This New AI Competes GPT-6 and Claude

In partnership with

MiniMax’s New AI Competes GPT-6 and Claude

MiniMax just launched M3.1 Flash, a new AI model built for coding, computer use, and autonomous agents. In Plotly's independent data-analysis benchmark, it scored 95% (41/43), tying GPT-6 Sol and Claude Opus 5.5. MiniMax claims nearly 3× faster inference and 70% lower usage costs compared with its previous model. API pricing starts at just $0.15 per million input tokens and $0.60 per million output tokens. But there's a catch: despite its fast token generation, Plotly found it considerably slower at completing full analytical tasks. M3.1 Flash is available now through MiniMax's API and coding tools, with open model weights planned for later. ^{_{Source: Minimax}}

The $0 AI Side Hustle Smart People Are Quietly Starting in 2026

Arnold Schwarzenegger. Jay Shetty. And You.

Codie Sanchez, Colin & Samir, Joanna Stern, and Oliver Darcy are here too. They all built newsletters (and podcasts and communities and websites) on beehiiv.

Some came with massive followings. Others built theirs along the way. That’s the thing about beehiiv: with so many ways to earn and grow, the platform is ready for whatever you’re building next.

Get started for free.

Tesla Roadster on Road passing through railway crossing

(Prompt is attached below)

Grab this prompt to reverse engineer this video

TESLA ROADSTER | CHICAGO RAILWAY CROSSING
Note: I use invideo to generate this, you can try it here… 
Final 30-Second Realistic iPhone Found-Footage Prompt

MASTER IMAGE LIBRARY
Use these exact uploaded photographs to preserve the original Roadster design. Each short identifier represents one file.
Exterior and body
@Image 1 = image1 = ROADSTER_TOP_FRONT_OVERHEAD_DOORS_CLOSED_STUDIO.jpg | Front overhead geometry
@Image 2 = image2 = ROADSTER_REAR_THREE_QUARTER_SIDE_DOORS_CLOSED_STUDIO.jpg | Side and rear proportions
@Image 3 = image3 = ROADSTER_REAR_STRAIGHT_ON_DARK_STUDIO.jpg | Direct rear design
@Image 4 = image4 = ROADSTER_TOP_REAR_OVERHEAD_WING_DEPLOYED_STUDIO.jpg | Rear wing and roof
@Image 5 = image5 = ROADSTER_REAR_STRAIGHT_ON_DAYLIGHT_WING_DEPLOYED_DIFFUSER.jpg | Rear mechanical layout
@Image 6 = image6 = ROADSTER_TOP_DOWN_OVERHEAD_DESERT_FULL_BODY.jpg | Complete overhead footprint
@Image 7 =image7= ROADSTER_TOP_FRONT_OVERHEAD_BUTTERFLY_DOORS_OPEN.jpg | Door and roof geometry
@Image 8 =image8= ROADSTER_REAR_THREE_QUARTER_DOORS_OPEN_WING_DEPLOYED_STUDIO.jpg | Rear-quarter panel geometry
@Image 9 = image9 = ROADSTER_FRONT_THREE_QUARTER_BUTTERFLY_DOORS_OPEN_STUDIO.jpg | Front three-quarter shape
@Image 10 = image10 = ROADSTER_REAR_THREE_QUARTER_DESERT_PILOT_APPROACH_HELMET.jpg | Vehicle scale and exterior
Wheels, bodywork and propulsion
@Image 11 = image11 = ROADSTER_FRONT_WHEEL_AND_FRONT_FENDER_CLOSEUP.jpg | Front wheel and fender
@Image 12 = image12 = ROADSTER_WHEEL_RIM_TESLA_CENTER_CAP_MACRO.jpg | Exact wheel spokes and Tesla logo
@Image 13 = image13 = ROADSTER_FRONT_HOOD_NOSE_PANEL_SEAMS_CLOSEUP.jpg | Front hood geometry
@Image 14 = image14 = ROADSTER_REAR_TOP_DOWN_TAILLIGHT_REAR_DECK_CLOSEUP.jpg | Rear light and deck
@Image 15 = image15 = ROADSTER_REAR_WING_AIRFOIL_HINGE_CLOSEUP_DEPLOYED.jpg | Wing hinge and mounting
@Image 16 = image16 = ROADSTER_REAR_DIFFUSER_FOUR_CIRCULAR_NOZZLES_CLOSEUP.jpg | Four rearward outlets
@Image 17 = image17 = ROADSTER_REAR_VIEW_AIRBORNE_DOWNWARD_GAS_PLUMES.jpg | Lift vapor appearance
Cockpit and interior
@Image 18 = image18 = ROADSTER_INTERIOR_COCKPIT_WIDE_PASSENGER_SIDE_THREE_QUARTER.jpg | Complete front cockpit
@Image 19 = image19= ROADSTER_INTERIOR_DASHBOARD_STRAIGHT_AHEAD_VIEW_FROM_BEHIND_SEATS.jpg | Dashboard layout
@Image 20 = image20 = ROADSTER_INTERIOR_DRIVER_YOKE_DASHBOARD_CENTER_SCREEN_CLOSEUP.jpg | Yoke and display
@Image 21 = image21 = ROADSTER_INTERIOR_STEERING_YOKE_COLUMN_REAR_ANGLE_DETAIL.jpg | Steering controls
@Image 22 = image22 = ROADSTER_INTERIOR_TWO_WHITE_BUCKET_SEATS_SIDE_VIEW_PANORAMIC_ROOF.jpg | Seats and canopy
@Image 23 = image23 = ROADSTER_INTERIOR_WHITE_BUCKET_SEATS_FRONT_FACE_AND_CENTER_DISPLAY.jpg | Seat and display geometry
@Image 24 = image24 = ROADSTER_INTERIOR_SEAT_HEADREST_RED_ROADSTER_WORDMARK_CLOSEUP.jpg | Upholstery and branding
Airborne movement
@Image 25 = image25 = ROADSTER_FLIGHT_AERIAL_HIGH_ANGLE_OVER_PARKED_VEHICLES_VAPOR_JETS.jpg | Airborne obstacle crossing
@Image 26 = image26 = ROADSTER_FLIGHT_SIDE_PROFILE_LEVEL_HOVER_DOWNWARD_THRUST_PLUMES.jpg | Level lift and downward jets
@Image 27 = image27 = ROADSTER_FLIGHT_REAR_VIEW_BESIDE_VERTICAL_RAMP_DOWNWARD_JETS.jpg | Rear flight and jet behavior
@Image 28 = DRIVER_MASTER = My separately uploaded driver identity reference
Open-door, night and desert references establish vehicle geometry only. Do not reproduce those environments or open the doors during driving.

