‘We're able to explore more behaviors a lot more quickly': How one local robotics company uses AI
for investigating its 'continued eligibility' for a program that gives visas to international students and faculty. Columbia
to restore its federal funding after the administration accused it of allowing antisemitism.
President Trump's name appears in documents related to Jeffrey Epstein that his administration is refusing to release,
. Separately, a House committee voted to subpoena
and files from the administration, while a judge rejected an administration request to unseal other documents in the case.
Two fired Market Basket executives accused the grocery chain's board of dismissing them
Write to us at
. To subscribe,
.
TODAY'S STARTING POINT
In conversations about artificial intelligence, the future tense tends to get a workout. Will AI lead to mass layoffs? Which AI company will end up on top? Will the US's new AI strategy, which the White House
Yet for many Americans, AI is already a daily reality. AI-powered chatbots provide customer service,
That's also true for local businesses. Earlier this year, the Greater Boston Chamber of Commerce was looking for ways to help firms navigate AI adoption. 'People seem to be hungry for, 'Well, how are other people using it?'' said Jim Rooney, who leads the Boston Chamber.
Advertisement
Rooney's solution was a survey. The anonymized results,
I wanted to better understand what local businesses' use of AI looks like in practice. Today's newsletter is the first in a two-part series about what I found.
Advertisement
Better, faster, smarter
When I visited Piaggio Fast Forward on a recent Friday, the Charlestown-based robotics company's brick-walled offices were largely devoid of people. But they were populated by something else: a fleet of squat, rotund machines.
This was the gita (pronounced like a Bostonian saying Derek Jeter's surname), which Piaggio unveiled in 2019. Essentially a backpack on wheels, the gita
mini
and gita
plus
— and stores groceries, gear, or other cargo beneath a central hatch. The robot's camera and sensors detect color and depth, which, with the push of a button and a warbly chime, lets it 'pair' with a specific person and automatically trundle along behind them during errands. The experience feels like having R2-D2 at your heels. (The resemblance is no longer implicit; Piaggio recently reached a licensing agreement with Lucasfilm and Disney to sell $2,875 gita
mini
s that
So where does AI come in? When Tyson Phillips first joined Piaggio to lead its research and development team, the technology wasn't on the menu. But about two and a half years ago, he came to see it as a necessity. Phillips's engineers built their own AI models to help train the company's robots to interact with people and the environment. 'It's actually very difficult to program a robot to do something,' Phillips said. 'AI is shortening that process a lot.'
To train the machines, Phillips invites paid human guinea pigs into a high-ceilinged space in Piaggio's offices. Its floor features lines of colorful tape, mannequins, and other obstacles that simulate what a robot might encounter in the outside world, like doors, walls, and people. Using motion capture cameras, Phillips records the volunteers navigating those obstacles, then distills the data into algorithms to program the robots.
Advertisement
A Piaggio Fast Forward employee presses a button to "pair" with one of the robotics company's gitamini robots.
David L. Ryan/Globe Staff
Part of AI's advantage is that it
isn't
human. Where a flesh-and-blood observer might write off a volunteer's subtle turn or weight shift while opening a door as intuitive, AI can recognize such moves as potentially valuable datapoints. 'With AI tools, we are hoping to identify those much smaller, more nuanced behaviors,' Phillips said. And for a relatively small company like Piaggio, the added analytical firepower helps. 'We're able to explore more behaviors a lot more quickly.'
That has come in handy to train Piaggio's other robot, kilo, which looks a bit like if Apple designed a flatbed cart. Built for warehouses and factory floors, kilo
'I think we'll use it to allow us to explore situations that we would've been previously uncomfortable in,' he said.
Evolutionary vs. revolutionary
Piaggio isn't alone. 'Every robotics company's using AI in some way,' Phillips told me. Yet some prognosticators worry that widespread adoption will cause layoffs, particularly among coders.
So far, Piaggio says, AI hasn't replaced anyone on Phillips's team, which includes people with backgrounds in AI as well as in biomechanics and neuroscience. In the Boston Chamber's survey, just 7 percent of companies reported job reductions because of AI.
