double-Digit Gains Last Week. In this article, we are going to take a look at where SoundHound AI Inc. (NASDAQ:SOUN) stands against other firms that recorded double-digit gains last week.

Ten companies finished the past trading week on a high note, outperforming the three major indices with double-digit gains amid a flurry of catalysts, including upcoming earnings releases, that sparked buying appetite.

Week-on-week, the Nasdaq rallied by 6.7 percent, the S&P 500 clocked in a 4.59-percent gain, while the Dow Jones was up by 2.48 percent.

In this article, we have listed last week’s top 10 performing stocks and detailed the reasons behind their gains. The stocks were chosen based on the highest percentage increase in closing prices on April 25 as against their prices on April 17, or the week prior.

To come up with the list, we considered only the companies with a $2-billion market capitalization and a $5-million trading volume.

A software engineer focused on a computer screen, writing code to create a conversational assistant.

SoundHound AI Inc. (NASDAQ:SOUN)

SoundHound AI jumped by 21.7 percent week-on-week to close Friday’s trading at $9.52 apiece versus the $7.82 finish on April 17 as investors cheered news that it partnered with Tencent Intelligent Mobility to enhance the latter’s cloud-based solutions for the automotive industry.

In a statement, SoundHound AI Inc. (NASDAQ:SOUN) said that Tencent is working on integrating SoundHound Chat AI Automotive to give drivers and passengers access to a range of applications, car controls, and entertainment domains just through natural speech.

SoundHound AI Inc.’s (NASDAQ:SOUN) conversational voice intelligence is capable of handling complex conversational questions and intelligently filtering responses to ensure users get relevant and useful responses.

“The collaboration will enable Tencent Intelligent Mobility to deliver dynamic, localized user experiences to automotive players across geographies that are looking to provide end users with seamless hands-free access to a range of in-vehicle applications,” it said.

Overall, SOUN ranks 7th on our list of firms that recorded double-digit gains last week. While we acknowledge the potential of SOUN as an investment, our conviction lies in the belief that AI stocks hold greater promise for delivering higher returns and doing so within a shorter time frame. There is an AI stock that went up since the beginning of 2025, while popular AI stocks lost around 25%. If you are looking for an AI stock that is more promising than SOUN but that trades at less than 5 times its earnings, check out our report about this cheapest AI stock.

If you’re a sports fan, April is a smorgasbord.

Baseball is back. The NBA playoffs are in full swing. The NFL draft is turning college stars into household names. Even bowling is rolling into the digital age (although it still can’t quite command the couch the way football does).

But here’s the real headline: Artificial intelligence isn’t just in the owner’s box. It’s now in the dugout, on the field, in your running shoes, and even in your local bowling alley. AI and edge computing have become the ultimate utility players, reshaping how pros compete, how fans watch, and how weekend warriors chase their own personal bests.

Football

AI’s influence on football is as pervasive as a well-executed zone defense. Today’s NFL teams deploy AI-powered sensors in helmets and pads, collecting granular data on player movement and collisions. Machine learning models analyze this information to optimize training, minimize injury risk, and suggest play calls based on real-time opponent tendencies. On the sidelines, AI sifts through mountains of game film, helping coaches anticipate rival strategies and adjust in the heat of battle. For fans, AI overlays in broadcasts provide instant statistical insights and predictive analytics, transforming the viewing experience from passive to interactive.

Soccer

The world’s game has embraced AI for everything from scouting to tactics. Clubs use AI-driven video analysis to break down thousands of hours of match footage, identifying emerging talent and subtle tactical patterns. Wearable tech tracks player stamina and positioning, while edge computing enables instant feedback on the pitch. AI-powered apps like AiScout let young players upload their drills for algorithmic evaluation, democratizing access to elite-level scouting and opening new doors for aspiring stars. On match day, AI-enhanced camera systems automate offside calls and track ball movement, improving officiating accuracy and reducing controversy.

Baseball

Tradition? Pinstripes? Baseball has become a playground for AI innovation. Teams employ data scientists to simulate every conceivable pitcher-batter matchup, using deep learning to forecast outcomes and tailor training regimens. AI models monitor player mechanics and workloads, flagging potential injury risks before they sideline a star. In the broadcast booth, AI-driven highlight generators serve up instant replays and personalized content, keeping fans engaged inning by inning. The sport’s embrace of AI doesn’t come without ethical dilemmas — data privacy and competitive balance are now as much a part of the conversation as batting averages and ERA.

Bowling

And then there’s bowling, which has quietly undergone a digital transformation. AI devices now power real-time scoring and deliver instant feedback on ball speed, spin and trajectory. Systems like Gemini AI offer personalized coaching tips and even generate dynamic commentary, making the bowling alley a hub of data-driven improvement. Verified remote competitions and global leaderboards connect bowlers worldwide, turning a local pastime into a global digital sport.

Running

Mobile apps equipped with computer vision and biomechanics algorithms analyze a runner’s gait, posture and stride, offering personalized drills to improve efficiency and reduce injury risk. AI-driven coaching platforms adapt training plans on the fly, responding to fatigue, weather and performance data in real time. Wearable devices process data at the edge, delivering instant feedback without the need for cloud connectivity — a boon for runners pounding the pavement in parks or on remote trails.