FILMING STYLE AND ENVIRONMENT
Create a 30-second ultra-realistic fictional Tesla Roadster encounter, filmed entirely on a modern iPhone by an ordinary passenger inside another car.
Location: Same Chicago urban industrial neighborhood, late afternoon. Busy multilane street, ordinary shops, brick buildings, sidewalks, pedestrians, delivery vehicles, buses and natural city activity.
Camera: Handheld iPhone, vertical 9:16, 4K, 30 fps. Natural smartphone HDR, realistic autofocus, slight exposure fluctuations, ordinary motion blur, subtle optical stabilization and imperfect human handling.
The recording vehicle contains four people: one driver, the main person recording from the front passenger seat, and two passengers in the back.
Only the front passenger operates the camera. The other occupants are heard naturally but do not need to appear on screen.
The person filming is not a professional cameraman. He notices things late, turns his phone instinctively, sometimes zooms too much, briefly loses framing and immediately corrects it.
Camera behavior must be motivated by human attention and surprise, never by preplanned cinematography.
The recording vehicle travels east in the left lane. The Tesla passes in the right lane and remains ahead-right. The iPhone films through the passenger-side window and forward windshield.
The railway crossing has fully lowered barriers blocking the entire usable roadway, visible warning lights and a moving freight train. Long queues of vehicles wait naturally in both approaching directions.
The location, traffic layout, train direction and Tesla route remain unchanged.
Human behavior and audio
The other passengers carry on ordinary background conversations. When something unusual happens, they independently react, interrupt one another, speak over each other, laugh nervously or fall briefly silent.
Voices should sound different according to their physical positions. The main cameraman is closest to the microphone. Rear-seat voices are softer and slightly muffled. The driver's voice comes from the left, and the driver remains attentive to the road.
Use one continuous live smartphone microphone perspective. Include HVAC hum, road and tire noise, passing engines, horns, train wheels, railway bells and conversations. Loud voices and thrust discharge cause realistic microphone compression and occasional slight distortion.
No studio dialogue, artificial crowd cheering, music or voiceover.
00:00–00:04 | THE TESLA APPEARS
The iPhone is casually filming ordinary traffic from inside the moving passenger car. The framing is imperfect, with part of the window frame visible and faint reflections across the glass.
Suddenly, the white Tesla Roadster accelerates past in the adjacent right lane.
Its rear three-quarter profile follows image2, the rear body matches image3 and image5, its spinning wheels match image11 and image12, and its taillight and wing match image14 and image15.
The recording passenger notices the car slightly late. His wrist turns quickly, the phone overshoots, and he corrects the framing to capture the Tesla from behind and slightly to the left.
The phone briefly zooms toward 1.5x as he tries to identify the vehicle.
Natural overlapping conversation begins.
Rear passenger, casually:
"What car is that?"
Another passenger:
"Wait, is that a Tesla?"
Main cameraman, recognizing it:
"Oh man, that's the new Tesla Roadster! The 2026 Roadster!"
He speaks while turning the phone, excited but not performing for the camera.
The Tesla pulls ahead aggressively with believable wheel rotation, road contact, suspension movement and electric motor whine.
00:04–00:09 | THE RAILWAY CROSSING AND TRAFFIC
The cameraman continues following the Tesla from the rear three-quarter angle as both vehicles approach the railway crossing.
The camera shifts naturally from the side window toward the front-right windshield area. The phone momentarily catches the A-pillar and window reflections before finding the car again.
Ahead, both eastbound lanes are congested with numerous waiting vehicles: sedans, SUVs, a bus, delivery trucks and ordinary city traffic.
The full-width railway barriers are already lowered. Red warning lights flash. A freight train moves across the road from north to south.
No gaps around the barrier system permit vehicles to bypass the crossing.
Traffic remains alive. Brake lights illuminate at slightly different moments. One vehicle creeps forward a short distance. A driver glances toward the train. Pedestrians wait behind the railway fencing.
The Tesla begins slowing as it approaches the congestion.
The cameraman adjusts the zoom slightly and briefly loses sharp focus as reflections cross the window.
Main cameraman, excited:
"Dude, look at that thing!"
A rear passenger responds with a short, indistinct comment. Someone chuckles quietly.
The voices overlap the train noise and railway alarm without becoming a conversation staged for the audience.
00:09–00:12 | THE ROADSTER BRAKES
The Tesla brakes decisively behind the stationary traffic in the right lane.
Its front suspension compresses naturally, brake lights illuminate and wheel rotation slows.
The recording vehicle brakes behind and slightly to the left.
The iPhone dips subtly from the deceleration. The cameraman tightens his grip and corrects the phone with a small upward movement.
One rear passenger leans forward slightly, audible through shifting clothing and a short breath.
The driver of the recording car stays focused on traffic.
The freight train continues moving. Crossing bells ring. The barriers remain down.
The Tesla does not wait unnecessarily.