Advertisement
Instead, just as Piaggio's robots are designed to work alongside people, Phillips hopes that AI will supplement rather than supplant. His engineers still develop algorithms by hand, using AI to check their work. 'We have an idea, we test it, we check it with AI, we tweak the original idea, we tweak the AI, and it bounces back and forth,' he said.
Still, that collaborative spirit helps explain why Piaggio's use of AI is mostly evolutionary, facilitating its business without radically reshaping it. But there are companies in Greater Boston trying to use the technology in more revolutionary ways. Next week, in part two of this series, I'll explore that.
Related:
As Massachusetts lawmakers try to lure AI data centers, environmental advocates from elsewhere warn that the centers
🧩
9 Across:
90°
POINTS OF INTEREST
UNITE HERE Local 26 represents Fenway Park concession workers who are threatening to strike.
Cassandra Klos/Bloomberg
Receipts:
Massachusetts lawmakers say court-appointed lawyers should've asked the Legislature for raises sooner, rather than launching a work stoppage that has paralyzed state courts. Emails show the attorneys have been asking
Labor dispute:
Fenway Park concession workers
Piling up:
Boston
'Pawtriots':
The Patriots' first day of training camp saw the team practicing without pads —
Advertisement
Feeling better:
Eight members of a Parisian youth choir are OK
Dwight Evans:
The former Red Sox right fielder has another shot at making the Hall of Fame this year.
Grant Watch:
How a researcher from Medfield created the go-to database
Changing the subject:
Trump and Tulsi Gabbard, his national intelligence director, baselessly claim that Barack Obama, Hillary Clinton, and other Democrats plotted a 'coup' against Trump years ago and that they could face criminal charges. (
Wares the beef:
Beef prices are climbing across the country as drought
Sentenced:
A judge gave Bryan Kohberger life in prison without the possibility of parole for killing four University of Idaho students in 2022. The victims' family members excoriated Kohberger. (
BESIDE THE POINT
🌽
Cultivating community:
An 18th-century farm is flourishing
🎙️
Big deal:
'The Big Dig,' GBH and PRX's nine-episode series about the infamous Boston megaproject, made Time magazine's list of the 100 best podcasts of all time. (
🤦🏻♀️
A Miss Conduct classic:
Their neighbors' floodlights shine directly into their bedroom. The neighbors
🐻
'Da Pope':
A Chicago family on vacation gave Pope Leo XIV a T-shirt inspired by the nickname for the city's NFL team. (
Advertisement
📺
Don't be a Debbie Downer:
'SNL' alumna Rachel Dratch discusses
💃
A step ahead:
A Boston soul line dancing group is
🎶
Don't worry:
Does live music at Logan baggage claim actually reduce travel stress?
Thanks for reading Starting Point.
This newsletter was edited by
❓ Have a question for the team? Email us at
✍🏼 If someone sent you this newsletter, you can
📬 Delivered Monday through Friday.