Tennis

Systems like PlaySight and SmartCourt analyze every shot, spin and movement, providing players and coaches with actionable insights during matches. AI-powered virtual coaches offer feedback on technique and tactics, making elite-level analysis accessible to club players and juniors. Automated line-calling systems like Hawk-Eye have largely eliminated human error, while AI-enhanced broadcasts deliver real-time stats and predictive analytics to fans worldwide.

Aerobics

Smart mirrors and fitness apps now use AI-powered motion tracking to assess form, count reps and provide instant corrections during workouts. These systems adapt routines based on individual progress, fatigue and goals, offering a level of personalization once reserved for private trainers. Edge computing ensures that feedback is delivered instantly, keeping participants motivated and engaged. AI-driven platforms also enable virtual group classes, where participants receive real-time encouragement and performance analytics, turning solitary workouts into interactive, community-driven experiences.

The Bottom Line

Across these seven sports, AI has become the silent MVP, analyzing data, optimizing performance, preventing injuries and enhancing the fan experience. Whether you’re chasing a championship, a personal best or just a little fun at the bowling alley, the digital revolution is here, and it’s making every game smarter, safer and more exciting. So next time you tune in or lace up, remember: The future of sports isn’t just on the field; it’s in the algorithms running quietly behind the scenes.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

Ahead of a grand prix weekend, most of us like to share predictions or try and guess who will come out on top on a Sunday. Data scientist Mariana Antaya took those chats one stage further and built a machine learning model to try and predict F1 race results. So far, her model has correctly called the winners of three grands prix this season.

“I’m a really big Formula 1 fan,” says Antaya when speaking with Motorsport.com. “Machine learning and all these algorithms are really widely used in Formula 1 by the teams. I don’t think as many people know, but the race engineers are using this for their strategy in real time.

“So, I wanted to try to predict the winner as a fun exercise, just to see, like, how good we can get with the data that’s available.”

To do this, Antaya started building a model of her own. Armed with lap times from last year’s Australian Grand Prix, which was sourced from the FastF1 API data store, Antaya set about comparing the 2024 race result with qualifying performances in 2025.

Once the rookies were removed from the program, which Antaya admits is the one factor she “interfered with” as there was no data to benchmark against, she began training her model. Using a gradient boosting tool, Antaya predicted the lap times for the race in Albert Park, and her program correctly picked Lando Norris as the winner.

“I said at the end of the video, this is obviously a simple model, and I didn’t know it was going to predict right,” Antaya says. From there, the project started growing as the F1 community gathered around to see how many more races Antaya could correctly call.

“I wanted it to be a crowdsourced type of thing,” she adds. “So, all of the audience could say ‘I really want you to include weather data in it,’ or ‘I really want you to include the practice sessions in the model.’

“I wanted people to tell me what other features they wanted to add to the model to improve it over the course of the season.”

Formula 1 Fan Mariana Antaya

Photo by: Mariana Antaya

And improve it has, as the machine learning model is continuing to predict race winners correctly. This doesn’t mean it’s perfect, however, and Antaya is now adding more datapoints to the program to help increase its accuracy.

“Having more data is going to help the model learn more and it’s going to be able to make better predictions,” she explains. “If you only have so much data, it’s going to have a very small mind, I guess, and it won’t be able to understand as much.”

In order to expand the mind of her model, Antaya added weather data ahead of the Japanese Grand Prix, which included the chance of rain during the race and track temperatures at Suzuka. In addition to this, wet-weather performance of the drivers was also added, and the program used this to correctly predict Max Verstappen’s victory at the race.

The next big step for the model came ahead of the Saudi Arabian Grand Prix this weekend, when it was trained on each team’s performance so far this year. Antaya explained that the extra strand of data would help her program understand that teams like McLaren and Williams have made a step forward in 2025, while others such as Red Bull aren’t performing consistently as well as they were in 2024.

“Now we’re taking into consideration more of a holistic picture of how well the car and the team is performing,” she explains.

‘Surprised’ by the series

The series of posts on Instagram and TikTok has been growing in popularity with each successive upload, and the clips have even reached Formula 1 itself. A handful of engineers from F1 teams on the grid reportedly reached out to Antaya after she started uploading, and she’s now looking forward to finding out how close she got to the prediction models used in the series.

“I’ve been shocked [by the response]. I’ve been really, really surprised,” she says. “I honestly have no idea [how the teams do it]. That’s a black box to me, I wish I knew. But I hope I’m doing it correctly or something similar. They are using, probably, much more complex models and much more data that they have on the car though, for sure.”

Hannah Schmitz, Principal Strategy Engineer of Red Bull Racing

Photo by: Peter Fox – Getty Images

With three out of five race winners correctly predicted, Antaya isn’t resting on her laurels as she hopes to make the predicter even more accurate. Ahead of the Miami Grand Prix, the data scientist says she wants to start experimenting with more complex machine learning processes to increase the accuracy of her predictions and reduce the mean absolute error of the model, which can be thought of as the average difference between the model’s predictions and the race result.

But while the accuracy of the model could increase thanks to additional datapoints and new processes being implemented, Antaya is aware that in F1 there will always be unpredictable elements.

“I think there’s always going to be that barrier,” she adds. “It’s really hard to be able to tell that there’s going to be a safety car this lap, and that this is then going to trigger some other stream of events.

“Maybe we could pull past data on crash percentage during the race, and that’s something that we can add as another feature. But it’s also a sport, so it’s not like we can look into the future and see what’s going to happen all the time.”