A faint but unusual mechanical pressure-building sound becomes audible from its direction.
The cameraman holds the Tesla loosely framed in the center-right, unaware of what will happen next.
00:12–00:18 | THE SUDDEN HIGH-PRESSURE LIFT
Almost immediately, the Roadster activates its fictional high-pressure cold-gas lift system.
Four fixed downward-facing underbody outlets begin discharging powerful compressed gas. Their positions remain constant near the four wheel regions, mechanically separate from the four rear-facing outlets of image16. The lift-outlet locations are a cinematic concept, not verified production mounting positions.
Dense white condensation develops around the expanding gas, guided by image17 and image26.
The jets strike the asphalt and drive dust outward in turbulent ground-hugging sheets. Tiny condensed moisture droplets catch the late-afternoon light.
The sudden discharge creates a loud, harsh, high-pressure roar.
The sound happens first. The human reaction follows immediately.
The cameraman involuntarily jerks the phone sideways and upward, briefly losing the lower part of the Tesla.
Main cameraman, startled:
"Oh, fuck! What the hell?!"
Rear passenger, overlapping:
"Yo, what the hell?!"
Someone in the back gasps sharply.
The iPhone microphone compresses under the combined shouting and pressure noise.
The camera immediately tries to recover its framing, tilting upward too quickly before settling on the rising car.
Meanwhile, the Tesla's suspension unloads. The wheels extend, the tires gradually separate from the road, and the entire car rises decisively as one heavy rigid object.
The vehicle remains nearly level with small stabilization corrections. The downward thrust remains visibly anchored to the same outlets.
Nearby people begin noticing the event independently.
A driver in the adjacent queue turns toward the Tesla. Another lowers a window. A pedestrian near the crossing looks up after hearing the loud discharge.
Reactions are staggered and subtle. No synchronized crowd movement.
The cameraman instinctively zooms back out to keep the rising Tesla within the frame.
00:18–00:25 | THE ROADSTER FLIES OVER THE MOVING TRAIN
The Roadster continues rising above the roofs of the queued vehicles, with strong downward thrust and a stable forward-facing body.
Its underside clears the moving train's roofline with a clearly visible safety gap before crossing the tracks.
The train continues moving underneath, and both crossing barriers remain lowered.
The car flies forward eastbound, maintaining level attitude. Downward jets support the lift, while the separate image16 rear-facing outlets can provide forward thrust.
Image25 establishes the airborne crossing arrangement, image26 the level flight behavior, and image27 the coherent rear discharge appearance.
The vehicle geometry remains identical to image2, image3, image4, image5, image12, image14, image15 and image16.
The iPhone follows from the same stationary recording vehicle.
The cameraman tilts upward, momentarily overshoots, then corrects. As the car moves farther away, he zooms toward 2x, causing a brief exposure and framing adjustment.
The Tesla briefly approaches the upper edge of the screen before the cameraman recovers it.
Rear passenger, shocked:
"Bro, it's actually flying!"
Another passenger, partly overlapping:
"No way! No way!"
A short nervous laugh follows.
Main cameraman, raising his voice:
"Bro, that's insane! Did you see that?!"
The different voices overlap unpredictably. Some words are partly masked by the thrust and train noise.
Other people around the crossing now notice the flying vehicle.
One person inside a nearby stopped car points upward. Another passenger leans toward a window. A pedestrian raises a phone after seeing the Roadster lift.
A few drivers turn their heads, while others continue watching the railway crossing or remain inside their vehicles.
Nobody reacts in perfect synchronization.
The distant reactions remain visually small and partly obscured by real traffic. Any faint exterior shouting is masked by ambient sound.
The Tesla maintains physical separation from the train and barriers. Its moving shadow, forward momentum, fixed thruster locations and dissipating vapor remain coherent.
The train continues along the same tracks without changing direction.
00:25–00:30 | THE LANDING AND DISAPPEARANCE
After clearing the railway crossing, the Roadster reaches the clear eastbound section beyond the tracks.
The downward gas jets gradually reduce pressure.
The Tesla descends naturally with a nearly level body.
Its wheels contact the asphalt, the suspension compresses and settles, and normal tire-supported movement resumes.
The Roadster rapidly accelerates away between the Chicago buildings.
The cameraman struggles to keep it visible as its distance increases.
He zooms slightly, moves the phone toward the windshield, then corrects its angle after nearly losing the Roadster behind the edge of the frame.
Main cameraman, breathless:
"No way... no fucking way."
Rear passenger, speaking at the same time:
"Did you get that?"
A brief laugh and another indistinct excited comment overlap.
The driver remains focused on the blocked road and the active railway crossing.
Through the window, some nearby pedestrians and drivers are still looking toward the direction the Tesla flew.
The Roadster becomes smaller in the distance until it disappears between the buildings.
The cameraman holds the empty street in frame for a brief moment, still processing what happened.
The recording ends abruptly.
No title card, cinematic ending, artificial zoom or logo reveal.