Ian Prasad Philbrick can be reached at
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
17 minutes ago
- Yahoo
Exclusive-Trump administration to formally axe Elon Musk's 'five things' email
WASHINGTON (Reuters) -The Trump administration plans as soon as Tuesday to formally axe a program launched by billionaire former Trump adviser Elon Musk requiring federal employees to summarize their five workplace achievements from the prior week, two people familiar with the matter said. The Office of Personnel Management, the federal human resources agency that implemented Musk's push to slash the federal workforce, plans to announce the end of the "five things" email to HR representatives across the federal government later on Tuesday, the two people said, declining to be named because the matter was not public. While many federal agencies had already phased out compliance with the weekly email, the move, not previously reported, signals the Trump administration is turning the page on one of Musk's most unpopular initiatives following a dramatic row between the two men in early June. The White House and OPM did not immediately respond to requests for comment. Musk, who spent over a quarter of a billion dollars to help Trump win November's presidential election, led the Department of Government Efficiency's efforts to slash the budget and cut the federal workforce until his departure in May to refocus on his tech empire. Musk initially received a warm White House sendoff from Trump, but then incurred the president's wrath by describing Trump's tax cut and spending bill as an abomination. Trump pulled the nomination of Musk ally and tech entrepreneur Jared Isaacman to lead NASA and later threatened to cancel billions of dollars worth of federal contracts with Musk's companies after the blowup between the two men. The "five things" email, launched by Musk in February to boost accountability, sparked tensions with department chiefs who were blindsided by the weekend email mandating the move. It also fueled confusion among government workers who received mixed messages about whether and how to comply. Reuters reported in March that the White House installed two Trump loyalists at OPM to ensure better policy coordination between the White House and the agency. Scott Kupor, a venture capitalist who took the helm at OPM in July, foreshadowed the end of the initiative last month, describing processing of the weekly response emails as "very manual" and "not efficient." It is "something that we should look at and see, like, are we getting the value out of it that at least the people who put it in place thought they were," he said.
USA Today
18 minutes ago
- USA Today
Worried about AI at work? Avoid these 5 leadership mistakes with your team
Artificial intelligence may be transforming the workplace, but for many employees, it's fueling uncertainty instead of excitement. According to a 2025 Pew Research Center study, 52% of U.S. workers worry AI could disrupt or replace their jobs. And an August 2024 SHRM survey found that nearly half feel unprepared for automation, while 95% say they don't trust their organization to manage the shift in a way that benefits everyone. How managers address these concerns can make or break team morale and productivity. Experts say clear, honest communication is critical, but the wrong message can backfire, fueling fear instead of trust. Whether you're rolling out new tools or just starting the conversation, it's important to engage your team with transparency, context and empathy. Below, two human resources experts break down five common mistakes to avoid when discussing AI with your team and provide guidance on navigating the discussion more effectively. 1. Acting like it's no big deal According to the World Economic Forum, when managers dismiss or avoid discussing AI concerns, they often create bigger problems down the road. 'Business leaders can't bury their heads in the sand and hope for the best,' says Eric Mochnacz, director of operations at Red Clover HR in New Jersey. 'They must have up-front discussions about the benefits of AI in their business, the drawbacks, the potential impacts and the areas where they'll not allow AI usage.' Chad V. Sorenson, president of Florida-based Adaptive HR Solutions, agrees that direct communication is key. 'Employees may feel AI threatens their jobs and may question leaders' motives for introducing AI tools,' he explains. So, 'address the fear and explore how AI can augment workflows and streamline repetitive tasks rather than replace workers.' Takeaway: Don't downplay concerns about AI. Acknowledge employee fears openly and explain how AI will support, not replace, their work. 2. Throwing around 'AI' without defining it AI isn't just one thing. Mochnacz explains that there's a significant difference between using generative AI to enhance email communications and utilizing AI chatbots to manage all customer interactions. Without these distinctions, employees don't understand what to expect from their workplace changes or how they can remain relevant. 'AI is such a buzzword, and leaders haven't taken the time to define it and understand the differences,' says Mochnacz. 'I've been in meetings where people ask, 'Can you do this with AI?' or 'Everyone's talking about AI, so we have to do something with it.'' He emphasizes the importance of clarifying the fundamentals: Takeaway: Don't use 'AI' as a vague catchall. Clearly define what types of AI you're using, what they do and why they matter to your team. 3. Failing to explain the why 'Any time leaders announce a new program or procedure without employee buy-in, there could be fear, skepticism or anger,' Sorenson cautions. Instead of simply telling employees what's changing, explain why the company needs AI. How does it fit into broader business goals? This context helps employees understand their role in the transition rather than viewing it as a threat. 