ABSOLUTE REALISM AND CONTINUITY RULES
Smartphone camera
This is real-person handheld behavior, not deliberate camera choreography.
Maintain subtle continuous hand movement, real vehicle vibration, occasional grip changes, imperfect framing and minor rotational drift.
When something unexpected happens, camera movement follows the cameraman's perception with a natural reaction delay.
A sudden noise can cause a quick hand jerk; an airborne Tesla causes an upward tilt; increasing distance causes an imperfect zoom.
Modern phone stabilization absorbs much of the micro-shake but does not eliminate larger human movements.
Do not add constant exaggerated shaking. Maintain enough clarity to understand the Tesla, traffic, train and railway crossing.
Human reactions
Four people occupy the filming car, but only one records.
Keep voices distinct and spatially consistent. People react at slightly different times, interrupt one another and occasionally speak simultaneously.
The main cameraman is closest to the microphone; rear passengers sound farther away.
Do not make every reaction perfectly timed, perfectly articulated or equally loud.
Driver attention stays primarily on traffic and safety.
Background activity
The surrounding city remains active before, during and after the event.
Cars brake and creep forward naturally. Pedestrians wait near the crossing. Some drivers watch the train.
After propulsion begins, people notice the Tesla at different times. A few point or raise phones. Most remain inside their vehicles.
No staged crowd cheering, identical head turns, exaggerated gestures or unnecessary camera cutaways to bystanders.
Background reactions are visible only when they naturally fall within the passenger's phone view.
Vehicle and flight
The Roadster's exterior and interior design must match all original numbered references.
Keep the wheels, taillight, wing, rear diffuser, body panels, roof and branding unchanged.
Four fixed fictional underbody outlets provide downward thrust. The four established image16 rear-facing outlets remain separate.
The discharge creates realistic condensation and ground turbulence, not flames or fuel spray.
Suspension unloading precedes liftoff. The vehicle climbs, stabilizes, crosses above the moving train, descends, lands and accelerates away without discontinuity.
Railway and traffic
The same full-width barriers remain lowered throughout the train's passage.
Train direction, track geometry, warning lights, queued vehicle positions, road markings and surrounding architecture remain fixed.
No gaps around the crossing, disappearing traffic, reversing train, clipping objects or sudden changes in vehicle placement.
Live smartphone audio
Record all sound from one iPhone microphone inside the passenger car.
Maintain low-frequency cabin rumble, HVAC hum, road noise, ordinary traffic, train clatter, railway alarm and distant environmental activity.
The Tesla's propulsion becomes louder and more forceful as it activates nearby, then gradually fades as the car moves away.
Nearby speech can briefly overwhelm background sounds; loud thrust and shouting trigger phone microphone limiting or mild clipping.
Do not isolate voices cleanly or add artificial reverb.
No soundtrack, voiceover, studio-quality dialogue or cinematic impact effects.

GENERATION EXECUTION
Generate six connected video segments using consistent starting and ending frames.
Segment 1 | 00:00–00:04
The Tesla passes the recording car. Passenger notices it, turns the iPhone and identifies the Roadster. Other passengers comment naturally.
Segment 2 | 00:04–00:09
The cameraman follows the Tesla toward the railway crossing. Establish extensive traffic, lowered full-width barriers, passing train and ordinary background behavior.
Segment 3 | 00:09–00:12
Both vehicles brake. The recording phone dips with deceleration. Tesla prepares propulsion immediately, while the train remains in motion.
Segment 4 | 00:12–00:18
Cold-gas discharge begins. Camera jerks in reaction to the pressure noise. Tesla lifts, multiple passengers shout, and nearby people start noticing.
Segment 5 | 00:18–00:25
Tesla crosses above the moving train. The cameraman tilts, reframes and zooms while passengers react simultaneously. Background people independently react.
Segment 6 | 00:25–00:30
Tesla lands, accelerates away, and the passenger struggles to follow it. Excited voices continue as the recording ends.
For every segment, retain the preceding segment's final vehicle position, direction, velocity, rail-crossing status, camera orientation, passenger voices, lighting and background geography.
Treat the six segments as one uninterrupted recording from the same iPhone. Hide generation joins through continuity, not artificial camera transitions.
Final visual target: Genuine, accidental smartphone footage. Ordinary people witnessing an extraordinary fictional Roadster flight demonstration. Natural imperfections, believable human behavior, strong mechanical realism and a physically coherent environment.
The footage must feel observed rather than directed.