'Ongoing two-way feedback is critical for continued refinement of how teams use AI to improve workflows, processes and results,' notes Sorenson. Takeaway: Don't skip the context. Explain why AI is being adopted and how it supports your team's goals to build trust and buy-in. 4. Overhyping what AI can do 'Leaders promising that AI will handle everything don't have a clear understanding of its possibilities and limitations,' Sorenson says. For instance, AI can help employees understand benefit plans. But it can't handle nuanced harassment complaints or mental health concerns. 'AI systems must be trained to understand when a human must intervene,' he adds. Mochnacz says the problem gets worse when leaders promise capabilities that may never materialize. 'We have no idea what AI is going to be able to do in a month, three months or a year,' he emphasizes. So, it's better to focus on specific, tested use cases rather than grand predictions about AI replacing everything. Takeaway: Avoid making big promises. Focus on what AI can realistically do today, not speculative future capabilities. 5. Leaving people out of the process 'Whenever there's a business, industry or technology shift, involve those it may impact,' stresses Mochnacz. 'Have up-front, direct conversations with your people about their roles and how they see AI helping them be more effective.' A recent MIT Sloan working paper finds that the most successful generative AI deployments consistently involve frontline workers from the earliest stages through rollout. Drawing on over 50 in-depth interviews, MIT researchers demonstrate that when employees help define the problem, co-design workflows, experiment with tools, and shape fair transition policies, not only does adoption improve, but worker productivity and job quality also rise. The key here is framing AI as a collaboration partner rather than a threat. Mochnacz explains that when leaders present AI as a good reality for everyone, employees will engage with the technology. But when the message becomes "prove AI can't replace you," workers resist because it feels like an ultimatum. Takeaway: Don't make AI decisions in a vacuum. Engage employees early and frame AI as a tool to support their work, rather than compete with it. What successful AI communication looks like Sorenson says poor AI communication shows up in obvious ways. You might notice more pushback in meetings, higher employee turnover or a spike in anxious watercooler conversations. These signals suggest that employees feel excluded or uncertain and may be bracing for the worst. In contrast, when communication is clear and inclusive, team engagement improves. 'If your AI communication strategy is successful, employees should engage in the conversation,' Sorenson notes. 'They'll make suggestions to continue to refine its use, and demonstrate an increased productivity level.' Look for those signs of healthy adoption: employees asking questions, suggesting improvements and using AI to work more efficiently. When teams feel empowered, not threatened, you know you've struck the right balance. What is USA TODAY Top Workplaces 2025? Do you work for a great company? Each year, USA TODAY Top Workplaces, a collaboration between Energage and USA TODAY, ranks organizations across the U.S. that excel at creating a positive work environment for their employees. Employee feedback determines the winners. In 2025, over 1,500 companies earned recognition as top workplaces. Check out our overall U.S. rankings. You can also gain insights into top-ranked employers by checking out the links below.
The Hill
18 minutes ago
- The Hill
OpenAI, Google, Anthropic AI models added to government purchasing system
Artificial intelligence (AI) models from OpenAI, Google and Anthropic have been added to a government purchasing system, allowing federal agencies to buy and use the AI products. The General Services Administration (GSA) announced Tuesday that ChatGPT, Gemini and Claude had been added to the agency's Multiple Award Schedule for purchase. 'America's global leadership in AI is paramount, and the Trump Administration is committed to advancing it,' GSA acting administrator Michael Rigas said in a statement. 'By making these cutting-edge AI solutions available to federal agencies, we're leveraging the private sector's innovation to transform every facet of government operations,' he continued. This follows the addition of xAI's Grok to the GSA schedule, which it announced last month after unveiling a new suite of products for U.S. government customers and scoring a Pentagon contract alongside the three other tech firms. The agency pointed to President Trump's AI Action Plan for the new additions to its purchasing system. The AI framework, released last month, called for accelerating AI adoption in the federal government. It specifically advocated for the creation of an AI procurement toolbox managed by the GSA that would 'allow any Federal agency to easily choose among multiple models in a manner compliant with relevant privacy, data governance, and transparency laws.' The recommendations for federal AI adoption represent one small portion of Trump's wide-ranging AI Action Plan, which also called for limiting state and federal regulations, fast-tracking permitting for data center and energy construction and creating export packages of U.S. technology.