Tesla’s new Roadster just drove upside down and flew

After nearly a decade of missed deadlines, Elon Musk unveiled the next-generation Roadster at a SpaceX facility in Texas. In the livestreamed demo, the car climbed a ramp that curled overhead, clung to it upside down, launched off the end, rolled 180 degrees in midair and landed on its wheels. Tesla says a fan suction system underneath creates up to 2 Gs of downforce, and SpaceX thruster technology handles the rest. The airborne version will be invite-only, seat one person, and its flying features aren’t approved for public roads. Production isn’t planned until 2028 and no price was announced, and Musk himself told the crowd success is not guaranteed.

50 Tips To Save Video Credits In AI Filmmaking

Best AI Agents For AI Filmmaking

AI & Tech News

Meta says it now has the best American open weight model

Meta launched Muse Spark Open, an open weight model it claims matches and sometimes beats the top Chinese offerings. Alibaba’s Qwen, Moonshot AI’s Kimi K3 and Z.ai’s GLM have been seen as the global leaders in open weight models. Meta says the new model is designed for coding and agentic tasks, with strengths in long horizon work and complex software engineering. It arrived just days after Meta unveiled Wonder and Wonder Full, two closed frontier models it says rival the best of ChatGPT and Claude. Developers who want to run capable models on their own hardware now have a serious US option again.

Amazon enters wearable AI with Bee devices

Amazon has launched two Bee AI devices designed to listen throughout the day and help users recall conversations, tasks, and commitments. The $75 Frontier is a wrist wearable, while the $100 Eclipse attaches to a smartphone. The move signals Amazon’s push to make ambient AI assistants a mainstream consumer category, though it will likely renew debate over always-on audio privacy. _{^{Source: USAToday}}

Eight US states take OpenAI to court, saying ChatGPT is hurting kids

A bipartisan group of eight states, including Iowa, Kansas and Arizona, sued OpenAI in separate state courts. They allege ChatGPT was built to be agreeable and engaging so users would trust it, and that it has coached people, including teenagers, toward harming themselves or others. The states want civil penalties and damages, plus court orders forcing changes to how ChatGPT works. OpenAI says it already has protections for children and wants to work with states on policies that apply to the entire AI industry, not just one company. These are allegations, not rulings, but this is the biggest coordinated legal challenge yet to a consumer AI product, and it could shape safety rules for every chatbot. ^{_{Source: Associated Press via Fox10}}

Nvidia commits $250 million to UK AI safety

Nvidia pledged computing resources valued at $250 million over five years to support AI safety and security research in the United Kingdom. The commitment provides $50 million in annual compute access, with support directed to Oxford-led BOLD, UCL-led SOFAIR, and Imperial College London. It is a notable example of a major chipmaker funding independent research into agentic AI risks. _{^{Source: reuters}}

Google awards $36 million for AI science projects

Google is awarding $36 million to 17 organizations using AI to address challenges in climate, health, and other scientific fields. The funding comes through its Impact Challenge: AI for Science program. The initiative points to a growing shift in AI investment toward applied research with direct public and scientific benefits.
^{_{Source: Google}}

Google expands AI access for European researchers

Google DeepMind and Google Cloud will give up to 200 leading European researchers direct access to frontier models along with $3 million in compute credits. Researchers will also be able to use tools including AlphaFold 3, WeatherNext, AlphaGenome, AlphaEvolve, and AlphaEarth Foundations. The program could accelerate European work in biology, climate modeling, genomics, and earth science. ^{_{Source: Google blog}}

OpenAI launches an academic plan for education

OpenAI introduced the Rosalind Academic Plan, a new education-focused offering aimed at helping students learn through questioning rather than simply receiving answers.

The launch reflects a broader effort by AI companies to position their tools as learning aids while addressing concerns about overreliance and critical thinking. It may also shape how universities adopt AI in classrooms and research programs. ^{_{Source: OpenAI}}

Top AI tools to check out…

I’ve been building a best AI tool directory where you’ll find the best of the best tool and their USECASES.

Thanks for your time!

Shailesh & OpenAILearning team